| Competition Tribunal
|
|
Tribunal de la concurrence
|
PUBLIC VERSION
Citation: Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2022 Comp Trib 18
File No.: CT-2019-005
Registry Document No.: 296
IN THE MATTER OF an application by the Commissioner of Competition for one or more orders pursuant to section 92 of the Competition Act, RSC 1985, c C-34 as amended;
| BETWEEN:
(applicant)
and
(respondent)
|
|
Dates of hearing: January 6-7, 11-15, 19-21, and 25 and February 3-4, 2021
Before: D. Gascon J. (Chairperson), A.D. Little J. and Ms. R. Samrout
Date of Reasons for Order and Order: October 31, 2022
REASONS FOR ORDER AND ORDER
Table of Contents
C. The merger provisions of the Act
A. The Canadian grain industry
F. The Moosomin and Virden Elevators
A. Challenges to the Commissioner’s experts
(2) Objectivity of the Commissioner’s experts
B. The Commissioner’s duty of fairness and adverse inferences
(b) Discovery under the current CT Rules
(c) Disclosure and the Commissioner’s duty of fairness
(a) The “general” adverse inference
(b) The adverse inferences related to efficiencies under section 96
C. Legal and evidentiary burden applicable to sections 92 and 96 of the Act
A. What is or are the relevant product market(s) for the purposes of this proceeding?
(a) The purpose of market definition
(b) Rationale and tools for market definition
(c) The language of section 92
(a) The product and the price at issue
(b) The “value-added” approach
(c) The HMT framework and the SSNIP test
(4) Conclusion on relevant product market(s)
B. What is or are the relevant geographic market(s) for the purposes of this proceeding?
(b) Distance, transportation costs, and farms’ preferences
(e) Evidence related to prices and price negotiations
(f) Mr. Heimbecker’s testimony and P&H business records identifying competitors
(4) Conclusion on relevant geographic market(s)
(b) The “substantial lessening” analysis
(a) P&H’s alleged pre-existing market power
(c) Market concentration and market shares
(d) Removal of a vigorous and effective competitor
(e) Effective remaining competitors
(h) Barriers to entry and expansion
(k) Magnitude, duration, and scope of the anti-competitive effects
(4) Conclusion on substantial lessening of competition
(b) Evidentiary burden on efficiencies
(a) The anti-competitive effects of any lessening of competition
(3) Conclusion on efficiencies and section 96
A. Legal principles applicable to costs
Schedule “A” – Relevant provisions of the Act
Schedule “B” – List of Exhibits
I.
Executive Summary
[1] On December 19, 2019, the Commissioner of Competition (“Commissioner”) filed a Notice of Application (“Application”) against the Respondent Parrish & Heimbecker, Limited (“P&H”), pursuant to section 92 of the Competition Act, RSC 1985, c C-34 (“Act”), following the acquisition by P&H of 10 primary grain elevators (“Elevators”) located in Western Canada (“Transaction”). Prior to the Transaction, these 10 Elevators were owned and operated by Louis Dreyfus Company Canada ULC (“LDC”), one of P&H’s competitors in the grain business. In his Application, the Commissioner challenges the acquisition by P&H of one of these Elevators, namely, the LDC Elevator located on the Trans-Canada Highway in Virden, Manitoba (“Virden Elevator”), near the Manitoba-Saskatchewan border.
[2] The Commissioner claims that by acquiring the Virden Elevator (“Virden Acquisition” or “Acquisition”), P&H causes or is likely to cause a substantial reduction of competition in the supply of grain handling services (“GHS”) for wheat and canola for those farms that benefited from competition between the Virden Elevator and the nearby elevator owned by P&H and located in Moosomin, Saskatchewan, also on the Trans-Canada Highway (“Moosomin Elevator”). The Virden Acquisition is the only portion of the Transaction challenged by the Commissioner in this Application.
[3] The Commissioner’s Application alleges that the anti-competitive effects caused by the Virden Acquisition require a remedy under section 92 of the Act. The Commissioner submits that farms in the area which had previously benefited from the competition between P&H and LDC are likely to pay materially more to obtain GHS from the Moosomin and Virden Elevators, and will thus receive less money for their wheat and canola. The Commissioner maintains that canola crushing plants (“Crushers”) and more distant Elevators are not sufficient to constrain an exercise of market power by P&H, due to higher transportation costs for farms to deliver their grain to these competitors.
[4] P&H disputes the Commissioner’s position. P&H submits that the Commissioner’s Application improperly defines both the relevant product market and the relevant geographic market affected by the Virden Acquisition. According to P&H, the relevant product market is the purchase of wheat or canola from the farms, as P&H does not supply GHS. As to the relevant geographic market, P&H submits that it is much broader than the Commissioner alleges since the purchase prices set by the Moosomin and Virden Elevators are influenced by rival Elevators and Crushers located far beyond the respective individual draw areas of these two Elevators. P&H contends that in the face of vigorous and effective competition from competing Elevators, as well as from canola Crushers and other direct purchasers of wheat and canola, P&H’s control of the Virden Elevator gives it neither the ability nor the incentive to exercise monopsony power in any properly defined market. Hence, says P&H, the Virden Acquisition does not lessen competition substantially in any relevant market, and is not likely to do so. Moreover, P&H argues that in any event, the efficiencies that the Virden Acquisition is likely to bring about will be greater than, and will offset, the effects of any alleged lessening or prevention of competition.
[5] For the reasons that follow, the Tribunal will dismiss the Application brought by the Commissioner. The Commissioner has failed to establish, on a balance of probabilities, that the elements of section 92 have been satisfied.
[6] The Tribunal
[1]
first concludes that in the circumstances of this case, the relevant product is not the sale of GHS to farms, as alleged by the Commissioner, but the purchase of wheat and canola by P&H. The definition of the relevant product market was a fundamental point of disagreement between the parties, and was highly influential in the Tribunal’s overall analysis. The Tribunal finds that the Commissioner’s proposed product market is not grounded in commercial reality and in the evidence. Moreover, in this case, the “value-added” approach to product market definition advanced by the Commissioner fails on the facts, from a precedential and legal standpoint, and from a conceptual and economic perspective. Turning to the geographic market, the Tribunal is of the view that the relevant geographic market for the purchase of wheat is more likely than not to be comprised of at least the Virden, Moosomin, Fairlight, Whitewood, Oakner, Elva, and Shoal Lake Elevators. As to the relevant geographic market for the purchase of canola, it includes at least the Moosomin, Virden, Fairlight, Oakner, Whitewood, Brandon (Richardson), Melville, Souris East, Shoal Lake, and Elva Elevators, as well as the Crushers at Harrowby, Yorkton (LDC), Velva, and Yorkton (Richardson). These relevant markets are somewhat closer to the geographic markets proposed by the Commissioner but are larger than the narrow “corridor of concern” between the Moosomin and Virden Elevators that he originally identified (discussed below).
[7] The Tribunal also finds that the Commissioner has not established that the Acquisition lessens competition substantially in any relevant market, or is likely to do so. The Tribunal reaches that conclusion after finding that the Virden Acquisition does not materially reduce, and is not likely to reduce materially, the degree of price or non-price competition in the purchase of wheat and canola in the relevant geographic markets, relative to the degree that would likely have existed in the absence of the merger. In particular, the evidence shows that the price effects of the Acquisition are immaterial for the purchase of both wheat and canola, that several effective remaining competitors will remain to constrain P&H’s ability to exercise market power, and that the post-merger market shares are below the 35% safe harbour threshold. The Tribunal finds that the Virden Acquisition causes some lessening of competition for the purchase of wheat, but the evidence does not allow it to conclude that such lessening reaches the substantiality level required by section 92.
[8] In light of those conclusions, the Tribunal does not need to determine the issue of efficiencies claimed by P&H. However, considering the extensive submissions made by the parties on efficiencies and the nature of the issues raised, the Tribunal addresses the matter. The Tribunal concludes that P&H has not proven, with clear and convincing evidence, that the Virden Acquisition is likely to bring about cognizable gains in efficiency. As a result, P&H would not have met its burden of demonstrating, on a balance of probabilities, that its claimed gains in efficiency would be greater than, and would offset, the anti-competitive effects of any lessening of competition resulting from the Acquisition.
II.
Introduction and Overview
A.
The parties
[9] The Commissioner is the public official appointed by the Governor in Council under section 7 of the Act and is responsible for the enforcement and administration of the Act.
[10] P&H is a private, family-owned Canadian agribusiness headquartered in Winnipeg, Manitoba. P&H buys many varieties of grain, including wheat and canola, from farms and sells them to customers located in Canada, Europe, Asia, and South America. P&H has vertically integrated operations spanning across Canada in grain trading, handling, and merchandising, as well as in crop inputs retail, flour milling, and feed mills. It employs over 1,500 people with customers in 24 countries. Prior to the Transaction, P&H owned 19 Elevators in Western Canada. It also has ownership interests in a number of export terminals at Canadian ports located near Vancouver, British Columbia and in Thunder Bay, Ontario.
B.
The Transaction
[11] Pursuant to an asset purchase agreement dated September 3, 2019, P&H agreed to purchase from LDC 10 Elevators and related assets in Western Canada, including the Virden Elevator. On December 10, 2019, P&H and LDC closed the Transaction, bringing the total number of Elevators owned by P&H to 29. The grain volumes purchased through the former LDC Elevators in the last full crop year when they were owned and operated by LDC was 1.6 million metric tonnes (“MT”).
[12] The Transaction is part of P&H’s growth strategy. P&H claims that it will improve its efficiency and effectiveness in competing with other grain companies in Western Canada.
C.
The merger provisions of the Act
[13] A merger is defined by section 91 of the Act as referring to the acquisition or establishment, by one or more persons, of “control over or significant interest in the whole or a part of a business of a competitor, supplier, customer, or other person.” It is not disputed that the Transaction is a merger covered by the Act.
[14] Mergers, along with matters such as restrictive trade practices, are reviewable by the Tribunal under Part VIII of the Act if they have anti-competitive effects (Tervita Corp v Canada (Commissioner of Competition), 2015 SCC 3 (“Tervita SCC”) at para 43). With respect to mergers, section 92 identifies these anti-competitive effects as either substantially lessening competition or substantially preventing competition. More specifically, subsection 92(1) allows the Tribunal to intervene with respect to a merger or proposed merger if it finds that the merger prevents or lessens, or is likely to prevent or lessen, competition substantially “(a) in a trade, industry or profession; (b) among the sources from which a trade, industry or profession obtains a product; (c) among the outlets through which a trade, industry or profession disposes of a product; or (d) otherwise than as described in paragraphs (a) to (c).” The Tribunal is empowered to make a remedial order when a merger is found to either lessen or prevent competition substantially.
[15] Subsection 92(2) provides that the Tribunal shall not find that a merger or proposed merger prevents or lessens, or is likely to prevent or lessen, competition substantially “solely on the basis of evidence of concentration or market share.” However, the Tribunal has found that these two factors nonetheless may help in assessing whether or not a merger or proposed merger could result in a substantial lessening or prevention of competition (The Commissioner of Competition v CCS Corporation et al, 2012 Comp Trib 14 (“Tervita CT”) at para 360, rev’d 2013 FCA 28, rev’d 2015 SCC 3; The Commissioner of Competition v Superior Propane Inc, 2000 Comp Trib 15 (“Superior Propane I”) at paras 126, 304–313; Director of Investigation and Research v Hillsdown Holdings (Canada) Ltd (1992), 41 CPR (3d) 289 (Comp Trib) (“Hillsdown”) at pp 315–316, 318).
[16] Section 93 sets out a non-exhaustive list of market-specific factors that the Tribunal may consider in determining whether a merger prevents or lessens, or is likely to prevent or lessen, competition substantially. These factors include the following: foreign products as effective competition; failing firm considerations; availability of acceptable substitutes; removal of a vigorous and effective competitor; barriers to entry; remaining effective competitors; and change and innovation. The list is open-ended, as it includes at paragraph (h) “any other factor that is relevant to competition in a market that is or would be affected by the merger or proposed merger.”
[17] The Act also carves out certain exceptions to the application of the Tribunal’s section 92 remedial powers. One such exception, which is relevant in this case, is what is commonly named the “efficiencies defence,” in section 96 of the Act. This exception provides that the Tribunal shall not make an order under section 92 if it finds that the merger in respect of which the application is made is likely to bring about efficiency gains which are greater than and likely to offset the anti-competitive effects resulting from the merger.
[18] The Commissioner bears the burden of satisfying the elements of section 92, and the Tribunal must make a positive determination in respect of those elements before it may issue a remedial order. However, as will be discussed in more detail below, P&H bears most of the burden of proof under the efficiencies defence in section 96.
[19] The burden of proof is the civil standard, that is, the balance of probabilities. In that respect, the Tribunal remains guided by the principles established in FH v McDougall, 2008 SCC 53 (“McDougall”), where the Supreme Court of Canada (“SCC”) held that there is only one civil standard of proof in Canada, the balance of probabilities (see also Tervita SCC at para 66). Speaking for a unanimous court, Justice Rothstein stated in his reasons that the only legal rule in all cases is that “evidence must be scrutinized with care by the trial judge” and that “evidence must always be sufficiently clear, convincing and cogent to satisfy the balance of probabilities test” (McDougall at paras 45–46). In all civil cases, the trier of fact “must scrutinize the relevant evidence with care to determine whether it is more likely than not that an alleged event occurred” (McDougall at para 49).
[20] The full text of the relevant provisions of the Act is reproduced in Schedule “A” to these Reasons.
D.
The parties’ pleadings
[21] In his Application, the Commissioner seeks an order requiring P&H to divest either the Virden Elevator or the Moosomin Elevator, as well as an order prohibiting P&H from acquiring any Elevator in the relevant markets for a certain period of time.
[22] The Commissioner submits that the relevant product is the supply of GHS. According to the Commissioner, GHS includes the following services: the elevation, grading, and segregation of the grain performed by the Elevators, as well as the cleaning, drying, blending, and storage that may be offered. The Commissioner pleads that the relevant markets should be defined as “the supply of [GHS] for wheat and the supply of [GHS] for canola for the aggregated locations of farmers that benefited from competition between the Virden Elevator and Moosomin Elevator.” The Commissioner says that there are no functional substitutes for these GHS.
[23] Turning to the geographic dimension of the relevant markets, the Commissioner pleads that the wheat and canola purchased by an Elevator usually originate from nearby farms, and that the relevant geographic market is therefore local due to transportation costs, with the most affected farms being located in a narrow corridor between the Virden and Moosomin Elevators, within a one-hour drive of each Elevator.
[24] The Commissioner contends that the Virden Acquisition causes, or is likely to cause, a substantial lessening of competition in the relevant markets, due to the elimination of an important competitor
[2]
. The Commissioner alleges that, with the acquisition of the Virden Elevator, P&H can unilaterally exercise enhanced market power in the relevant markets, at the expense of farms located in certain parts of Saskatchewan and Manitoba. According to the Commissioner, P&H will be able to materially raise the implicit price that farms pay for GHS for wheat and canola in the Virden-Moosomin corridor, and farmers will be paid less for their wheat and canola.
[25] The Commissioner maintains that canola Crushers and more distant Elevators are not sufficient to constrain an exercise of market power by P&H owing to higher transportation costs for farms to deliver their grain.
[26] The Commissioner further claims that several section 93 factors support these conclusions, in that: 1) Elevators and direct purchasers in other countries cannot compete directly for the purchase of wheat and canola from farms in the relevant markets because of transportation costs; 2) for the vast majority of farms in the relevant markets, there are no viable substitutes; 3) barriers to entry and expansion are high, owing to significant capital costs and difficulty finding a suitable location to build an Elevator and accompanying access to rail transportation; 4) P&H no longer intends to expand the rail car capacity at the Moosomin Elevator, which would have increased this Elevator’s ability to handle more wheat and canola and the level of competition in the relevant markets; 5) the closest remaining Elevator to the Virden and Moosomin Elevators is an Elevator owned by Viterra Inc. (“Viterra”) in Fairlight, Saskatchewan (“Fairlight Elevator”), but it is insufficient to constrain an exercise of market power by P&H due to its location on a secondary road, 35 kilometers south of the Trans-Canada Highway; 6) the Virden Elevator, which has now been removed as a competitor, was previously a vigorous and effective competitor to P&H; and 7) the market for the delivery of GHS is not subject to material change through innovation.
[27] The Commissioner adds that, even if the relevant product markets were more broadly defined to be the purchase by Elevators of wheat and canola from farms, the Acquisition still causes, or is likely to cause, a substantial lessening of competition in these product markets due to P&H’s ability to materially decrease the price of wheat and canola paid to farms.
[28] P&H opposes the Commissioner’s Application and asks the Tribunal to dismiss it with costs. In P&H’s view, the Commissioner improperly defines both the relevant product market and the relevant geographic market. Furthermore, P&H submits that the Acquisition does not enable it to materially lower the prices it pays to farms for their wheat or canola, nor does it lead to a substantial lessening of competition in any relevant and properly defined market.
[29] P&H submits that the relevant product market is the purchase of wheat or canola. It states that, contrary to what the Commissioner advances, it does not supply GHS to farms.
[30] P&H argues that the prices it pays for grain at the Virden or Moosomin Elevators are largely dependent on global prices, which are independent of changes to the local competitive landscape around the Virden and Moosomin Elevators. According to P&H, the prices that it offers to pay farms for grain are centrally set: they are derived from the demand and prices it receives from its sales to customers in international and domestic markets, as well as by the costs to transport grain from its network of Elevators to export terminals or to domestic buyers.
[31] P&H also disagrees with the relevant geographic market as defined by the Commissioner. P&H maintains that Elevators purchase grain from farms located farther away than what the Commissioner alleges. P&H contends that the Virden and Moosomin Elevators each purchase grain from hundreds of farms mostly located outside the geographic area between these two Elevators along the Trans-Canada Highway, well beyond a one-hour drive. According to P&H, the Virden and Moosomin Elevators must purchase grain at competitive prices against many other rival Elevators whose draw areas extend farther than the narrow “corridor of concern” and the proposed geographic markets identified by the Commissioner. Therefore, in P&H’s view, the relevant geographic market is much broader than the Commissioner alleges since the purchase prices set by the Moosomin and Virden Elevators are influenced by rival Elevators located far beyond the respective individual draw areas of the two Elevators at issue. P&H claims that it does not hold or exercise monopsony power in a relevant geographic market as alleged by the Commissioner, or even in the broader area of Southeastern Saskatchewan and Southwestern Manitoba.
[32] P&H contends that in the face of vigorous and effective competition from competing Elevators, as well as from canola Crushers and other direct purchasers of wheat and canola, P&H’s control of the Virden Elevator gives it neither the ability nor the incentive to exercise monopsony power in any properly defined market. Rival Elevators and other purchasers within and beyond the draw areas of the Virden and Moosomin Elevators already purchase grain from farms that also sell to the Virden and Moosomin Elevators, have significant excess capacity to purchase additional grain, and can increase their purchases from those farms at low cost. In other words, says P&H, the Virden Acquisition will not substantially lessen competition.
[33] P&H further argues that barriers to entry and expansion are low, with the result that P&H’s ability to exercise any monopsony power would be constrained by the expansion of existing Elevators’ purchases and/or by new entry.
[34] Moreover, even if the Virden Acquisition were found to substantially lessen competition, P&H argues that the gains in efficiency that the Acquisition is likely to bring about will be greater than, and will offset, the effects of any alleged lessening of competition. According to P&H, it will not likely attain such gains in efficiency if the Tribunal makes the orders sought by the Commissioner. The efficiencies claimed by P&H from the Acquisition include the following: improved scale economies and cost savings at the Fraser Grain Terminal (“FGT”) located in British Columbia; elimination of the margin that LDC formerly paid to use the Vancouver export terminal owned by Kinder Morgan; output expansion and improved scale economies at the Virden Elevator; and administrative synergies.
[35] In his reply, the Commissioner opposes P&H on this last point and submits that the Virden Acquisition will not generate cognizable gains in efficiencies to the extent alleged by P&H. The Commissioner further maintains that, if the Tribunal makes the orders sought, P&H’s ability to achieve the alleged efficiencies being claimed would not be impacted. In any event, the Commissioner holds that any cognizable efficiencies that P&H may obtain through the Virden Acquisition and that would be lost if the orders sought were made will not be greater than or offset the anti-competitive effects of the Acquisition.
E.
Procedural history
[36] Around the time the Commissioner filed the Application in December 2019, he stated that he would request an expedited scheduling order in accordance with the Tribunal’s Practice Direction regarding an Expedited Proceeding Process before the Tribunal, dated January 2019. Under an expedited scheduling order, an application will typically be heard by the Tribunal within five to six months after the filing of the notice of application.
[37] P&H opposed the Commissioner’s request and asserted that procedural fairness concerns would arise under an expedited process. P&H proposed an alternative schedule pursuant to which the hearing would take place approximately three to four months later than the hearing dates contemplated under the expedited process.
[38] On January 13, 2020, the Tribunal denied the Commissioner’s request for an expedited process (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2020 Comp Trib 1). The Tribunal was not persuaded that in the absence of P&H’s consent, the expedited process was a reasonable option given the circumstances and fairness considerations arising in this case. Moreover, the period of three to four months that could be gained with the expedited process did not justify the imposition of the process over P&H’s strong objections. The Tribunal adopted the alternative schedule proposed by P&H and issued a scheduling order in early March 2020, pursuant to which the hearing of the Commissioner’s Application would start in November 2020 (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2020 Comp Trib 2 (“Scheduling Order”)). Adjustments were subsequently made to various steps of the Scheduling Order as a result of the COVID-19 pandemic. The parties nonetheless continued to work towards the November 2020 hearing dates.
[39] In October 2020, P&H advised the Tribunal that its expert was no longer available in November because of unforeseen personal circumstances. The Tribunal agreed to adjourn the hearing with the consent of both parties. Eventually, the Tribunal issued an amended Scheduling Order, pursuant to which the hearing would now proceed in early January 2021 (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2020 Comp Trib 13).
[40] In the course of the proceedings leading up to the hearing, counsel for P&H insisted on various occasions on an in-person hearing notwithstanding the COVID-19 pandemic and the implementation of various lockdowns. While counsel for the Commissioner initially accommodated P&H’s request and agreed to a hybrid hearing, the Commissioner eventually opposed the request as the pandemic worsened. In December 2020 and early January 2021, the Tribunal ordered that the hearing would take place remotely by way of videoconference using the Zoom platform (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2020 Comp Trib 14; Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2021 Comp Trib 1).
[41] In anticipation of the hearing, the parties exchanged witness statements in accordance with the schedule fixed by the Tribunal. These witness statements included statements from farmers in Western Canada, as well as initial and reply witness statements by John Heimbecker, the Chief Executive Officer (“CEO”) of P&H.
[42] On November 27, 2020, the Commissioner moved to strike some paragraphs of the initial witness statement of Mr. Heimbecker on the basis that it contained inadmissible hearsay and inadmissible lay opinion evidence. In December 2020, the Tribunal granted this motion in part and ordered P&H to prepare a revised witness statement from Mr. Heimbecker (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2020 Comp Trib 15 (“Parrish & Heimbecker”)).
[43] Initially, both parties agreed to designate the identity of their respective farmer witnesses as Confidential Level B in accordance with the Confidentiality Order issued by the Tribunal (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2020 Comp Trib 3). As the hearing approached, however, P&H revised its position. By way of letter, P&H advised the Tribunal at the end of November 2020 that witness statements prepared by three farmers on behalf of P&H would no longer be designated confidential. Moreover, P&H expressed doubts about the merits of the Commissioner’s confidentiality designations and eventually asserted that the Commissioner should file a formal motion to designate as confidential the identities of his farmer witnesses. On December 7, 2020, the Commissioner moved for an order designating the identities of five farmers as confidential.
[44] On December 29, 2020, the Tribunal dismissed the Commissioner’s motion and reasons for this decision were issued in early January 2021 (Canada (Commissioner of Competition) v Parrish & Heimbecker, Limited, 2021 Comp Trib 2). The Tribunal found that the Commissioner had failed to present clear and convincing evidence sufficient to satisfy the Tribunal that the requirements for the confidentiality designations were met. Further to that decision, only three of the five farmer witnesses originally identified by the Commissioner appeared at the hearing in a public setting.
[45] The hearing was held virtually between January 6 and February 4, 2021, and the witnesses testified by videoconference in accordance with a witness protocol that was developed by the Tribunal with the parties’ input.
[46] Not only was this the first virtual hearing for the Tribunal, but this was also the first time that experts testified together as part of a panel of expert witnesses formed in accordance with Rules 75 and 76 of the Competition Tribunal Rules, SOR/2008-141 (“CT Rules”). CT Rule 76 provides that the Tribunal “shall direct the manner in which the panel [of witnesses] shall testify” and that counsel can cross-examine and re-examine the witnesses. The protocol for this concurrent expert evidence session (also known as “hot-tubbing”) was set out in a specific Direction issued by the Tribunal with the parties’ consent.
[47] The purpose of this “hot-tubbing” process is to streamline the testimonies of expert witnesses, and to allow experts to ask questions from each other and highlight their areas of agreement and disagreement. Pursuant to the Tribunal’s Direction, the experts and counsel for the parties agreed on a list of five main issues to be addressed by the experts at the concurrent evidence session, and the experts identified their areas of agreement and disagreement on each issue. The parties also exchanged short statements of each expert’s proposed expertise. Each expert was granted a full and fair opportunity to present and explain their respective position on each issue, and opposing counsel were able to cross-examine the experts. A significant benefit that flowed from this concurrent evidence session was that experts were able to rapidly focus on the key areas of disagreement between them. In the view of all members of the Tribunal, the process worked well and helped the Tribunal to have a solid understanding of each expert’s position, while allowing the Tribunal and the parties to narrow the disputed issues between the experts.
III.
Factual Background
A.
The Canadian grain industry
[48] The grain supply chain in Canada involves an interconnected network of businesses and infrastructure that moves grain from individual farms to end customers, such as companies that manufacture food or feeds. The main participants include farmers who produce grain, grain companies that purchase grain from farmers, railways that transport grain from Elevators to export terminals or to domestic customers, and export terminals where the grain is delivered for storage and shipping.
[49] Canadian farmers grow a variety of grains such as wheat, barley, soybeans, peas, and canola. The Commissioner’s Application in this case focuses solely on two types of grain, namely, wheat and canola. Wheat and canola are both commodity products.
[50] Farmers can sell their wheat to Elevators, and their canola to Elevators or Crushers. For many years before 2012, when the Canadian Wheat Board (“CWB”) was in existence, grain companies bought wheat and barley on behalf of the CWB on a toll basis. At the time, the CWB was, by law, the sole marketer of wheat and barley for export and domestic human consumption. Grain companies then acted as the agents of and service providers to the CWB. Grain companies purchased other grains such as canola directly from farmers, without the intervention of the CWB. However, in August 2012, the CWB’s role ended and grain companies ceased being service providers to the CWB. The grain companies now purchase and sell wheat and barley from farmers on their own account for sale to their own customers, as they do for other types of grain. With the end of the CWB’s role as a sole purchaser of certain grain, the historical tariffs and fees that had been in place for the service provided by grain companies ended. But, as will be discussed below, the heritage from the CWB days has an impact on certain purchasing and selling practices in the grain business.
[51] Canadian grain companies sell grain domestically or to overseas customers by transporting it by rail to export terminals located at Canadian ports. At the export terminals (and at some local Elevators), grain is segregated by type and quality attributes, stored, blended, and loaded onto vessels.
[52] In addition to P&H, there are several major grain companies that purchase wheat and canola in competition with P&H in Western Canada. The two largest are Viterra and Richardson International Limited (“Richardson”). Viterra is a privately-held subsidiary of Glencore, a British-Swiss multinational corporation; it has 79 Elevators and six port facilities across Canada and parts of the United States (“U.S.”). Richardson is a privately-held Canadian subsidiary of James Richardson & Sons, Limited which owns 73 grain Elevators and has ownership or partnership interests in the largest three grain terminals in Canada.
[53] Other major grain companies operating in Western Canada include Cargill Limited (“Cargill”), Paterson Grain Limited (“Paterson”), Ceres Global Ag Corp. (“Ceres”), Bunge Ltd (“Bunge”), Archer-Daniels-Midland Limited (“ADM”), and G3 Canada Limited (“G3”). Cargill is a vertically-integrated company with 31 Elevators and port terminals across Canada. Paterson operates more than 40 Elevators whereas G3 has 17 Elevators and four export terminals.
[54] In addition to these major players, other local grain companies such as GrainsConnect Canada also compete in Western Canada.
B.
Elevators and Crushers
[55] Elevators are designed to stockpile and store the grain they purchase from spatially dispersed farms. The Elevators, upon receiving the grain from a farm, will grade it, elevate it, and segregate it; they may also clean, blend, dry, and store the grain at the Elevator until a railcar or a truck comes to take the grain to its next destination. This is what the Commissioner refers to as GHS. Elevators’ staff will typically examine grain samples from the farms’ trucks, assess for dockage as needed, grade the grain, unload the trucks delivering the grain, elevate the grain to the appropriate storage bins, store the grain and keep it in condition, blend the grain as appropriate, assist with weighover (i.e., inventory counts), dry the grain as needed, prepare cash settlements for farms, and load the grain into railcars for shipment to a port terminal or to a further processing mill such as flour mills. Grain companies incur costs for those activities, such as costs related to any cleaning or drying, transportation from Elevators to export terminals or domestic locations, developing export and domestic customers, and managing risk with respect to fluctuations in exchange rates and commodity prices.
[56] Grain is graded in accordance with the Official Grain Grading Guide published by the Canadian Grain Commission (“CGC”). A grading factor is a physical condition of grain that indicates a certain quality level. For wheat, the highest quality grade under the CGC’s classification system is grade 1 Canadian Western Red Spring Wheat (“1CWRS”). Turning to canola, the most common grade for harvested canola is 1CAN CANOLA. In the case of wheat, the protein content also affects the price. The base protein content commonly used by grain companies is 13.5%, and a higher protein wheat commands a higher price relative to 1CWRS 13.5. Protein spreads reflect the cash price adjustments (either up or down from the cash price for 1CWRS 13.5) based on the protein content of the wheat.
[57] Elevators have varying grain storage capacities. The storage capacity of P&H’s Elevators ranges from 22,000 MT at the Glossop Elevator (located in Glossop, Manitoba) to 106,000 MT at the Weyburn Elevator (located in Weyburn, Saskatchewan).
[58] Elevators are often located close to railways, as the grain is typically loaded onto railcars and transported by rail. The term “rail car spots” is commonly used within the industry and refers to the number of railcars at an Elevator that can be accommodated for loading on a sidetrack (or spur line) off the main track line.
C.
Farms
[59] Even though some farms will have storage and elevating capacity, farms typically rely on Elevators as they could not achieve the same efficiencies in moving their grain from the farm to the domestic customers or to export terminals for delivery to international end customers. Farms can sell their wheat and canola to multiple grain companies and are offered prices by Elevators and Crushers for their grain.
[60] In most instances, farms are responsible for hauling their grain to the Elevators. Some farms have their own trucks to transport their grain, while others employ commercial trucking companies to load, ship, and unload their grain. In certain circumstances, some Elevators or Crushers might offer pick-up service, which is charged to farms through a trucking allowance.
[61] The transportation costs incurred by farms to bring their grain to an Elevator will vary with distance but also with travel time, road conditions, and seasonal road weight restrictions that may affect certain secondary roads. All else being equal, most farms prefer to sell their grain to closer Elevators.
D.
Pricing and contracts
[62] Grain companies such as P&H buy wheat or canola at their Elevators by paying farms a “net” or “cash” price for their grain (“Cash Price” or “CP”). The Cash Price is also sometimes referred to as the “flat” or “bid” price for the grain. No matter how it is worded or expressed, the Cash Price represents the actual amount of money (per MT or per bushel) received by a farm for the net quantity of grain delivered and sold at an Elevator. P&H posts its Cash Price for grain for each of its Elevators. Farms can also use P&H’s mobile application, named “P&H Direct,” to see the Cash Prices at each of P&H’s Elevators across Western Canada.
[63] The price of grain can be expressed in terms of dollars per MT or dollars per bushel. There are 36.744 bushels of wheat to the MT and 44.092 bushels of canola to the MT.
[64] The Cash Price that a farmer receives for grain is comprised of two components: the futures price (“Futures Price” or “FP”) and what is commonly known in the grain industry as the “basis.” The term “basis” refers to the difference between the Futures Price and the Cash Price (“Basis” or “B”)
[3]
.
[65] The Futures Price reflects the global commodity market price for the grain, set by global supply and demand forces. Neither the farms nor the Elevators have control over the Futures Prices, as these are global commodity prices. The world Futures Prices for wheat and canola are determinative of P&H’s prices for those commodities. For wheat, P&H uses the Minneapolis Hard Red Spring wheat futures contract price for CWRS. This price trades in U.S. dollars (“USD”) per MT. For canola, P&H uses the Intercontinental Exchange futures price for canola in Saskatchewan. This price trades in Canadian dollars (“CAD”) per MT. Grain companies (including P&H) typically use 1CWRS as their base grade for wheat pricing and 1CAN CANOLA as their base grade for canola pricing.
[66] While both the Commissioner and P&H agree that the Cash Price, the Futures Price and the Basis are the three components of the pricing process for grain, they fundamentally disagree on the interrelation between these three components. The Commissioner claims that P&H has no control over the Futures Price and sets the Basis, and that the Cash Price paid to farms is the resulting amount. In other words, the Commissioner argues that FP - B = CP. P&H instead argues that the Basis numerically results from the difference between the Cash Price it sets and the Futures Price over which it has no control. In sum, P&H submits that FP - CP = B. The Commissioner claims that the relevant price for the purposes of a competition analysis is the price for GHS — which, he says, equates to the Basis —, whereas P&H is of the view that the relevant price is the Cash Price effectively paid to the farms.
[67] Farms can sell and deliver their grain at different times throughout the year and they can sell a portion of their crop before it is harvested. Some farms can store some or all of their grain on their farm if they have the proper elevating capacity, which allows them to sell their grain at a time of their choosing.
[68] The Cash Price ultimately received by the farms can sometimes be adjusted upwards when Elevators offer limited-tonne or limited-time pricing “specials” to fill remaining space in a train or a vessel or to obtain additional grain supplies to meet sales commitments. From time to time, the Cash Price or the Basis can also be adjusted to reflect individual negotiations between farms and the Elevators. P&H estimates that this occurs in approximately XXX of its grain purchase transactions.
[69] In terms of contracts with Elevators and Crushers, farms can enter into different types of agreements to sell their grain. They can enter into fixed price contracts, grain pricing order agreements (“GPOs”) — also known as grain purchase orders or target contracts —, and basis contracts.
[70] Under a fixed price contract, the Cash Price, Futures Price and Basis are fixed. Similarly, the quantity and quality of grain to be delivered, as well as the delivery period, are determined in the fixed price contract. Fixed price contracts are used for forward or deferred delivery purchase transactions as well as for spot purchase transactions. Forward or deferred delivery refers to a delivery of grain at some point in the future. Farms can enter into forward or deferred delivery contracts to deliver a specific quantity and quality of grain to an Elevator for an agreed Cash Price within a prescribed delivery window in the future. In P&H’s fixed price contracts, the Cash Price appears as the “net” price.
[71] Under a GPO, a farm sets a targeted Cash Price above an Elevator’s posted Cash Price (“Target Cash Price”) at which the farm agrees to sell and deliver to that Elevator a specific type of grain in a specified delivery month. If the Elevator’s posted Cash Price reaches a farm’s Target Cash Price, the GPO is triggered and the Elevator must purchase the farm’s grain at the Target Cash Price. If a GPO is triggered, it becomes a fixed price contract. Farms always keep the option to amend or cancel a GPO at any time before it is triggered. A farm chooses the expiry date for the GPO, which may be in effect for days, weeks, or months.
[72] The third type of agreement that farms can enter into is a basis contract. Under such a contract, the Basis is agreed upon and fixed in the contract, but the Futures Price for contracting purposes is taken from the international markets and fixed by the farms’ actions at a later date. Such agreements allow farms to lock in what they consider to be a favorable Basis. Under a basis contract, the quantity and quality of grain to be delivered, as well as the delivery period, are set, but the Cash Price is determined once the farm triggers the basis contract, which sets the Futures Price.
[73] When P&H buys wheat or canola from a farm, it takes title to the grain at the time the farm delivers the grain to the Elevator. At that point in time, the farm receives the contracted Cash Price for its grain and ownership of the grain then passes to P&H. The Cash Price may be adjusted at the time of delivery of the grain to the Elevator if the quality of the wheat or canola delivered is different from the quality the parties had agreed upon in the contract.
E.
P&H’s business
[74] P&H operates within the grain business by buying and selling grain for its own account throughout the crop year, which spans from August 1 to July 31 of the following year.
[75] P&H buys wheat and canola from farmers via a network of 29 Elevators located throughout Western Canada, including the Moosomin Elevator and the 10 Elevators purchased from LDC in December 2019. P&H’s 29 Elevators are the entry points to its grain network in Western Canada.
[76] P&H sells the varieties of grain it purchases, such as wheat and canola, to customers located in Canada, Europe, Asia, and South America. Just over half of P&H’s total wheat and canola sales are for export. P&H’s export customers pay for wheat and canola at the Canadian port.
[77] In order to move the wheat and canola it sells to its customers located overseas, P&H utilizes the rail network to ship grain from its Elevators in Western Canada to its export terminals located on the West Coast and in Thunder Bay, Ontario.
[78] P&H has an interest in three export terminals located near Vancouver in British Columbia, namely, the Alliance Grain Terminal (“AGT”), the Fraser Surrey Docks (“FSD”), and the FGT, where P&H has recently invested XXXXXXXXXXx. P&H also has an interest in the Superior terminal located in the port of Thunder Bay in Ontario (“Superior”). The vast majority of grain exported by P&H moves through its export terminals. The storage capacities are 102,000 MT at AGT, 18,000 MT at FSD, 176,000 MT at Superior, and 92,000 MT at FGT, where P&H has a partial entitlement to storage and throughput capacity.
[79] Export terminals are used to receive grain from rail, segregate and store grain by type and quality attributes, clean grain when required, blend grain, and load grain onto vessels. As with other commodities, wheat and canola of the same grade received from different P&H Elevators are commingled at the terminals. The cleaning and blending of grain occur principally at P&H’s export terminals, rather than at its Elevators, given the greater economies of scale available at these terminals.
[80] P&H also operates a milling group that sources Canadian wheat to produce flour and cereal products. P&H moves the wheat supplied to its milling group by rail or truck from its Elevators to its mills in Western and Eastern Canada.
[81] Additionally, P&H operates a Crop Inputs and Services business, which supplies fertilizer, seed, and pesticides as well as agronomic services to farms through dual crop inputs and grain facilities at its Elevators across Canada. P&H has a “one-stop-shop” crop inputs retail and grain purchase business model. The former LDC Elevators purchased by P&H did not offer crop inputs services.
[82] P&H’s audited consolidated financial statements for the 2018 fiscal year indicate that, across all of its lines of businesses, it generated consolidated revenues of approximately XXX XXXXXXXXX and gross profit of XXXXXXXXXXX. By comparison, P&H reported consolidated revenues of approximately XXXXXXXXXXXXX and gross profit of XXX XXXXXXXX for the 2017 fiscal year.
[83] In March of every year, P&H sets its annual grain-purchasing budget for Western Canada for the upcoming fiscal year, which begins on May 1 of each year. Its grain purchase targets aim to increase P&H’s total grain volumes and share over time.
F.
The Moosomin and Virden Elevators
[84] Prior to the Transaction, P&H and LDC respectively owned and operated the Moosomin Elevator and the Virden Elevator, located in proximity to one another near the Manitoba-Saskatchewan border. Then, LDC would send grain from the Virden Elevator westward by rail to its export terminals on the West Coast. Following the Transaction, these two Elevators were re-assigned to P&H’s Thunder Bay catchment area, meaning that the grain purchased by these Elevators is shipped to the Superior terminal in Thunder Bay. However, the Moosomin Elevator, which is located west of the Virden Elevator, is also in a position to ship grain to P&H’s West Coast terminals.
[85] For rail transportation, the Moosomin Elevator has 56-car spots while the Virden Elevator has 112-car spots. In terms of storage capacity, the Moosomin Elevator has a capacity of 26,000 MT and an annual throughput capacity in the range of XXXXxx MT. For its part, the Virden Elevator has a storage capacity of 46,000 MT and an annual throughput capacity in the same range of XXXXxx MT.
IV.
Evidence – Overview
[86] Over the course of the hearing, the Tribunal heard from 16 lay witnesses and three expert witnesses. Over 250 exhibits were filed.
A.
Fact witnesses
(1)
The Commissioner
[87] The Commissioner led evidence from three farmer witnesses located in Manitoba or Saskatchewan, namely:
· Alistair Pethick: Mr. Pethick and his brother operate a farm located in McAuley, Manitoba. They mainly grow wheat and canola, but also soybeans, oats, and hay as well as other speciality crops in some years. Mr. Pethick sold his wheat to the Moosomin, Virden, and Fairlight Elevators, as well as to the Ceres Elevator located in Northgate, Manitoba;
· Chris Lincoln: Mr. Lincoln and his family own and operate two farms located in Maryfield and Wawota, Saskatchewan. They grow wheat and canola. Mr. Lincoln’s farms have the capacity to store 80-85% of his grain. The Fairlight Elevator operated by Viterra is the closest Elevator to Mr. Lincoln’s farms. Since harvesting his crops in November 2019, Mr. Lincoln has sold all his crop to the Fairlight Elevator. In 2018, he sold 20% of his commodity crop to the Virden Elevator and the balance to the Fairlight Elevator; and
· Ian Wagstaff: Mr. Wagstaff owns a 6,000-acre farm approximately two miles south of Manson, Manitoba. He is a wheat and canola farmer. He harvests approximately 100,000 bushels of wheat and canola per year. Mr. Wagstaff can store 60,000 to 70,000 bushels of wheat at his farm, meaning that he must sell approximately 25-30% of his crop at harvest time. In the past two years, he has sold most of his crop to the Virden Elevator.
[88] The Commissioner had two other farmer witnesses, XXXXXXXXx and XXXXXx, who decided not to testify in public at the hearing. However, the parties filed an agreed statement of facts regarding the testimonies of these two farmer witnesses.
[89] The Commissioner also led evidence from Harvey Brooks, who is the General Manager of the Saskatchewan Wheat Development Commission (“Sask Wheat”). Sask Wheat is a producer-led organization established to grow Saskatchewan’s wheat industry through research, market development, and advocacy. Mr. Brooks has been General Manager of Sask Wheat since 2014. Prior to joining Sask Wheat, Mr. Brooks served as Deputy Minister of Agriculture for the Government of Saskatchewan, Director of Policy and Economic research with the Saskatchewan Wheat Pool, and Head of Corporate Policy at the CWB. He holds a Ph.D. in Economics from Iowa State University and a Masters degree in Agricultural Economics from the University of Saskatchewan.
[90] Eight representatives of grain companies other than P&H also testified before the Tribunal for the Commissioner. These companies had provided data to the Commissioner during his investigation of the Transaction. These witnesses were:
· Dean McQueen: Mr. McQueen is the Vice President, Grain Merchandising and Transportation (North America) at Viterra. Viterra markets and handles grain, oilseeds, and pulses. It operates grain elevators and special crop facilities, port terminals, and processing facilities. Mr. McQueen is responsible for overseeing the merchandising and transportation of grain, oilseeds, and pulses, including procurement, through the Viterra country grain Elevator network;
· Ray Elliot: Mr. Elliot is a Manager for Seed Procurement at Bunge’s Harrowby Crusher facility located in Russell, Manitoba. Bunge is an agribusiness and food company that buys oilseeds and softseeds from producers and sells finished products to customers. Mr. Elliot is responsible for managing all the seed purchases for Bunge’s crushing plants in Western Canada;
· Brett Malkoske: Mr. Malkoske is the Chief Financial Officer of G3. He previously was the Vice President of Business Development and Communications at G3, where he was responsible for external communications and facilitating the development and execution of G3’s strategic plans in Canada;
· Darcy Jordan: Ms. Jordan has been a Management Accounting and Reporting Senior Analyst at Cargill since 2019. Cargill is a merchandiser and processor involved in crop inputs product retailing, grain handling, milling, salt distribution, and merchandising. In her role, Ms. Jordan is responsible for Cargill’s management reporting, supporting the Manitoba region for margins, and implementing the controls framework and profit and loss statements;
· Kara Hawryluk: Ms. Hawryluk is the Canada Operational Controller at LDC. Along with its parent company Louis Dreyfus Company B.V., LDC is a global merchant and processor of agricultural goods. Ms. Hawryluk is responsible for working with LDC’s commercial and operational teams to ensure timely and accurate reporting of Elevator and trading information;
· Jeff Wildeman: Mr. Wildeman is the Origination and Supply Chain Solutions Manager at Ceres. Ceres is involved in the procurement and provision of North American agricultural commodities, industrial products, fertilizers, energy products, and supply chain logistics services. Mr. Wildeman is responsible for the origination of Canadian agricultural commodities for Ceres’s grain merchandising operations;
· Mark Irons: Mr. Irons is the Vice-President, Softseed Crush for ADM, an American global food processing and commodities trading corporation. Mr. Irons oversees the management of commercial activities related to ADM’s softseed crush assets in North America; and
· Bryce Geddes: Mr. Geddes is a Marketing Specialist at Richardson, a worldwide handler and merchandiser of major grains and oilseeds, and a vertically integrated processor and manufacturer of oats and canola-based products. Mr. Geddes is responsible for collecting and analyzing transactional data for Western Canadian markets in which Richardson conducts its grain and crop inputs businesses.
[91] The Tribunal notes that the Commissioner obtained data from nine grain handling companies including 15 Elevators and five Crushers. This data was used in the preparation of the expert evidence filed by the Commissioner.
[92] The Tribunal generally found the Commissioner’s farmer witnesses and Mr. Brooks to be credible, forthright, helpful, and impartial. They were knowledgeable about their respective businesses and farm operations. With respect to the representatives of competing grain companies and Crushers, the Tribunal found that these witnesses were reliable and gave no reasons to doubt the accuracy of the transaction data they provided.
(2)
P&H
[93] Turning to P&H, it led evidence from the following three farmer witnesses, who are all based within the Commissioner’s proposed geographic market and his narrower “corridor of concern” in Manitoba and Saskatchewan:
· Kristjan Hebert: Mr. Hebert owns a 22,000-acre farm located in Fairlight, Saskatchewan, which is operated through Hebert Grain Ventures. Mr. Hebert grows wheat and canola as well as malt barley, hybrid rye, and yellow peas;
· Tim Duncan: Mr. Duncan owns and operates an approximately 3,000-acre farm located west of Cromer, Manitoba. He grows wheat, canola, and oats. From year-to-year, he will also grow barley, peas, and/or soybeans; and
· Edward Paull: Mr. Paull owns and operates an approximately 3,400-acre farm located 4.5 miles outside of Elkhorn, Manitoba, a town located between the Moosomin and Virden Elevators. He grows wheat and canola every year.
[94] Mr. Heimbecker, the CEO of P&H, also testified at the hearing and was the only witness representing P&H itself. In addition to being CEO, Mr. Heimbecker is the President of P&H’s Grain Division Canada. Mr. Heimbecker has been at P&H and in the grain business for his entire professional career, which started in May 1987. He was named CEO of P&H in September 2019. As President of Grain Division Canada, he is in charge of P&H’s grain business for all of Canada. Mr. Heimbecker also acted as P&H’s main witness on the issue of efficiencies.
[95] As was the case for the Commissioner’s farmer witnesses, the Tribunal generally found P&H’s farmer witnesses to be credible, forthcoming, helpful, and impartial. As to Mr. Heimbecker, the Tribunal also found him forthcoming and knowledgeable about P&H’s business. The Tribunal, however, observes that Mr. Heimbecker was not close to the day-to-day operations of P&H’s Elevators, and was of more limited assistance to the panel in this respect. In addition, some of his evidence was distinctly oriented towards a successful outcome for P&H in this proceeding and was therefore less helpful to the Tribunal in such instances.
B.
Expert witnesses
[96] Three expert witnesses provided expert reports and testified at the hearing.
(1)
The Commissioner
[97] Dr. Nathan Miller and Mr. Andrew Harington testified on behalf of the Commissioner.
(a)
Dr. Miller
[98] Dr. Miller is the Saleh Romeih Associate Professor at the McDonough School of Business at Georgetown University in Washington, DC. He holds a B.A. in Economics and History from the University of Virginia and a Ph.D. in Economics from the University of California at Berkeley. He served as a Visiting Professor at Toulouse School of Economics in 2019-2020. Prior to joining Georgetown University in 2013, he served as a Staff Economist in the U.S. Department of Justice from 2008 to 2013. Dr. Miller’s area of expertise is in industrial organization, with a specialization in antitrust economics and a focus on collusion and the competitive effects of mergers.
[99] The Commissioner asked Dr. Miller to prepare a report examining the competitive effects and the deadweight loss (“DWL”), if any, with respect to the acquisition of grain Elevators and related assets from LDC by P&H (namely, the Transaction). His report focused specifically on the Virden Acquisition. Dr. Miller was also asked to reply to the report filed by P&H’s expert, Ms. Margaret Sanderson, in response to his initial expert report.
[100] With the parties’ agreement, Dr. Miller was accepted as an expert qualified to give opinion evidence in industrial organization and competition law economics. The Tribunal generally found Dr. Miller to be credible, forthright, objective, and impartial. Dr. Miller was a cooperative witness and explained his models and analyses with clarity.
(b)
Mr. Harington
[101] Mr. Harington is a Chartered Professional Accountant, a Chartered Financial Analyst charterholder, and a Chartered Business Valuator. He is a Principal in the Toronto office of The Brattle Group, an economic consulting firm with offices around the world. Mr. Harington has provided business and intellectual property valuation and merger and acquisition advisory services for over 25 years.
[102] Mr. Harington’s mandate was to comment on the witness statements of Mr. Heimbecker as they relate to an assessment of efficiencies under section 96 of the Act. Mr. Harington was asked in particular to comment on whether, and if so the extent to which, the efficiencies that Mr. Heimbecker identified are cognizable under section 96 of the Act and would likely be lost if the Tribunal made the orders sought by the Commissioner.
[103] At the hearing, Mr. Harington was qualified as an expert in the quantification of efficiencies. The Tribunal found Mr. Harington to be credible, forthright, objective, and impartial, as well as willing to acknowledge the weaknesses/shortcomings in his own evidence or in the Commissioner’s case. He was a reliable and knowledgeable expert.
(2)
P&H
[104] Ms. Margaret Sanderson appeared on behalf of P&H as an expert witness.
[105] Ms. Sanderson is the Vice President and the global practice leader of the Competition and Antitrust Economics practice for the consulting firm Charles River Associates International Limited, a multinational firm that provides economic, financial, and business strategy consulting. She holds a M.A. in Economics and a B.Sc. in Economics and Quantitative Methods from the University of Toronto. Prior to joining Charles River Associates, Ms. Sanderson was Assistant Deputy Director of Investigation and Research within the Economics and International Affairs Branch of the Competition Bureau. She has 30 years of experience addressing the competitive effects of mergers and other firm conduct.
[106] Ms. Sanderson’s mandate was to provide her opinion on the likely anti-competitive effects of P&H’s Acquisition of the Virden Elevator and to respond to the initial expert report of Dr. Miller.
[107] With the parties’ agreement, Ms. Sanderson was accepted as an expert qualified to give opinion evidence in industrial organization and competition law economics. The Tribunal generally found Ms. Sanderson to be credible, forthright, objective, and impartial. Ms. Sanderson was helpful to the panel in her explanations.
C.
Documentary evidence
[108] The list of exhibits that were admitted in this proceeding is attached as Schedule “B” to these Reasons.
V.
Preliminary issues
[109] At the hearing, counsel for P&H raised issues regarding the Commissioner’s evidence and obligations in these proceedings. These preliminary matters must be addressed before dealing with the main issues in dispute in the Commissioner’s Application. They are as follows: 1) challenges to the evidence provided by the Commissioner’s experts; 2) adverse inferences and the Commissioner’s duty of fairness and obligations regarding the gathering of evidence; and 3) the legal burden of proof in this Application. Each will be dealt with in turn.
A.
Challenges to the Commissioner’s experts
(1)
Mr. Harington’s evidence
[110] At the hearing, P&H challenged a number of paragraphs found in Mr. Harington’s expert report on the issue of efficiencies. In particular, P&H asked the Tribunal to strike or give no weight to approximately 49 paragraphs of Mr. Harington’s report, on the basis that they express opinions of law related to statutory construction or the interpretation of cases. P&H further asserted that a number of other paragraphs of Mr. Harington’s expert report constitute inappropriate legal opinion evidence or inappropriate hearsay evidence related to the grain industry.
[111] The Commissioner responds that none of the challenged paragraphs should be struck. He submits that Mr. Harington set out his understanding of the legal framework as it informed his opinion on efficiencies. With respect to P&H’s claim that some paragraphs of Mr. Harington’s report should be struck because they constitute opinion evidence related to the grain industry, the Commissioner explains that efficiencies and economic experts need to set out their factual understanding of the industry before they can give their opinion. The Commissioner further notes that in this case, Mr. Harington cited all sources in support of the factual statements contained in his report.
[112] For the reasons that follow, the Tribunal agrees in part with P&H and will give limited weight to the legal opinions expressed by Mr. Harington as part of his expert report.
[113] As the Tribunal noted in The Commissioner of Competition v Vancouver Airport Authority, 2019 Comp Trib 6 (“VAA CT”), it has consistently applied the principles articulated by the SCC in R v Mohan, [1994] 2 SCR 9 (“Mohan”) and its progeny when it is tasked with determining the admissibility of expert evidence (VAA CT at para 107). In White Burgess Langille Inman v Abbott and Haliburton Co, 2015 SCC 23 (“White Burgess”), the SCC set out a two-step test for determining the admissibility of expert evidence. It held that in order to be admissible, expert opinion evidence must first meet the four threshold requirements established in Mohan, namely, relevance, necessity in assisting the trier of fact, absence of any exclusionary rule, and a properly qualified expert. At the second step, the decision maker engages in a balancing exercise and weighs the potential benefits of admitting the proposed evidence against the risks.
[114] It is well recognized that, under the principle of “necessity,” expert evidence must provide the courts with information that is considered as being “outside the experience and knowledge of a judge” (Mohan at p 23). The proposed expert opinion evidence must be necessary to assist the trier of fact, bearing in mind that necessity should not be judged strictly. This is notably the case where the expert evidence is needed to assist a court or a tribunal due to the technical nature of the issues at stake, or where the expertise is required to enable the decision maker to appreciate a matter at issue and to help it form a judgment on a matter where ordinary persons are unlikely to do so without the help of those with special knowledge.
[115] Experts, however, must not substitute themselves for the trier of fact (Mohan at p 24). As the Tribunal stated in VAA CT, “evidence that provides legal conclusions or opinions on issues and questions of fact to be decided by the court is inadmissible because it is unnecessary and usurps the role and functions of the trier of fact” (VAA CT at para 109, referring to Quebec (Attorney General) v Canada, 2008 FC 713 at para 161, aff’d 2009 FCA 361, 2011 SCC 11 and to Mohan at p 24). In sum, expert witnesses are not entitled to opine on legal matters, which fall within the scope of the court or Tribunal’s experience and knowledge. An expert opinion that is analogous to a memorandum of fact and law can become inadmissible as it “merely summarizes legal decisions, offers legal submissions on those decisions, and then expresses the author’s personal views on the ultimate issue that is for this Court to decide” (Es-Sayyid v Canada (Public Safety and Emergency Preparedness), 2012 FCA 59 at para 41). The closer the expert evidence approaches an opinion on the ultimate issue to be decided, the stricter the application of the principle will be.
[116] In many paragraphs of his expert report, Mr. Harington examines in detail the framework for quantifying cognizable efficiencies under section 96 of the Act. He does an extensive review of the provisions of the Act, of the case law, and of the Competition Bureau’s 2011 Merger Enforcement Guidelines, Competition Bureau Canada, October 6, 2011 (“2011 MEGs”)
[4]
. Relying on these legal sources, he provides his interpretation of section 96 of the Act dealing with efficiencies.
[117] It is not disputed that Mr. Harington is not a legal expert. The Tribunal agrees with P&H that the impugned paragraphs of his report constitute legal conclusions and opinion on an important issue that is up to the Tribunal to decide upon, namely, efficiencies. There is no doubt that the interpretation of section 96 and the determination of the proper legal framework to assess the efficiencies defence advanced by P&H falls within the Tribunal’s experience, expertise, and knowledge. The legal opinion expressed by Mr. Harington on this issue, strictly speaking, intrudes on the role and functions of the Tribunal.
[118] At the same time, the Tribunal acknowledges the extensive and well-recognized experience and expertise of Mr. Harington regarding the complex issue of efficiencies in merger reviews. Section 96 of the Act is a very technical provision and the Tribunal appreciates that Mr. Harington’s comments on how the jurisprudence has been thought through were made to provide the background of his analysis and to help the panel understand his reasoning. The Tribunal accepts that it would have been difficult for Mr. Harington to prepare his expert report and offer his opinion on P&H’s claimed efficiencies without providing some legal assumptions or basis to anchor his assessment of the particular facts in this case. In these circumstances, the Tribunal will not declare the impugned paragraphs of Mr. Harington’s report inadmissible as they are necessary to understand his opinion on efficiencies, but it will give them limited weight in the determination that the Tribunal is called upon to make on the appropriate legal framework under section 96 of the Act.
[119] Turning to P&H’s complaint about Mr. Harington’s comments on the Canadian grain industry, the Tribunal accepts that Mr. Harington is not a grain industry expert. The Commissioner was indeed not offering Mr. Harington’s evidence as such. However, the Tribunal is satisfied that, in making comments on the grain industry in his expert report, Mr. Harington was simply providing his factual understanding of the grain industry based on the documents contained in the evidence. P&H also had an opportunity to cross-examine him to expose the limits of his knowledge on this front. The factual references to the grain industry made by Mr. Harington are grounded on various portions of the evidence, and the Tribunal is not convinced that they should be declared inadmissible or given no weight. The Tribunal is not accepting what Mr. Harington says on the grain industry as a fact. It is simply taking note of the factual sources Mr. Harington relied on for his opinions.
(2)
Objectivity of the Commissioner’s experts
[120] P&H also asserted in its closing submissions that Dr. Miller failed to provide his expert opinion in an objective manner because he advanced a product market based on GHS, without examining the possibility of alternative product markets. P&H further submitted that neither Dr. Miller nor Mr. Harington opined objectively in their expert reports because of what it termed “their speculative approach” to what has occurred since the Transaction was completed. Echoing an observation made by Justice Moldaver (then at the Ontario Court (General Division)) in R v Clarke Transport Canada Inc, 1995 CanLII 7327, P&H claimed that the Commissioner’s experts were “hired guns.”
[121] The Tribunal does not agree.
[122] Nothing in Dr. Miller’s and Mr. Harington’s expert reports and testimonies, including in their respective cross-examinations, allows the Tribunal to conclude that these two experts did not provide their evidence objectively and in an impartial manner. Experts have a duty to provide independent assistance to a court at common law (White Burgess at para 26). Like many courts at the federal level and in provinces and territories, the Tribunal has also provided explicit guidance on the duty of experts by issuing its Notice on Acknowledgement of Expert Witnesses in December 2010. Pursuant to that Notice, experts appearing before the Tribunal have the obligation to sign a form acknowledging that they will comply with the Tribunal’s code of conduct for expert witnesses.
[123] The Tribunal’s code of conduct provides that experts must assist the Tribunal impartially, that they must be independent and objective, and that their role should not be conflated with that of an advocate for a party. In the Tribunal’s opinion, this is exactly what both Dr. Miller and Mr. Harington have done in this case. P&H’s claim that they were “hired guns” is entirely without merit and finds no support in the evidence heard by the Tribunal in this case.
B.
The Commissioner’s duty of fairness and adverse inferences
[124] A second area of preliminary issues relates to the Commissioner’s duty of fairness and his obligations regarding the gathering of evidence in the context of this Application. More specifically, P&H asked the Tribunal to draw adverse inferences against the Commissioner. P&H’s position is two-fold. First, P&H submitted that the Tribunal should draw an adverse inference against the Commissioner “generally” in this proceeding. Second, P&H argued that an adverse inference should be drawn against the Commissioner because he failed to obtain and produce evidence that was “peculiarly” within his power with respect to subsection 96(3) of the Act on efficiencies and the counterfactual test established in subsection 96(1).
[125] In its written submissions, P&H submitted that there are some circumstances in which a party who bears a burden of proof is not the party best situated to adduce the evidence related to the issue at stake, because the relevant facts lie particularly within the knowledge of the other party. The failure of a party to adduce evidence within its power may be considered as a matter of evidentiary weight and can lead to an adverse inference against it. In support of its position, P&H relied on the SCC’s decision in R v Jolivet, 2000 SCC 29 (“Jolivet”).
[126] P&H argued that in weighing the evidence in the record in this Application generally, and more specifically under section 96, the Tribunal must be “alive to what evidence is not in the record.” P&H maintained that, if there are gaps in the evidence, and the missing evidence was uniquely within the ability of the Commissioner to obtain, the Tribunal should weigh this consideration and be prepared to draw an adverse inference that such evidence, had it been produced, would not support the Commissioner’s position with respect to the Application generally and to efficiencies under section 96.
[127] In addition to the legal principles set out in Jolivet, P&H also referred to Tribunal decisions which, according to P&H, established a general duty of fairness owed by the Commissioner during proceedings under the Act (Commissioner of Competition v Canada Pipe Company, 2004 Comp Trib 2 (“Canada Pipe 2004”) at paras 60–64, aff’d 2004 FCA 76; see also Commissioner of Competition v Canada Pipe Company, 2003 Comp Trib 15 (“Canada Pipe 2003”) at para 53).
[128] During cross-examination of the Commissioner’s witnesses, counsel for P&H posed questions designed to show that the witnesses had additional documents, or information, or both, that the Commissioner had not elected to obtain and disclose, or had not included in the individual’s witness statement.
[129] During oral argument at the hearing, P&H further submitted that while the Commissioner had collected documents from the merging parties, made market contacts, and collected data from grain companies and Crushers prior to commencing this proceeding, the more important question was what the Commissioner did not obtain and file before the Tribunal. P&H contended that the Commissioner did not request nor obtain, from the grain companies or Crushers, any contemporary business documents related to market shares, markets, rail capacity and expansions, excess capacity, barriers to entry, or competition generally. According to P&H, it was incumbent on the Commissioner, acting in the public interest, to investigate the matter fully before commencing this proceeding and to put a full and proper evidentiary record before the Tribunal. The Commissioner having failed to ask for and obtain the evidence, P&H claims that an adverse inference should be drawn by the Tribunal against him.
[130] Not surprisingly, the Commissioner disagrees with P&H’s submissions. During the hearing, the Commissioner submitted that he had complied with his obligations. The Commissioner disagreed with P&H’s characterization of Justice Blanchard’s reasons in Canada Pipe 2004 because in that case, the Tribunal considered the 1994 Competition Tribunal Rules, SOR/94-290, which are no longer in effect. Relying on McIlvenna v Viebig, 2012 BCSC 218, aff’d 2013 BCCA 411, the Commissioner argued that the decision to draw an adverse inference is discretionary and should not occur unless it is warranted in all the circumstances.
[131] At the hearing, the Commissioner further argued that Jolivet was a criminal case about whether the Crown’s failure to call a witness at a criminal trial could be the subject of comment in the address to a jury by defence counsel. In the Commissioner’s submission, the decision in Jolivet confirmed that the Crown was under no obligation to call a witness it considered unnecessary to its case.
[132] The Commissioner also countered P&H’s arguments about best evidence with his own submission, stating that the best evidence of how P&H competes on a day-to-day basis at an Elevator through pricing should have come from grain merchants such as P&H’s employees, rather than relying solely on the evidence of Mr. Heimbecker, a senior executive of P&H. The Commissioner noted that two specific P&H employees were exclusively within the control of P&H and that there was no legitimate explanation for not calling them as witnesses.
(1)
Legal principles
(a)
Adverse inferences
[133] The drawing of an adverse inference from the absence of evidence relies on the reasoning that the failure by a party to call certain evidence may, depending on the circumstances, amount to an implied admission that the evidence would be contrary to the party’s case, or at least would not support it (Jolivet at para 28).
[134] In Jolivet, the SCC considered whether a jury was entitled to draw an adverse inference from the Crown’s failure to call a witness. During the trial, the Crown had indicated to the jury, twice, that it would be calling the witness to corroborate important admissions allegedly made by the accused. Just prior to the close of the prosecution’s case, the Crown advised the court that it no longer intended to call that witness and provided an explanation for this decision. Speaking for the SCC, Justice Binnie referred to the general rule developed in civil cases about adverse inferences from the failure to tender a witness, noting that a party may provide a satisfactory explanation for not doing so. A party may have no special access to the potential witness, or the missing proof may lie in the peculiar power of the party against whom the adverse inference is proposed — in which case the argument for an adverse inference is stronger (Jolivet at paras 25–27). The SCC also held that one “must be precise about the exact nature” of the adverse inference to be drawn. The SCC concluded that, because Crown counsel had announced to the jury its intention to call the allegedly corroborative witness, an adverse inference of “unhelpfulness” would have been a fair result owing to the Crown’s failure to substantiate its assertion of the existence of corroborative evidence (Jolivet at paras 29–30).
[135] The authors of Sopinka, Lederman & Bryant, The Law of Evidence in Canada, 5th ed (LexisNexis Canada Inc, 2018) describe the situations in which an adverse inference may be drawn as follows:
§6.471 In civil cases, an unfavourable inference can be drawn when, in the absence of an explanation, a party litigant does not testify, or fails to provide affidavit evidence on an application, or fails to call a witness who would have knowledge of the facts and would be assumed to be willing to assist that party. In the same vein, an adverse inference may be drawn against a party who does not call a material witness over whom he or she has exclusive control and does not explain it away. The inference should only be drawn in circumstances where the evidence of the person who was not called would have been superior to other similar evidence. The failure to call a material witness amounts to an implied admission that the evidence of the absent witness would be contrary to the party's case, or at least would not support it.
§6.472 An adverse inference should be drawn only after a prima facie case has been established by the party bearing the burden of proof.
[136] The Federal Court of Appeal (“FCA”) has applied this passage in Deyab v Canada, 2020 FCA 222 at para 46 and Caron Transport Ltd v Williams, 2020 FCA 106 at para 10.
[137] The FCA also considered adverse inferences in Toronto Real Estate Board v Commissioner of Competition, 2017 FCA 236 (“TREB FCA”) and Apotex Inc v Canada (Health), 2018 FCA 147 (“Apotex”). In TREB FCA, the court concluded that the Tribunal made no error in declining to draw an adverse inference against the Commissioner in the circumstances. The FCA held that the requested inference was tantamount to finding that the Commissioner had a legal obligation to quantify anti-competitive effects under section 92, which he had not because of the binding precedent issued by the SCC in Tervita SCC. In addition, the FCA stated as follows with respect to the Commissioner’s and the Tribunal’s roles in the proceeding:
[104] Considering that the Commissioner had no such legal obligation, he, like any other plaintiff, had to decide what evidence he had to put forward to prove his case. As we know, he chose to do so by way of qualitative evidence and in so doing, he took the risk of failing to persuade the Tribunal that the anti-competitive effects of TREB’s practice resulted in a substantial prevention of competition. As it turned out, the Tribunal was persuaded by the qualitative evidence adduced by the Commissioner.
[105] We have carefully considered the case law and cannot see any basis to accept TREB’s and CREA’s proposition that the Tribunal ought to have drawn an adverse inference against the Commissioner for failing to conduct an empirical assessment of markets in the United States and in Nova Scotia, or for that matter in the GTA. That, in our respectful view, would be akin to giving the Tribunal the power to dictate to the Commissioner how he should present his case. There is no authority for such a proposition.
[Emphasis added.]
[138] In Apotex, the FCA confirmed that recent decisions have treated the drawing of an adverse inference as a matter of discretion, to be exercised only where warranted in all of the circumstances. The court identified two reasons for this evolution. First, court rules now go a long way towards rendering witnesses and documents available to both sides, through discovery and other procedural mechanisms. Second, courts have recognized that whether or not an adverse inference is warranted on particular facts is bound up inextricably with the adjudication of the facts (Apotex at para 68, citing TREB FCA at para 107).
(b)
Discovery under the current CT Rules
[139] Pursuant to Rule 60 of the CT Rules, a party to a proceeding has to serve an affidavit of documents on each other party, identifying the documents that are “relevant” to any matter in issue and that are or were in the possession, power, or control of the party. CT Rule 60 does not distinguish between the Commissioner and the other parties for the purposes of discovery, and parties are all subject to the requirement of disclosing what is “relevant.” CT Rule 65 adds that access to what is disclosed must be provided.
[140] Relevance is determined by the way the issues are framed in the pleadings. A document of a party is considered relevant if the party intends to rely on it, if the document tends to adversely affect or support another party’s case, or if the document might fairly lead a party to a “train of inquiry” that could have either of these consequences (Antonio Di Domenico, Competition Enforcement and Litigation in Canada, (Toronto: Emond Montgomery Publications Limited, 2019) (“Di Domenico”) at p 736, referring to subsection 222(2) of the Federal Courts Rules, SOR/98-106 (“FC Rules”)). The definition of relevance is therefore quite broad and applies to all parties.
[141] FC Rule 226 further provides that the disclosure obligation is continuous. This requirement has been imported by the Tribunal in its proceedings (Tervita v Canada (Commissioner of Competition), 2013 FCA 28 (“Tervita FCA”) at para 74; The Commissioner of Competition v Air Canada, 2012 Comp Trib 20 at para 22). The continuous disclosure obligation entails that the initial disclosure affidavit must be updated any time a party becomes aware that it is deficient.
[142] The most recent court decision to have considered the Commissioner’s disclosure obligations is Vancouver Airport Authority v Commissioner of Competition, 2018 FCA 24 (“VAA FCA”), in which the FCA said the following:
[30] The procedural fairness obligations require the Commissioner of Competition to disclose to the Airport Authority evidence that is relevant to issues in the proceedings. This is necessary for the Airport Authority to know the case it has to meet and to fairly defend itself against the allegations. Often — as the Commissioner has recognized in this case by releasing roughly 8,300 documents from his investigatory file — this includes exculpatory material or other material resting in the investigatory file that could assist the party whose conduct is impugned in testing the evidence called by the Commissioner or in building its own case. […] In some cases, there may be limits on the obligation to disclose based on materiality, proportionality, applicable legislative standards and the nature of the proceedings. […]
[Citations omitted.]
[143] The FCA further noted that the Tribunal proceedings are adjudicative in nature, which typically commanded high procedural fairness requirements (VAA FCA at para 29).
(c)
Disclosure and the Commissioner’s duty of fairness
[144] In light of P&H’s submissions, it is also important to consider the issue of adverse inferences in the context of the more general legal principles governing the Commissioner’s disclosure obligations and duty of fairness. These go back to the Tribunal’s decision in Canada Pipe 2004.
[145] In Canada Pipe 2004, the respondent had requested additional disclosure from the Commissioner and to examine witnesses before the hearing. The procedural rules governing Tribunal proceedings back then were different from today’s; they applied a standard of reliance for the Commissioner’s general disclosure obligations, as opposed to the standard of relevance currently in place. In that case, the Tribunal dismissed the motion for additional discovery of documents and persons.
[146] In the Tribunal’s reasons, Justice Blanchard addressed the “duty of fairness” of the Commissioner (Canada Pipe 2004 at paras 60–64). He found that, although the Commissioner’s disclosure obligation was dictated by a standard of reliance under the then-rules, the Commissioner was “nonetheless required to act fairly in the exercise of her duties.” He noted that the Commissioner is a public officer with significant statutory powers to gather information and exercise public interest privilege, and there was a presumption that the Commissioner was acting in good faith. He further found that in those proceedings, the Commissioner was not a normal adversary, but rather a public officer with a statutory obligation to act fairly (Canada Pipe 2004 at para 62; see also Canada Pipe 2003 at para 53). Justice Blanchard likened the Commissioner’s obligations to that of a prosecutor who must act fairly, referring to the criminal law decisions in Boucher v The Queen, [1955] SCR 16 (“Boucher”) at pp 23–24 and R v O’Connor, [1995] 4 SCR 411 (“O’Connor”) at pp 477–478. He then stated:
[64] It naturally follows that just as the Crown prosecutor must be motivated by fairness and not the notion of winning or losing, so too the Commissioner must be motivated by goals of fundamental fairness and not by achieving strategic advantage on the proceeding. This is not to say that the duties articulated in such landmark criminal cases as Boucher, supra, or O'Connor, supra, should be directly imported into an administrative law setting. The Tribunal is an administrative Tribunal with an administrative process and procedural fairness must be customized to accommodate the expedited process required by the legislation and rules which govern its proceedings. Though the standard of disclosure may justifiably be different in proceedings before the Tribunal than in criminal proceedings, the underlying notion of fairness must remain constant for both. It is in this context that the reliance standard is to be applied.
[147] The Tribunal pauses to note that Justice Rand’s opinion in Boucher made comments about all available proof of facts being presented in a criminal matter. The passage from Justice L’Heureux-Dubé’s reasons for the SCC in O’Connor further referred to “full and fair disclosure as a fundamental aspect of the Crown’s duty to serve the Court as a faithful public agent, entrusted not with winning or losing trials but rather with seeing that justice is served” (O’Connor at para 101).
[148] P&H’s arguments in this proceeding do not concern disclosure obligations of the Commissioner so much as whether the Commissioner has an obligation to collect evidence (i.e., information, documents, and data) from third parties during an investigation or inquiry, and then to present that evidence fully to the Tribunal during proceedings commenced under section 92. The Tribunal observes that P&H cited no case dealing specifically with the Commissioner’s obligation to gather evidence during an investigation or inquiry, nor about whether the Commissioner may have an obligation to obtain an order under section 11 of the Act before a hearing, in order to assist a respondent with its case. P&H also did not refer to any cases involving other statutory officers’ or law enforcement officials’ obligations to carry out full and fair investigations or to obtain court orders to gather information for a party whose transaction or conduct is under review.
[149] Neither party referred to any prior determination of the Tribunal or appellate courts about the scope of the Commissioner’s obligation to present a full evidentiary record to the Tribunal, nor the obligations of any comparable statutory or law enforcement official (other than Boucher). Indeed, neither party referred to the remarks made by the FCA in TREB FCA at paragraphs 104–105, about the Commissioner’s decisions in presenting a case to the Tribunal.
[150] In addition, since Justice Blanchard’s decision, the procedural landscape during Tribunal proceedings, including disclosure rules, has changed. The Tribunal’s procedural rules passed in 2002 have been replaced by the CT Rules issued in 2008, which now contemplate a relevance-based approach to documentary discovery of the Commissioner. Since Canada Pipe 2004, the FCA has also revised the characteristics of the public interest privilege that existed in 2004 and examined procedural fairness obligations during Tribunal proceedings (VAA FCA at paras 28–35).
[151] In this context, it is fair to consider whether, and how, the Commissioner’s duty of fairness may have changed since Canada Pipe 2004, owing to a respondent’s right to disclosure and production of all non-privileged records in the possession or control of the Commissioner under the current CT Rules and the respondent’s ability to make its own comprehensive submissions and call its own evidence based on that same body of evidence. The Tribunal did not receive meaningful legal submissions on that question, nor does it have submissions on how the absence of the third party evidence in the present case adversely affected P&H’s already-vigorous defence against the Commissioner’s case — apart from general submissions criticizing the Commissioner’s efforts to collect the evidence and examples of what else could have been requested.
(2)
Tribunal’s assessment
[152] With these considerations in mind, the Tribunal will now consider the adverse inferences requested by P&H.
(a)
The “general” adverse inference
[153] For the following reasons, the Tribunal declines P&H’s request to draw an adverse inference against the Commissioner “generally” in this proceeding.
[154] First, P&H provided no specifics as to the exact nature of the adverse inference to be drawn. During the hearing, it made submissions that criticized the Commissioner’s investigation and lack of document production and data gathering, and it added generalized submissions of the same nature in oral argument. However, it did not specify that, because a particular piece of evidence was not tendered to the Tribunal or because a certain witness did not testify, the Tribunal should infer that some particular fact did occur, or that the Tribunal should draw an adverse inference of a specific nature against the Commissioner.
[155] The Tribunal finds it preferable to be asked much more precisely what inference to draw and on what basis, before deciding whether to draw an adverse inference (Jolivet at para 28). In the Tribunal’s view, such specificity is particularly important when a party asks the Tribunal to draw an inference against a party based on the absence of evidence or the absence of a witness. In the case at bar, the generalized adverse inference requested by P&H is too amorphous for meaningful adjudication.
[156] Second, the Tribunal is unaware of anything that prevented P&H from attempting to obtain documents or information itself (setting aside additional data, discussed separately below). P&H could have interviewed the farmer witnesses and could have attempted to interview or send written questions to the grain companies’ witnesses in advance of the hearing (or even while the merger review was occurring), and could have asked them for documents. There was no suggestion that P&H attempted to do so and was rebuffed, or that the Commissioner tried to interfere with any such attempts.
[157] Third, the Commissioner does not bear the exclusive or entire burden of adducing evidence for the Tribunal. In litigation in respect of a merger under section 92, the Commissioner is not required to present every bit of evidence at the hearing. Contested proceedings under section 92 are adversarial by nature. The Commissioner called some farmer witnesses to support his case under section 92 in relation to issues for which he had the evidentiary and legal burden of proof. It was the Commissioner’s risk not to obtain and present specific evidence from them (TREB FCA at paras 104–105).
[158] This is also not a situation where the witnesses were not called to testify at all. P&H had the opportunity to cross-examine each of the Commissioner’s farmer witnesses to expose missing or incomplete information, and it did so in several respects. The cross-examination revealed that there were additional inquiries that could have been made to the farmer witnesses and there were documents that could have been requested from them.
[159] In these circumstances, the Tribunal prefers a more surgical alternative instead of a general adverse inference against the party that called the witness to testify. Incomplete evidence gathered from or presented by a witness during examination-in-chief may adversely affect the credibility or reliability of the witness’s testimony. Exposed during cross-examination at a hearing, it can sometimes be damaging to a party’s case. Given that P&H could also have easily sought the same information and documents, apparently did not do so, but exposed the issues at the hearing, the Tribunal concludes that the Commissioner took on the risk of failing to discharge his burden under section 92 and of having adverse reliability or credibility findings made by the Tribunal against the witnesses.
[160] Fourth, the absence of data from rival Elevators is addressed separately, below. That analysis supports the Tribunal’s conclusions on the requested general adverse inference.
[161] Finally, having considered the parties’ submissions, the Tribunal is disinclined in this case to make a legal ruling with potentially far-reaching consequences concerning the Commissioner’s general duty of fairness as it concerns either gathering evidence for a proceeding under section 92 or presenting that evidence. The Tribunal notes that P&H’s pleading in response to the Application (“Response”) did not express any concerns about the Commissioner’s investigation or inquiry into the proposed merger. P&H did not later seek to amend its pleading after it received the Commissioner’s Affidavit of Documents or after its oral discovery of the Commissioner. Nor did P&H raise any concerns to the Tribunal on receipt of the witness statements, or else seek any further order before the hearing. Considering how and when P&H raised the issue and the scope of the parties’ submissions, the Tribunal considers it unnecessary and inappropriate to make more detailed comments.
[162] In light of the foregoing, the Tribunal exercises its discretion not to make a generalized adverse inference against the Commissioner. In stating this conclusion, the Tribunal should not be understood to express a view on the scope of the Commissioner’s fairness duties as submitted by P&H and denied by the Commissioner in this case. Resolving issues related to the Commissioner’s general fairness obligations in the disclosure process will be for another day.
(b)
The adverse inferences related to efficiencies under section 96
[163] P&H also argued that the Tribunal should draw a more specific adverse inference against the Commissioner for his failure to obtain certain data from third-party grain companies that compete with it at the Virden Elevator, and which had an impact on the evidence relating to the efficiencies defence.
[164] P&H took the position that the Transaction would increase throughput at the Virden Elevator, resulting in cognizable efficiencies for the purposes of section 96. During the cross-examination of Mr. Harington, the Commissioner’s expert on efficiencies, P&H drew attention to paragraph 130 of Mr. Harington’s expert report. In that paragraph, Mr. Harington stated that the only way a redistribution of throughput between competing Elevators would result in an efficiency to the Canadian economy is if “the entity from which the increased throughput is being taken operates at a higher per unit variable operating cost” than P&H (Exhibits P-A-195, CA-A-196 and CB-A-197, Expert Report of Mr. Andrew Harington (“Mr. Harington Report”), at para 130). Mr. Harington testified under cross-examination that he did not have the variable operating costs of the rival Elevators to the Virden Elevator. Without that, he said, he could not do the comparison contemplated by his paragraph 130. Because his mandate was to respond to the alleged efficiencies claimed by P&H in Mr. Heimbecker’s initial witness statement on efficiencies, rather than to determine the efficiencies himself, Mr. Harington did not request or obtain the variable operating cost data of the rival Elevators.
[165] Mr. Harington testified that in fact, he would have required a lot more than the variable operating costs data: he would have needed all of the data on locations of farms that shifted volumes of grain from one Elevator to another, and what the transportation costs were for those farms. He would have looked at the efficiencies implications for all of Canada. Mr. Harington further noted that he would not reasonably expect P&H to have its competitors’ variable cost data. However, Mr. Harington testified that he had all of the evidence he needed to do the job he was asked to do (i.e., to respond to P&H’s position on increased throughput at the Virden Elevator as an efficiency under section 96).
[166] P&H submitted that the Tribunal should draw an adverse inference against the Commissioner owing to the Commissioner’s failure to request and obtain the variable operating costs data from rival Elevators because, without the data, a precise assessment could not be completed for the purposes of the counterfactual test contemplated in subsection 96(1) of the Act and the redistribution analysis under subsection 96(3). According to P&H, the Commissioner could have obtained the required data either by request or by obtaining an order under section 11 of the Act.
[167] The Commissioner responded that he had no such burden under section 96. Referring to paragraph 122 of the SCC’s decision in Tervita SCC, the Commissioner observed that the merging parties bear the onus of establishing all elements of the efficiencies defence after the Commissioner has discharged his initial burden to prove the anti-competitive effects and the DWL for the purposes of section 96. A respondent’s burden includes proof of the extent of the efficiency gains and whether the gains are greater than and offset the anti-competitive effects. The Commissioner noted that P&H’s position, according to which the Commissioner did not collect evidence enabling it to prove an efficiency claim, was not raised in its own initial Response pleading. The Commissioner noted that P&H could have sought, but did not seek, discovery from third parties to obtain the information it now requires. P&H decided to discharge its burden to quantify cognizable efficiencies through a witness statement from Mr. Heimbecker, rather than from an expert witness. According to the Commissioner, P&H cannot shift the burden onto the Commissioner for its own failure to discharge its burden.
[168] The Tribunal agrees with the Commissioner and will not draw the specific adverse inferences requested by P&H against the Commissioner in relation to efficiencies. There are three reasons for this determination.
[169] First, it is not clear what exactly P&H seeks from the Tribunal by way of adverse inference. Again, P&H did not specify which rival Elevators’ data were at issue, who owed them, what the variable costs data would necessarily or could reveal (by itself or in combination with other unidentified data), or what the outcome would be under subsections 96(1) or (3) following analysis and quantification.
[170] For example, P&H did not explain how the absence of variable operating costs data at one or more unnamed Elevators constitutes an implied admission against the Commissioner that the data will lead to a cognizable and quantifiable efficiency under section 96. It would be speculative to find that such an implied admission follows from the sole absence of unknown data. To do so would require making several assumptions about the contents of the data and the outcome of calculations using those data. As Mr. Harington’s testimony confirmed, significant additional data would be required to do the analysis he envisioned. Accordingly, no inference is warranted based on an absence of the variable operating costs data.
[171] Second, P&H has not demonstrated that the Commissioner had an obligation to obtain the data. It has cited no case nor pointed to a principled basis for such an obligation. Apart from the initial burden on the Commissioner under section 96 to show and quantify anti-competitive effects as established in Tervita SCC, the legal burden under section 96 is on the respondent. While P&H sought to argue that Tervita SCC did not settle the evidentiary burden under section 96, it provided no compelling legal or factual reason to shift a further burden onto the Commissioner on the facts of this case.
[172] Third and relatedly, P&H has not shown that the Commissioner knew or should have known that P&H needed the data in the present case. There is no evidence that P&H made any efforts itself to request or obtain the variable operating costs data. While P&H may well be correct that its competitors would not voluntarily provide that data to a rival, it did not try to obtain the data by way of request to them or by filing a motion with the Tribunal.
[173] On the evidence, the Tribunal does not accept that the Commissioner should (or even could) have known that P&H required the data. P&H acknowledged during argument that it did not ask the Commissioner to obtain it. When asked by the Tribunal how the Commissioner would have known that P&H needed it, or whether the Commissioner should have filed an application for an order under section 11 of the Act to obtain it, P&H did not provide a clear answer.
[174] Moreover, based on the events leading up to the hearing, the Tribunal sees no realistic basis on which the Commissioner could have known that he should obtain the impugned data:
At the pleadings stage in this proceeding, P&H did not raise possible efficiencies arising from increased throughput at the Virden Elevator, nor anything specific about subsection 96(3) of the Act. Its Response pleaded that the efficiencies from the Transaction “include: improved [FGT] scale economies and cost savings, elimination of the margin that [LDC] formerly paid to use the Vancouver export terminal owned by Kinder Morgan, outlay expansion and improved scale economies at the former [LDC] elevator and administrative synergies;”
There was no suggestion that P&H noted the absence of the data and raised it after receiving the Commissioner’s Affidavit of Documents;
At the oral examinations for discovery, P&H declined to provide the Commissioner with any specific insight about its efficiencies defence. Counsel for the Commissioner asked several questions requesting information about efficiencies to Mr. Heimbecker. The answers provided by P&H’s counsel were essentially that it was a matter for an expert report to be filed later and that otherwise, no substantive answers would be provided at discovery;
Mr. Heimbecker repeated that position in his subsequent responses to undertakings and questions taken under advisement;
However, P&H did not file an expert report concerning efficiencies;
Mr. Heimbecker’s reply witness statement, delivered over two months before the hearing started, set out evidence to advance P&H’s position on efficiencies. However, it made no reference to any need for variable operating costs data from rival Elevators;
P&H also did not raise any need for data after it received a copy of Mr. Harington Report, also more than two months before the hearing commenced. As noted above, this expert report referred directly to variable operating costs of other entities;
P&H did not file a motion to the Tribunal seeking an order to compel the Commissioner to obtain the data; and
The issue did not come to light until Mr. Harington’s cross-examination, near the end of the hearing.
[175] In these circumstances, the Tribunal finds it unrealistic to expect that the Commissioner would be or could have been aware that P&H required variable operating costs data of rival Elevators for its efficiencies defence. It was equally unrealistic to expect the Commissioner to be aware that P&H expected him to attempt to obtain that data either by request or under section 11 of the Act. Rather, the Tribunal finds P&H’s position on the need for this data to be late-blooming and tactical, rather than based on a substantive need to support its position on efficiencies arising at the Virden Elevator.
[176] Exercising its discretion based on the evidence and arguments made, the Tribunal therefore declines to make any specific adverse inferences on issues related to efficiencies. To draw an adverse inference against the Commissioner in the present circumstances would be demonstrably unfair.
C.
Legal and evidentiary burden applicable to sections 92 and 96 of the Act
[177] The last preliminary issue that needs to be briefly addressed is the legal burden of proof in this Application. In its submissions, P&H suggested that the allocation of the burden of proof established by the SCC in Tervita SCC has left some questions unanswered regarding the Commissioner’s burden under section 96 of the Act.
[178] With respect, the Tribunal disagrees.
[179] It is not disputed that, under section 92, the Commissioner bears the burden of proving that the merger will create, maintain, or enhance market power through the merged entity’s ability to profitably influence price, quality, service, or other dimensions of competition. However, there is no requirement for the Commissioner to prove that the merged entity will, in fact, exercise these powers (The Commissioner of Competition v Canadian Waste Services Holdings Inc, 2001 Comp Trib 3 (“Canadian Waste”) at para 108, aff’d 2003 FCA 131, leave to appeal refused, [2004] 1 SCR vii; Superior Propane I at para 258). In determining whether the Commissioner has met his burden on this point, a forward-looking analysis of whether the merger will give the merged entity the ability to prevent or lessen competition substantially compared to the pre-merger benchmark — or “but for” world — must be conducted (Tervita SCC at para 51).
[180] With respect to section 96, Justice Rothstein in Tervita SCC clearly stated that “the [Superior Propane cases] established that the Commissioner has the burden under s. 96 to prove the anti-competitive effects” of a merger (Tervita SCC at para 122). Conversely, the merging parties bear the onus of establishing all the other elements of the efficiencies defence, including the extent of the efficiency gains and whether the gains are greater than and offset the merger’s anti-competitive effects (Tervita SCC at para 122). To meet his burden, the Commissioner must quantify the quantifiable anti-competitive effects he relies upon. Where these effects are measurable, they must be calculated or at least estimated, and a failure to quantify quantifiable effects will not result in such effects being considered qualitatively or remaining undetermined (Tervita SCC at paras 125–133). Justice Rothstein explained that an approach that would permit the Commissioner to meet his burden without at least establishing estimates of the quantifiable anti-competitive effects would fail to provide the merging parties with the information they need to know the case they have to meet (Tervita SCC at para 124). Qualitative anti-competitive effects which are not quantifiable can also be taken into account, provided they are supported by the evidence and the reasoning for the reliance on the qualitative aspects is clearly articulated by the Tribunal (Tervita SCC at para 147).
[181] In the Tribunal’s view, there is at present no legal precedent for the Commissioner to have any additional burden under section 96 beyond that established by the SCC in Tervita SCC. P&H has not provided any argument or sufficient supporting evidence that could allow the Tribunal to revisit, revise or enlarge the clear standard set out in Tervita SCC on the legal and evidentiary burden of the Commissioner under the merger provisions of the Act.
VI.
Issues
[182] The following broad issues are raised in this proceeding:
- What is or are the relevant product market(s) for the purposes of this proceeding?;
- What is or are the relevant geographic market(s) for the purposes of this proceeding?;
- Has the Commissioner established, on a balance of probabilities, that the Virden Acquisition lessens, or is likely to lessen, competition substantially?;
- If the Commissioner has established that the Virden Acquisition lessens, or is likely to lessen, competition substantially, what is the remedy to be ordered?;
- Has P&H established, on a balance of probabilities, that the gains in efficiency will be greater than, and will likely offset, the effects of any lessening of competition pursuant to section 96 of the Act?;
- What costs should be awarded?
[183] Each of these issues will be discussed in turn.
VII.
Analysis
A.
What is or are the relevant product market(s) for the purposes of this proceeding?
[184] In order to determine whether the Virden Acquisition lessens competition substantially, or is likely to do so, the Tribunal must first identify the product and geographic dimensions of the relevant market(s) for the purposes of this proceeding. In this case, the fundamental dispute between the parties is how to properly characterize the product market — and more specifically, the relevant product and the relevant price — in a situation where the merging firms’ alleged specific contribution of value is only a component of the final price of the product. The Commissioner claims that P&H supplies GHS for wheat and canola to farmers, whereas P&H submits that it purchases wheat and canola from farmers. The Commissioner submits that the relevant price is the “imputed” price for GHS — which, he says, approximates the Basis —, while P&H argues that it is the Cash Price charged to farmers for their grain.
[185] As acknowledged by both Dr. Miller and Ms. Sanderson during their respective testimony, the definition of the relevant product market is a key element that has an impact on the rest of the Tribunal’s analysis in this case (i.e., the geographic market, the competitive effects analysis, the market shares, the surplus calculations, etc.).
(1)
Analytical framework
(a)
The purpose of market definition
[186] In assessing whether, under section 92 of the Act, a merger lessens competition substantially or is likely to do so, the focus is on whether the merger is likely to create, maintain or enhance the ability of the merged entity to exercise market power, unilaterally or in coordination with other firms (Tervita SCC at para 44).
[187] Market power is not defined in the Act. Market power has been described by the Tribunal as the ability to “profitably influence price, quality, variety, service, advertising, innovation or other dimensions of competition” (Canadian Waste at para 7) or as “the ability to maintain prices above the competitive level for a considerable period of time without such action being unprofitable” (Hillsdown at p 314). Both of these descriptions were cited with approval in Tervita SCC, at paragraph 44.
[188] The first step in measuring market power is to define the relevant market. Put differently, the purpose of identifying the relevant product (or geographic) market is to identify the possibility for the exercise of market power (Canadian Waste at para 39; Superior Propane I at para 47; Director of Investigation and Research v Southam (1992), 43 CPR (3d) 161 (Comp Trib) at pp 177–178). Market definition is often considered a critical component in assessing market power because it frames the context within which competitive effects can be analyzed (Di Domenico at p 408).
[189] The Tribunal and the courts have traditionally considered it necessary to define a relevant market before proceeding to assess the competitive effects of mergers under the Act (Canada (Director of Investigation and Research) v Southam Inc, [1997] 1 SCR 748 at para 79; Tervita CT at paras 92, 360–364; Superior Propane I at para 56; Hillsdown at p 297). The relevant market is typically a predicate to a finding of substantial lessening or prevention of competition in merger cases because the merger must be one that will substantially lessen or prevent competition, or is likely to do so, within an area of actual or potential competition.
[190] However, the Tribunal has cautioned against losing sight of the ultimate inquiry and task of the Tribunal, which is to determine whether the merger being assessed prevents or lessens, or is likely to prevent or lessen, competition substantially (Superior Propane 1 at para 48). Market definition is not an end in itself: it is merely an analytical tool to assist in evaluating anti-competitive effects (Superior Propane I at para 48; 2011 MEGs at para 3.2).
[191] It is further important to note that a competition market is an analytical construct, and neither the product market nor the geographic market needs to coincide with the market as it is considered by a business or industry (Superior Propane I at paras 67, 85, 101, 106). Relevant markets for the purpose of a merger analysis are not always intuitive and may not align with how industry participants use the term “market” or view their “market.”
[192] Market definition is typically the subject of contested submissions and can often be outcome-determinative in merger matters under section 92.
(b)
Rationale and tools for market definition
[193] When defining relevant markets in proceedings brought under section 92 of the Act, the Tribunal considers whether there are close substitutes for the product at issue. Market definition is based in part on substitutability, and it focuses primarily on demand responses to changes in relative prices after the merger. The ability of a firm to raise prices without losing sufficient sales to make the price increase unprofitable ultimately depends on the purchasers’ willingness to pay the higher price. This is determined by analyzing evidence of the ability of purchasers to switch their purchases to substitute products and locations in response to a price increase (Tervita CT at paras 58–60). Close substitutes have been defined in terms of whether “buyers are willing to switch from one product to another in response to a relative change in price, i.e., if there is buyer price sensitivity” (Canada (Commissioner of Competition) v Tele-Direct Publications Inc (1997), 73 CPR (3d) 1 (Comp Trib) (“Tele-Direct”) at p 35, citing the test adopted by the FCA in Canada (Director of Investigation and Research) v Southam Inc, [1995] 3 FC 557, 63 CPR (3d) 1 (FCA) (“Southam FCA”), rev’d on other grounds [1997] 1 SCR 748).
[194] In assessing the extent of the product (and geographic) dimensions of relevant markets in the context of proceedings under the Act, the Tribunal has generally followed the well-established hypothetical monopolist analytical framework, or hypothetical monopolist test (“HMT”) (VAA CT at para 300; The Commissioner of Competition v The Toronto Real Estate Board, 2016 Comp Trib 7 (“TREB CT”) at paras 121–124; The Commissioner of Competition v Visa Canada Corporation and MasterCard International Incorporated, 2013 Comp Trib 10 (“Visa Canada”) at para 173; Tervita CT at para 58; Superior Propane I at para 57).
[195] In Tervita CT and Superior Propane I, two merger cases, the Tribunal embraced the description of that framework set forth in the Competition Bureau’s MEGs (see, for example, 2011 MEGs at para 4.3). Under this approach, a relevant product market is defined as the smallest group of products (including at least one product of the merging parties) in respect of which a sole profit-maximizing seller — the hypothetical monopolist — controlling all suppliers in the proposed market would find it profitable to impose and sustain a small but significant and non-transitory increase in price (“SSNIP”) above levels that would likely exist in the absence of the merger. The purpose of the HMT is to determine the extent to which customers in the candidate market will switch to other products in response to a SSNIP (Visa Canada at para 198). In the determination of whether a SSNIP would be profitable, the HMT makes use of demand elasticity and cross-elasticity evidence as well as what are known as practical indicia. If a small price increase would drive purchasers to an alternative product, then that product must be reasonably substitutable for those in the proposed market and must therefore be part of the market, properly defined. The conceptual exercise is repeated to include a broader array of products until it defines a product set over which a hypothetical monopolist could profitably impose a SSNIP (Brian A. Facey and Cassandra Brown, Competition and Antitrust Laws in Canada: Mergers, Joint Ventures and Competitor Collaborations, 3rd ed (Toronto: LexisNexis Canada Inc, 2020) (“Facey and Brown”) at p 205).
[196] Pursuant to the HMT framework, the product dimension of a relevant market is defined in terms of the smallest group of products in respect of which a hypothetical monopolist would have the ability to impose and sustain a SSNIP above levels that would likely exist in the absence of the merger. The “smallest group” principle is an important component of the test because, without it, there would be no objective basis upon which to draw a distinction between a smaller group of products in respect of which a hypothetical monopolist would have the ability to profitably impose a SSNIP and a larger group of products in respect of which that monopolist may also have such an ability (VAA CT at para 326; TREB CT at para 124).
[197] The SSNIP will be applied to the price that is being paid by the purchasers of the candidate product (Visa Canada at para 198), often referred to as the “base price” (see, for example, 2011 MEGs at para 4.6).
[198] Generally, for the purposes of determining the SSNIP, the objective benchmarks are as follows (and are reflected as such in the 2011 MEGs): an “increase in price” is typically one of 5% or more, and a “non-transitory” price increase is typically one that is maintained for at least one year. This 5%/one-year approach is generally treated as a “threshold” used to identify market power at the market definition stage, where the objective is to define the smallest market in which a substantial lessening of competition would be possible. If sellers of a product or of a group of products in a provisionally defined market, acting as a hypothetical monopolist, would not have the ability to profitably impose and sustain a 5% price increase lasting one year, the product bounds of the relevant market will be progressively expanded until the point at which a hypothetical monopolist would have that ability and degree of market power. Essentially the same approach is applied to identify both the product and geographic dimensions of relevant markets.
[199] Indeed, the Commissioner and P&H both acknowledged that a 5% increase and a one-year time frame are the standard thresholds for a SSNIP, here in Canada and in many other jurisdictions. However, these benchmarks can be adjusted to reflect the specific realities of a given industry or business.
[200] The Tribunal agrees that the HMT approach adopted in previous Tribunal cases, consistent with the 2011 MEGs, should continue to be used in this case.
[201] Given the practical challenges associated with determining the base price in respect of which the SSNIP assessment must be conducted in a proceeding brought under section 92 of the Act, market definition will often include not only the analysis of prices through the HMT framework but also other evidence of substitutability or customer switching. Market definition may therefore involve assessing indirect evidence of substitutability, including factors such as: functional interchangeability in end-use of the products; switching costs; the views, strategies, behaviour and identity of buyers; trade views, strategies and behaviours of other market participants; physical and technical characteristics; and price relationships and relative price levels.
(c)
The language of section 92
[202] During final argument, the Tribunal raised an issue related to the interpretation of section 92. The Tribunal observed that, contrary to other provisions of the Act such as civil agreements between competitors (section 90.1), section 92 on mergers does not expressly refer to a substantial lessening or prevention of competition “in a market.” Subsection 92(1) rather uses broader language incorporating the phrase “trade, industry or profession” in four paragraphs referring to the effect of the merger on competition: “(a) in a trade, industry or profession; (b) among the sources from which a trade, industry or profession obtains a product; (c) among the outlets through which a trade, industry or profession disposes of a product; or (d) otherwise than as described in paragraphs (a) to (c).”
[203] Both parties expressed the view that the absence of the word “market” in subsection 92(1) makes no difference, that the merger provision clearly relates to a substantial lessening or prevention of competition “in a market,” and that determining the relevant “market” forms part of the analysis to be conducted by the Tribunal.
[204] The Tribunal agrees and finds that the absence of the word “market” in the opening part of section 92 should not be interpreted as implying that a relevant competition market does not need to be defined or utilized in merger analysis.
[205] The principled approach to statutory interpretation requires that section 92 be read in its entire context, in its grammatical and ordinary sense, harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament (see, for example, Bell ExpressVu Limited Partnership v Rex, 2002 SCC 42 (“Bell ExpressVu”) at para 26).
[206] Looking first at the wording of section 92, the Tribunal agrees with P&H that a merger as defined in section 92 clearly encompasses the concept of market or markets. A merger is defined in section 91 as the taking of control of a “business.” The word “business” is in turn defined broadly in section 2 of the Act as “the business of (a) manufacturing, producing, transporting, acquiring, supplying, storing and otherwise dealing in articles, and (b) acquiring, supplying and otherwise dealing in services” [emphasis added]. The French version of the Act uses the words “tout autre commerce” to translate the expression “otherwise dealing.” The use of the word “business” in the definition of “merger” therefore makes it clear that a merger for the purposes of a section 92 assessment is with respect to the market or commercial activity of the “business.” Moreover, since a “business” is defined in terms of activities dealing in articles or services, the definition must concern one or more articles or services (as each are defined in subsection 2(1)) that are bought or sold as part of a commercial activity.
[207] Importantly, there are also express references to the notion of “market” in subsection 92(2) and section 93 of the Act, which provide direction on how to assess whether a merger lessens or prevents competition substantially. Subsection 92(2) prohibits the Tribunal from making a finding and exercising its discretionary power to impose a remedy under subsection 92(1) solely on the basis of evidence of “concentration or market share.” The assessment under section 92 is further informed and limited by section 93, which sets out factors that may be considered in determining whether a merger affects competition in a significant way for the purposes of section 92. Section 93 contains sustained references to the concept of “market” in paragraphs 93(d), (e), (g), (g.1), (g.2), and (h), and implies that such a market has been defined to make the competitive assessment. For example, paragraph (h) refers to “any […] factor that is relevant to competition in a market that is or would be affected by the merger or proposed merger.”
[208] Turning to the object and purpose of section 92, there is no doubt that the Tribunal’s focus is on assessing the degree to which market power is created, maintained or enhanced by the merger at issue. The concept of a competition or antitrust market is implicit in many provisions of the Act for the identification of anti-competitive conduct and for the substantiality threshold that must be applied to the assessment of anti-competitive effects.
[209] The Tribunal further observes that since the Tribunal’s first decision in a contested merger proceeding (i.e., Hillsdown), subsection 92(1) has consistently been interpreted as synonymous with “market” by the Tribunal and the courts (Tervita SCC at para 44; Hillsdown at pp 297–314), and by the Commissioner. In Tervita SCC, the SCC made it clear that the assessment of the substantial effect on competition was an effect on the “market”: it “involves assessing the degree and duration of any effect it would have on the market” [emphasis added] (Tervita SCC at para 78).
[210] In sum, section 92 does not have a different scope in its relationship to the substantial lessening of competition even if it refers to the effect of competition in a “trade, industry or profession,” as opposed to a “market.” Given the definition of “merger” and the fact that several factors listed in subsection 92(2) and section 93 expressly contemplate an evaluation made in relation to a “market,” the Tribunal is satisfied that, even though section 92 does not expressly refer to a substantial lessening or prevention of competition in a market, “Parliament intended that competition must be shown to be likely to be prevented or lessened substantially in a competition law or antitrust market” (Paul S. Crampton, Mergers and the Competition Act, (Carswell, 1990) (“Crampton 1990”) at p 261).
(d)
HMT and monopsony
[211] Finally, the Tribunal observes that the approach to market power and to market definition, which has mostly developed in matters involving the sale of a product, is similar in the context of the purchase of a product: the market power of buyers is the “ability of a single firm (monopsony power) […] to profitably depress prices paid to sellers […] to a level that is below the competitive price for a significant period of time” (2011 MEGs at para 2.4).
[212] The HMT framework therefore applies to define relevant markets for both the sale and the purchase of a product. For monopsony power, the 2011 MEGs describe the analytical process as follows, at paragraph 9.2:
[…] The conceptual basis used for defining relevant markets is, mirroring the selling side, the hypothetical monopsonist test. A relevant market is defined as the smallest group of products and the smallest geographic area in which a sole profit‑maximizing buyer (a “hypothetical monopsonist”) would impose and sustain a significant and non‑transitory price decrease below levels that would likely exist in the absence of the merger. The relevant product market definition question is thus whether suppliers, in response to a decrease in the price of an input, would switch to alternative buyers or reposition or modify the product they sell in sufficient quantity to render the hypothetical monopsonist’s price decrease unprofitable.
[213] In such cases, the SSNIP becomes “a small but significant and non-transitory decrease in price” (“SSNDP”) below levels that would likely exist in the absence of the merger
[5]
.
(2)
Parties’ positions
(a)
The Commissioner
[214] The Commissioner submits that the proper way to characterize the relevant product market is to define it as the provision of GHS for wheat and canola to farmers who, prior to the Acquisition, benefited from competition between the Virden and Moosomin Elevators. He argues that the supply of GHS for wheat is a relevant product market and that the supply of GHS for canola is a separate one. According to the Commissioner, a hypothetical monopolist of GHS for each of wheat and canola could profitably impose a 5% SSNIP. Moreover, says the Commissioner, there are no functional substitutes for GHS for wheat or canola.
[215] The Commissioner maintains that local competition between Elevators for wheat or canola manifests itself through the Basis, and not through the final Cash Price that an Elevator pays to the farmer. He submits that, because P&H has no control over the Futures Price, the only component of the Cash Price it can control is the Basis, and that only the Basis component of the Cash Price is affected by local competition between the Elevators. The Basis, or the “imputed” price for GHS calculated by Dr. Miller, adds the Commissioner, represents the Elevators’ specific contribution of value in the final Cash Price paid to farmers. The Commissioner argues that the relevant product market should therefore reflect an economic framework that analyzes the competition actually impacted by the Virden Acquisition.
[216] The Commissioner contends that P&H’s approach to the product market allows it to mask and obscure the anti-competitive effects of its Acquisition by focusing on the grain itself and on the Cash Price it pays to the farmers, whereas the services effectively provided by P&H are only a small part of the overall value of the grain they purchase from the farmers and resell to their end customers.
[217] By the end of the hearing, both the Commissioner and Dr. Miller agreed that their proposed approach to the product market definition could be qualified or described as a “value-added” approach, even though the Commissioner did not use these specific terms in his pleadings or even in his opening submissions at the hearing, and even though Dr. Miller did not describe his analytical approach as such in his initial expert report.
[218] The Commissioner maintains that economics, the facts, and the law support his “value-added” approach to product market definition.
[219] With respect to economics, the Commissioner submits that defining the product market as the sale of GHS by P&H “facilitates” an economic analysis focused on the competition affected by the Virden Acquisition and the Elevators’ contribution of value. On this point, the Commissioner mostly relies on Dr. Miller’s initial and reply expert reports and on his testimony. Using common features of the market definition exercise, such as the HMT, diversion ratios, and upward pricing pressure (“UPP”) calculations, Dr. Miller testified that he relied on the “imputed” price of GHS because, in his view, this aligns with the service that the Elevators effectively give to farmers. Dr. Miller opined that receiving GHS from Elevators is what is enabling the farmers to access the worldwide market. According to Dr. Miller, the price for GHS is an important factor in a farm’s choice of Elevators, and price competition between Elevators is reflected in the Basis and the “imputed” price he calculated for GHS.
[220] With respect to facts, the Commissioner relies on four main elements to support his proposed product market: 1) P&H recognizes the role of the Basis in competition between Elevators, and as one of the two components of the Cash Price, along with the Futures Price; 2) the only component of the price of grain that P&H can set is the Basis. P&H claims that it uses its “Workback Algorithm” to determine its Cash Prices but, according to the Commissioner, this algorithm sets the Basis and not the Cash Price, which can fluctuate during the day with the variations of the Futures Price. Moreover, P&H sends mass communications about the Basis on a daily basis, through emails and on its P&H Direct application; 3) the Basis is also carried over to the contracts concluded between Elevators and farmers, and this is done by all grain companies with the exception of LDC; and 4) farmers are affected by local competition and they use the Basis to make comparisons between Elevators. In sum, argues the Commissioner, the provision of GHS is how the industry operates, and the Basis is an important industry-wide practice.
[221] Turning to the law, the Commissioner submits that characterizing the product market as GHS is consistent with the purpose clause of the Act, with previous merger cases, and with the 2011 MEGs. With respect to the so-called “value-added” approach, he claims that the U.S. Horizontal Merger Guidelines issued in 2010 (U.S. Department of Justice and the Federal Trade Commission, Horizontal Merger Guidelines, August 19, 2010 (“US HMEGs”)) actually recognize that in situations where the price of a merging firm’s contribution to value can be identified with “reasonable clarity,” the price used to assess the merger can be a component of the final price of the product that the firms effectively compete on.
(b)
P&H
[222] P&H responds that GHS is not a product supplied by P&H and the other grain companies. The relevant products, says P&H, are the wheat and canola purchased from farmers by P&H, for which the farmers receive the Cash Price. According to P&H, the so-called GHS are only internal processes that the grain companies may apply to the grain after the title to the grain has passed from the farmer to the grain company: in other words, the GHS claimed by the Commissioner are merely costs to the grain companies. Furthermore, it is export terminals that are used to do the following: to receive grain from rail; to grade, segregate, and store grains by type and quality attributes; to clean when required and blend; and to load grain onto vessels. Therefore, many of the GHS claimed by the Commissioner are not services that are performed at the Elevator level.
[223] P&H argues that the documentary and viva voce evidence shows that GHS are not actually transacted, contracted, or discussed between farms and P&H or other grain companies. P&H adds that Mr. Heimbecker’s testimony and evidence clearly establishes that P&H does not supply any GHS, and this evidence was unchallenged on cross-examination.
[224] P&H further submits that there are no precedents in Canadian law for utilizing an imputed price for the merging firms’ specific contribution to the value of a product and for determining the relevant market and the applicable SSNIP on the basis of the so-called “value-added” approach. In fact, adds P&H, the Commissioner has never publicly endorsed the “value-added” approach he is suddenly advancing in this case. Moreover, the Commissioner’s approach would suffer from numerous flaws identified by economists with respect to the “value-added” approach advocated in the US HMEGs.
[225] While Dr. Miller asserts that the price for GHS drives the farmers’ decisions, P&H points to the fact that the testimonies of farmer witnesses on both sides make it clear that farmers base their sale decisions on the posted Cash Price, which is what the grain companies actually compete on.
(3)
Tribunal’s assessment
[226] As the Commissioner rightly indicated, the fundamental disagreement in this case is how to define the product market, and more specifically the relevant product that is being exchanged when the farmers sell their grain to an Elevator. Ms. Sanderson indeed acknowledged that the main issue to be determined by the Tribunal is the choice between the Cash Price for grain and the “imputed” price for GHS.
[227] For the reasons set below, the Tribunal agrees with P&H on this issue and concludes that there are two relevant product markets in the present case, namely, the purchase of wheat and the purchase of canola by P&H.
[228] The Tribunal finds that the Commissioner’s proposed product market (i.e., the sale of GHS) is not grounded in commercial reality and in the evidence, and that in this case, the “value-added” approach advanced by the Commissioner fails on three fronts: on the facts, from a precedential and legal standpoint, and from a conceptual and economic perspective. It fails factually, because there is no clear and convincing evidence to conclude, on a balance of probabilities, that GHS effectively exist as a relevant “value-add” product or “add on feature” that is transacted and to which a price can be attached. The Commissioner created an analytical price for operational activities associated with the purchase of grain by Elevators. It further fails as a legal argument, because the Commissioner’s position finds no support in the authorities he cites, including the US HMEGs, cases in Canada and elsewhere, and his own MEGs. Finally, it fails conceptually, because the Commissioner did not address the issue of the appropriate SSNIP threshold, as described by the current MEGs, despite the fact that adopting a “value-added” approach would profoundly change the HMT framework and the effective SSNIP level used for market definition purposes.
(a)
The product and the price at issue
[229] At its core, four critical elements determine the characteristics and boundaries of a relevant competition market: 1) a product for which there is a source of demand and a source of supply; 2) a price at which the product is transacted; 3) a geography within which the product is transacted; and 4) the extent of available substitution between different sources of demand or supply for the product if there is a sufficient price incentive.
[230] Since products can be said to be in the same market if they are close substitutes (Tele-Direct at p 36), the debate around the product market definition has typically revolved around the issue of substitutability. However, this is not the case here, as the parties are not debating the extent to which there are functional substitutes to selling GHS to the farmers or to purchasing grain by the Elevators. The parties instead agree that there are no substitute products for either GHS or for grain (namely, each of wheat and canola).
[231] The debate around the product market definition in this case is about how to characterize what occurs when a farmer sells grain to an Elevator. The debate does not center on the definition of the product market, but rather on the definition of the relevant product itself, and of the relevant price attached to it.
[232] Before dealing with the product market — or the geographic market —, the Tribunal must first determine what the product at issue is and what its price is. It is important to distinguish the “product” and the “price” from the product “market.” The Tribunal does not dispute that the product market is an analytical framework and an artificial construct, where the Tribunal needs to determine the presence and extent of substitutes (Superior Propane I at para 101). However, the underlying product or the underlying price for it are not analytical or theoretical constructs themselves. The product and the price attached to a product need to be anchored in the evidence and, indeed, tied to commercial reality. Determining the relevant product and the relevant price is an inquiry that must be based on the evidence of each case.
[233] On the facts of this case, the relevant product is not the sale of GHS, but the purchase of grain.
(i)
The notions of product and price
[234] In the legal and economic context relevant to merger analysis, the term “product” refers to the output that a producer (seller) provides to a purchasing customer, or the input that a producer (purchaser) acquires from a supplying customer.
[235] To identify the relevant product for the purpose of defining a relevant competition market, the starting point is to determine what the customer actually buys or sells, and at what price. In the context of a merger, it starts with the product(s) in respect of which, prior to the merger, the merging entities were competitors. In other words, what is the product that P&H and LDC sold or purchased in the marketplace, in competition with one another, prior to the Transaction?
[236] For a product to exist in the economic sense and in the context of the relevant market definition for the purpose of an application under section 92 of the Act, there must be a separate and identifiable supply and demand for it. It must be transacted and it must have a price attached to it. Product market definition is based on substitutability and focuses on demand responses to changes in prices. The SSNIP must therefore be applied to the price that is being received by the sellers or paid by the purchasers of the candidate product (Visa Canada at para 198). In fact, in order to apply the HMT, a relevant price for the relevant product must be identified. The focus is on the price of the good or service effectively being sold or bought. There can be no product, and no price for such product, if the product has no independent existence and market presence. The Tribunal further accepts P&H’s view that an article or a service which would be neither bought nor sold cannot fall within the definition of a “business” under the Act (and hence, the definition of a merger in section 91), because it would not be a product that is acquired, supplied, or otherwise dealt in.
[237] According to the MEGs, the “base price” to be used for market definition purposes and to postulate a price increase in the HMT framework “is typically the prevailing price in the relevant market” (2011 MEGS at para 4.6). This, once again, refers to a price observed in a market. The base price is the price used in the normal course of business, namely, whatever price “is ordinarily considered to be the price of the product in the sector of the industry (e.g., manufacturing, wholesale, retail) being examined” (2011 MEGs at para 4.7). The MEGs therefore make it clear that the relevant price must echo what is effectively occurring in the industry being considered.
[238] The price that has been typically employed in the standard approach to market definition and in the HMT analysis is the “cumulative price,” namely, “the total value of the product when it leaves the stage of the industry in question. This simply represents the sum of input costs plus value added” (Crampton 1990 at p 265, fn 11). The author Crampton (now a judicial member of the Tribunal) also referred to alternative “prices” such as the “value added price” representing the difference between the cost of all inputs and the cumulative price, and he specified that this price “has been employed in situations where the value added is billed as a separate fee, with no mark-up being applied to the product in relation to when the service is performed” (Crampton 1990 at p 265).
[239] However, the Tribunal underlines that price and cost are two different notions. A price is the amount that a customer buying a product or a person supplying one is willing to pay or receive for it. The price is ascertained from the perspective of the customer or the purchaser. On the other hand, a cost is the expense incurred for making a product or offering a service that is sold. It is ascertained from the perspective of the producer. A cost cannot simply be equated with a price, though there is evidently a relationship between the two, since costs have a direct impact on the price of a product.
[240] The Tribunal has found no precedent, and the Commissioner has not referred to any, where it was recognized by the courts or by the economic literature that there can be a “market” for a product that is not sold and purchased as such, but is only a cost component of a final product. The Tribunal accepts that a specific part can be categorized as a separate product and can constitute the basis for a separate product market (see, for example, Canada (Director of Investigation and Research) v Xerox Canada Inc (1990), 33 CPR (3d) 83 (Comp Trib) (“Xerox”)). But a product must have a price attached to it. In other words, the notion of product for market definition purposes and the calculation of a SSNIP implies a transacted price, not solely a theoretical price derived from allocating revenue to the cost of selected activities associated with the purchase or sale of a product.
(ii)
GHS is not a transacted product
[241] The evidence indicates that GHS, as defined by the Commissioner, is not a product that actually exists in the grain industry or that is being supplied by the Elevators and purchased by the farmers: GHS are not transacted or contracted between the farms and the grain companies.
[242] The Tribunal has found no reference to GHS in P&H’s materials nor a description of any service provided by P&H or other grain companies under the label of GHS. Neither the Elevators, the Crushers, nor even the farmers themselves recognize GHS as a separate, identifiable product. It is not even a term of art used in the grain industry. There is no evidence that farmers and grain companies transact on the basis of the sale and purchase of GHS.
[243] More specifically, none of the six farmer witnesses who testified at the hearing on behalf of either the Commissioner or P&H referred to GHS as an actual service they receive from the Elevators and pay for. Farmers do not talk in terms of GHS.
[244] Similarly, the Elevators do not refer to the notion of GHS. Mr. Heimbecker, for P&H, and the witness from Cargill, Ms. Jordan, both confirmed that in the industry, grain companies do not charge farmers for GHS and rather consider the services covered by the Commissioner’s definition of GHS as costs. At one point in his submissions, the Commissioner mentioned the fact that P&H refers to its grain handling and trading business in its own financial reporting documents. However, the Commissioner has pointed to no evidence, whether from P&H or from any other grain company, containing a reference to grain handling services. The “grain handling business and trading” is not to be confused with “grain handling services” (as a product). P&H purchases grain for the purpose of trading it in national and international markets and grain handling is an operational process in the purchase and resale of the grain.
[245] While the Tribunal agrees with Dr. Miller that the participants and competitive constraints at each stage of an industry are distinct, the Tribunal finds that GHS is not a product that is transacted as such between farmers and Elevators, at any stage of the grain handling and trading business. It is an artificial bundle of services that the Commissioner and Dr. Miller have constructed for the purposes of this proceeding.
[246] Mr. Heimbecker testified that P&H does not supply GHS in the post-CWB era (Exhibits CA-R-115 and P-R-116, Witness Statement of Mr. John Heimbecker (“Heimbecker Statement”), at para 115). This evidence was not contradicted and the Tribunal accepts the evidence of Mr. Heimbecker on this point.
[247] In sum, GHS, or the “value-added” product identified by the Commissioner in this case, is not a product that farms and grain companies recognize and effectively buy and sell. Moreover, GHS is not a separate value-added product that farms transact separately from the grain, and for which they could find substitute sources of supply (i.e., sellers offering just the GHS value-added service).
[248] The Tribunal pauses to observe that, from an industry perspective, the notion of GHS advanced by the Commissioner corresponds to a historical nomenclature that disappeared when the CWB was dismantled on July 31, 2012 (Heimbecker Statement at paras 113–117). Before its dismantling, the CWB purchased certain types of grain — including wheat — from farms and grain companies handled that grain on behalf of the CWB on a “toll basis,” effectively providing the type of services included by the Commissioner in his definition of GHS. However, with the end of the CWB, grain companies are no longer intermediaries between the CWB and farmers, and the market interrelation between the farmers and the grain companies has changed. As a result, grain companies have assumed part of the role of the CWB in the supply chain, namely, as purchaser of grain with the risks attached to the marketing and sale of the product. With the end of the CWB, the historical tariffs and fees that used to be charged for a service also came to an end. In the post-CWB world, P&H (like other grain companies) buys wheat and canola from farms, taking title to the grain at the time the farm delivers it to the Elevator. At that point in time, the farm receives the contracted Cash Price for its grain and ownership of the grain passes to P&H or the other Elevators. From that point on, the farm has no right or interest in the grain and bears no risk in relation to the purchase transaction. Instead, P&H is fully responsible for the costs, risks, and rewards of aggregating, transporting, and selling the grain to a customer.
[249] True, as will be discussed below, the farmers and Elevators sometimes negotiate on one component of the Cash Price (namely, the Basis) because information on the Basis is provided by the Elevators. The Tribunal is satisfied that this reflects a historical heritage from the CWB days (or, as argued by P&H, a linguistic vestige of the CWB era). But what were services at the time of the CWB are now costs to the Elevators, not GHS to which a price can be attached.
[250] For the purpose of this Application, the only products that are exchanged and transacted between the farms and the Elevators are wheat and canola.
(iii)
There is no price for GHS
[251] Furthermore, the Tribunal concludes that there is no “price” associated with GHS.
[252] As was the case for GHS as a product, the Tribunal finds no evidence of a “price” for GHS. The evidence from the farmer witnesses indicates that they do not talk about receiving GHS and are not charged any fees or price for GHS. All farmer witnesses testified that there are no separate charges for GHS. In addition, no price exists for GHS in the contracts or agreements concluded by P&H and other grain companies with the farmers.
[253] Nor is a price for GHS recorded or kept in the transaction data reported by the grain companies in the usual course of business. The transaction data obtained from the grain companies provides address information for farms delivering canola or wheat to the Elevators and Crushers, with some exceptions. The farm address information allowed Dr. Miller and Ms. Sanderson to identify deliveries to the Elevators from farms with addresses in the various regions used in their respective analyses, as well as any other location. The transaction data report Cash Prices and deliveries to each grain company but they do not include any information on GHS or on a price for such alleged bundle of services.
[254] It is of note that, just as they do not consider GHS to be a separate product offered for purchase and sale, the grain companies do not keep an amount representing the price of GHS — or even the Basis, as will be discussed below — in their transaction data set. If GHS were meaningful for the Elevators or the farmers from a transactional perspective, the grain companies would likely keep track of its price in their transaction data.
[255] Indeed, the Commissioner recognized at the hearing and in his written submissions that the farmers are not charged for GHS as a separate, added service to the purchase of grain. Moreover, a witness for the Commissioner in the discovery process and Dr. Miller in his testimony each confirmed that the price for GHS had to be “imputed” and that it is a constructed price.
[256] As GHS is neither observed as an added service feature nor transacted as such, Dr. Miller had to create a measure for the price of it. In his expert report, Dr. Miller imputed the price of GHS for each transaction as the difference between the Cash Price and the Futures Price from the financial market, as of the transaction date, adjusting for exchange rates in the case of wheat. Dr. Miller sometimes referred to the price of GHS as the Basis, although he admitted that his constructed price of GHS is not equal to the Basis used by Elevators for wheat or canola for each transaction. Dr. Miller’s transaction-level prices of GHS were estimated with error, in part because the contract date (which often determines the relevant Futures Price) is unobserved in the transaction data.
[257] In sum, there is no observed or observable added feature for GHS, no price is attached to it, and none appears in the transaction data or can be otherwise identified. The evidence from the Commissioner reveals that the “imputed” price for GHS is a constructed price, and not an explicit or implicit price transacted between buyer and seller. Here, there is no doubt that it is not a situation where a farmer pays a price for an actual specific service, such as delivery, transportation or processing fees.
[258] The farmer witnesses (i.e., Mr. Lincoln, Mr. Wagstaff, Mr. Pethick, Mr. Paull, Mr. Duncan, and Mr. Hebert) all testified that the Cash Price they ultimately receive for their grain drives their decisions to sell their grain to an Elevator. As admitted by the Commissioner, farmers care about the money they receive from the sale of their grain, and they never talk about GHS as separately transacted.
[259] In his witness statement, Mr. Pethick stated that he considers the “basis price” but also the overall price (i.e., the Cash Price) communicated by the Elevators when deciding where to deliver his grain (Exhibit P-A-001, Witness Statement of Mr. Alistair Pethick (“Pethick Statement”), at paras 19–20). Similarly, Mr. Lincoln also referred to the importance of “competitive prices” and to being offered a “higher price,” generally, without suggesting that this was limited to the Basis (Exhibits CB-A-025 and P-A-026, Witness Statement of Mr. Chris Lincoln (“Lincoln Statement”), at paras 13, 15). The Agreed Statement of Facts for the two farmer witnesses who did not testify at the hearing indicates that XXXXX “regularly checks the prices” at various Elevators, without limiting his comment to the Basis (Exhibits CA-R-242 and P-R-243, Compilation of Additional Documents Added to Agreed Book, Agreed Statement of Facts re XXXX and XXXX). For their part, each of Mr. Hebert, Mr. Paull, and Mr. Duncan said in their respective witness statements that the Cash Price is what drives their decisions to sell to Elevators or Crushers.
[260] The evidence from Mr. Heimbecker also reveals that most pricing-related communications made by P&H to farmers refer to the Cash Price, with no mention of the Basis (Heimbecker Statement at para 61; Exhibits CB-A-134 and P-A-135, Read-in Brief of the Commissioner (“Commissioner Read-In”)). Industry participants also use the Basis and the Cash Price interchangeably, to refer to the net final price that farmers effectively receive for their grain. Mr. Hebert, for example, testified that when he is using the Basis, he is “calculating it back to the cash price that [he is] going to deposit” (Consolidated Transcript, Confidential B, at p 870). In short, he is using the Basis to compare the Elevators’ net prices for the grain.
[261] In the Tribunal’s view, on the facts of this case, the only “base price” for market definition purposes is the Cash Price paid to farmers to purchase their wheat or canola, not the “imputed” price of GHS that is never separately or identifiably charged to farmers by Elevators to handle their grain. Farms sell grain to Elevators and Crushers for a single Cash Price. Elevators and Crushers purchase grain from farms for a single Cash Price. The Cash Price paid to farms to purchase grain is the ordinary and prevailing price in the relevant markets.
(iv)
The Basis differs from GHS and is not the price of GHS
[262] In his submissions on the product market and on GHS, the Commissioner relied heavily on the Basis and on the evidence related to it.
[263] The Commissioner argues that the Basis is a component of the Cash Price, along with the Futures Price. Both combine to form the Cash Price paid to the farms. According to the Commissioner, the Basis component of the Cash Price is the mechanism that P&H uses to ensure that the price at which it buys the grain from the farmers is a price that allows it to cover the expenses of operating the Elevator where the farmers deliver their grain. The Commissioner maintains that the Basis “is an amount subtracted from the Futures Price in the case of canola (and added to the Futures Price in the case of wheat to account for exchange rate differences) that covers the grain company’s costs to operate the Elevator while also providing the grain company with a margin.” He claims that it is not a simple mathematical construct, as alleged by P&H and Mr. Heimbecker. According to Dr. Miller, the Basis covers the costs of the Elevators and allows for a profit. Dr. Miller claims that it is the relevant price “in terms of how competition plays out” in the grain industry, “in terms of the value added” by the Elevators, and “in terms of just the profitability of the [grain handling] business” (Consolidated Transcript, Public, at p 1430).
[264] The Commissioner further submits that all contracts entered into by the Elevators and the farms refer to the Basis, or even to the “basis price,” and that farmers use the Basis in their dealings with the Elevators. The Commissioner adds that the problem with GHS and the Basis is an “implementation issue” as neither the Basis nor the price of GHS appear in the transaction data of the grain companies. In sum, throughout his submissions, the Commissioner effectively equates GHS to the Basis to justify his approach to the product market definition.
[265] The Tribunal is not persuaded by the Commissioner’s arguments and finds that the existence of the Basis is not sufficient to transform GHS into a relevant product to which a relevant and reliable base price can be attached. The evidence does not support that the Basis can be equated with the notion of GHS advanced by the Commissioner and Dr. Miller. There is a distinction and a difference between the use of the Basis by farmers and grain companies in the sale and purchase of grain on the one hand, and GHS as an operational cost to the grain companies on the other.
[266] The Tribunal acknowledges that the Basis is an industry benchmark, recognized and used by the participants in the grain handling business. But the Basis is not a product that is transacted as such. The Basis is not a price either. Despite the repeated references made by the Commissioner to a “basis price” in his oral and written submissions, the Tribunal is not persuaded that, on a balance of probabilities, the evidence supports a conclusion that the Basis represents a price attached to a product. Mr. Heimbecker testified that the Basis is not a price expressed in dollar terms, even though there is often a dollar sign apposed to its numerical figure in some of P&H’s own documents. In the Tribunal’s view, the Basis is best described as the numerical difference between the Cash Price and the Futures Price, as determined by P&H’s Workback Algorithm. The Workback Algorithm is only run once a day by P&H, but the Cash Price that a farmer sees in P&H Direct will adjust instantaneously to reflect any changes in the Futures Price. This evidence has not been contradicted, and was confirmed by farmers such as Mr. Duncan and Mr. Hebert in their respective testimony (see, for example, Consolidated Transcript, Confidential B, at p 929). Yes, the Basis is a metric used in the grain business, but it is not a price for a product.
[267] One of the farmer witnesses, Mr. Paull, testified that the Basis is just a tool used to track or monitor the Cash Price that he will receive when he delivers his grain to the Elevators. One of P&H’s customer service representatives, since promoted to managing an Elevator, explained the Basis in emails sent to hundreds of customers as follows: “[t]his premium or discount to the futures value is commonly referred to as a basis. The basis reflects each grain company’s own particular handling, transportation and marketing costs, combined with the bid values from their own-end use customers” (Exhibit CB-A-149, P&H Email Subject: Gain From You Grain, dated February 16, 2017).
[268] The Tribunal observes that the Basis is not identified by the farms or the grain companies as a price associated with GHS or with any specific service. The grain industry participants instead refer to the Basis, without more. It is never the “Basis for GHS” or for any other specific product. Nowhere in the evidence is there a reference to a Basis for something. It is a concept and a notion used in the grain industry, but there is no evidence that the Basis refers to or is attached to a particular product. As is the case for GHS, and even though it is an information communicated by the Elevators in their dealings with the farms, the Basis is not recorded in the transaction data by the grain companies.
[269] Moreover, when the farmers refer to the effect of changes in the Basis, they always refer to changes expressed in terms of cents per bushel or per MT. The bushel or the MT is a unit of measurement for the grain, not for GHS. In other words, the farms always refer to the impact that a change in the Basis will have on the price for the grain itself, expressed in terms of a dollar value per bushel or MT. The changes to the Basis are never expressed in terms of cents for GHS or for any type of service provided by an Elevator.
[270] Even if a change to the Basis offered is the subject of negotiations between grain companies and farmers, what ultimately changes is the Cash Price from which the Basis is derived. What changes is the price expressed in dollars and cents per bushel or MT, for the grain being purchased by the Elevator and sold by the farmer.
[271] Further to its review of the evidence, the Tribunal therefore concludes that the Basis is best described as a component of the price of grain, not as a price attached to a specific product or a separate value added. In fact, the Tribunal finds that the evidence from P&H and the grain companies establishes that the Basis is not a price of a product by itself but a cost element for the Elevators.
[272] The Tribunal adds that merely displaying the components of a price, itemizing the price of a product, or providing the breakdown of the components of a product being sold or purchased, does not have the effect of creating a separate product if such an itemized component does not have an economic life of its own, and is not transacted in a market for a price. This is exactly the case for the Basis and the GHS.
[273] In his final argument, the Commissioner submitted that this is not even a case where the price for GHS is implicit. He claims that there is nothing implicit about GHS and that its price or value appears in the contracts that P&H and other grain companies enter into with farmers. The Commissioner claims that the price for GHS is in fact explicit and only appears to be implicit because the Basis is not subsequently recorded in the transaction data of the grain companies. The Commissioner argues that the explicit price underlying his product market is the Basis being charged to the farmers by the Elevators. And that the problem is strictly one of implementation because the Basis sets the price for GHS.
[274] The Tribunal does not agree. This is not a situation where there is an explicit price for GHS that is part of the contract price between the farmers and the Elevators.
[275] First, the Commissioner confuses the price of GHS and the Basis. The evidence clearly establishes that the Elevators do not post a price for GHS; they post a Basis. The Basis sometimes shows up in the information posted by P&H and other grain companies, as well as on some invoices and in some contracts. But not GHS. The Basis and GHS are not the same thing, and Dr. Miller clearly acknowledged that he had to “impute” the price of GHS from the Basis information (Exhibits P-A-169, CA-A-170 and CB-A-171, Expert Report of Dr. Nathan Miller (“Dr. Miller Report”), at para 173). Dr. Miller used a computed price as a proxy for the Basis, because neither the Basis, nor a price for GHS, appears in the transaction data. Dr. Miller and a representative of the Commissioner indeed acknowledged, in their testimony and on discovery, that the imputed price for GHS is a “subset of the Basis” (Exhibits CA-R-242 and P-R-243, Compilation of Additional Documents Added to the Agreed Book, P&H Read-in Brief (“P&H Read-In”), at pp 21–22). Dr. Miller further admitted that the imputed price for GHS does not always correspond to what the Basis component of the Cash Price actually is.
[276] Second, a calculated or imputed price is not a price that can be correctly described as being observed or observable. Here, Dr. Miller has not used the observed Basis; he has estimated a price for GHS (which he sometimes refers to as a Basis) for each transaction from the difference between the Cash Price and an assumed Futures Price at a given date. Dr. Miller testified that at a transaction level, the estimation leads to variance between the observed Basis of which a corresponding Cash Price was transacted and the constructed price for GHS for that transaction.
[277] The Commissioner submits that the price of GHS imputed by Dr. Miller is a good approximation of the Basis. For the Virden and Moosomin Elevators, Dr. Miller takes the median of the transaction-level prices of GHS and he uses the Virden Elevator median price to calculate his mark-up. The median prices calculated by Dr. Miller are the benchmark prices he uses in his HMT analysis.
[278] Dr. Miller and Ms. Sanderson agree that taking a median over thousands of observations measured with error reduces the measurement error. However, they disagree about whether the median prices for GHS obtained by Dr. Miller for the Virden and Moosomin Elevators are reliable estimates for the Basis.
[279] Referring to the contracts of XXXXXX, a farmer, Dr. Miller said that the difference between his computed median price of CAD $31.02 per MT for GHS for wheat and the corresponding value of the Basis in XXXXXXx contracts of CAD $34.83 per MT was “reasonably close,” using 37 deliveries made by XXXXXX. The Tribunal observes that it is still a significant difference of 12% — namely, CAD $3.76 per MT or 10.23 cents per bushel of wheat. And that difference is a variance of only a few transactions out of thousands. However, when Dr. Miller compares the median for all wheat transactions at the Moosomin Elevator, calculated by his own model as being CAD $34.78 per MT, to the actual Basis in the contracts of XXXXXX (i.e., CAD $ 34.83), the two values are very similar. In light of that, the Commissioner claims that the price for GHS imputed by Dr. Miller represents a good approximation of the actual Basis found in the contracts.
[280] Ms. Sanderson testified that the large difference in the median between the Moosomin and Virden Elevators for both wheat and canola is likely due to a larger error between Dr. Miller’s predicted median and the median of the actual Basis. Ms. Sanderson points to the fact that for wheat, the GHS median value is 20% lower at the Virden Elevator than at the Moosomin Elevator, whereas in the case of canola, the Virden Elevator median value is 60% higher than the Moosomin Elevator value, in a situation where the Virden Elevator had larger sales than Moosomin. Ms. Sanderson attributed these differences to the uncertainties in the methodology used by Dr. Miller.
[281] Like Ms. Sanderson, the Tribunal does not share the Commissioner’s confidence in Dr. Miller’s calculations. The Tribunal is not persuaded that using a median to correct for the error terms between the estimated and the actual values is sufficient, in this case, to correct for the high uncertainty in the estimated values. The median calculated by Dr. Miller includes several sources of uncertainty stemming from the fluctuation of the Futures Price and exchange rates within a day that may not be correlated with the actual transactions, the choice for delivery dates, and negotiations on delivery. The Tribunal agrees that in theory, using the median of a very large data set (in this case, thousands of transaction data) provides a much smaller error than the error of each observation. However, in his attempt to correlate a Futures Price for an estimated delivery date for each transaction, Dr. Miller did not provide convincing evidence to give the Tribunal the level of confidence or precision needed to conclude that the median predicts the actual Basis. The Tribunal adds that if the measurement errors are systematically skewed from actual delivery periods or Futures Price and exchange rates spot price reflected in the application at the time of sale, the results will also be skewed. Dr. Miller did not provide the Tribunal with evidence that this did not occur.
[282] In any event, whether Dr. Miller’s “imputed” price of GHS is a good or not so good approximation of the Basis is not a determinative issue in this case. The Tribunal finds that neither the “imputed” price of GHS nor the Basis can be used as the foundation of a product market definition since neither is a transacted price attached to an identifiable product.
[283] The Tribunal pauses to note that, if in a given case, the evidence enabled it to isolate or identify a value-added component or feature of a product with sufficient precision, and to find a value-added price from business records, the Tribunal may conclude that the computed price is the price of that value-added component. In this case, the evidence does not permit to make such a conclusion.
(v)
The Basis plays a role in competition
[284] That being said, the Tribunal appreciates that the Basis is a touchstone of competition between the Elevators, and that it is an important factor to understand the rivalry between Elevators as well as the competitive dynamics in the grain industry. One cannot ignore what is going on at the Elevator’s level (including competition on the Basis) when determining whether there is an increase in market power and a substantial lessening of competition.
[285] The evidence supports a conclusion that the Basis component of the price of grain is affected by local competition. Farmers such as Mr. Pethick, Mr. Paull, and Mr. Hebert testified that they can play Elevators against each other based, among other things, on the Basis. Mr. Paull and Mr. Hebert admitted on cross-examination that they use not only the Cash Price to compare Elevators, but also the Basis. Even Mr. Heimbecker admitted that farmers use the Basis to compare grain prices between Elevators. There was also evidence that P&H sends email blasts to customers containing references to the Basis. The vast majority of grain companies also incorporate references to the Basis as well as the Futures Price and the Cash Price in their contracts with farmers. Farmers can also enter into Basis contracts, though Mr. Heimbecker testified that these were rare in the grain industry. Furthermore, P&H can and does adjust the Basis to account for local competition from Elevators: P&H and other grain companies will offer limited-time or limited-tonnage specials to farms to attract their grain, and such premiums above the posted Cash Price are sometimes expressed through changes in the Basis.
[286] The Tribunal accepts that the Basis captures an aspect of competition between the Elevators. The Basis is one part of competition between the Elevators and, as will be discussed below, it is an element to consider in the substantial lessening of competition analysis. However, this does not mean that the Basis, or Dr. Miller’s “imputed” price for GHS, can constitute the base price for market definition purposes, or that the existence of the Basis is sufficient to transform GHS into a relevant product.
[287] The Commissioner is urging the Tribunal to adopt, for its product market definition, an economic framework that allows for and facilitates an analysis focused on the local competition allegedly impacted by the Acquisition. This is what led the Commissioner and Dr. Miller to identify GHS as a product and to “impute” a price for it, derived from the Basis. Dr. Miller opined that, economically, constructing an imputed price to approximate the Basis was the right way to analyze the “effects” of the Virden Acquisition. However, in an effort to define a relevant product market, the Commissioner developed a theoretical framework that does not reflect the commercial reality of the grain industry, where GHS does not exist as a product, does not have a “price,” and is not transacted. In short, the Commissioner’s position ignored the fundamental premise of product market definition, which requires the existence of a transacted product whose price is the ordinary price in the sector of the industry being examined.
[288] In sum, the Tribunal finds that the evidence does not support the existence of a separate relevant “market” for the sale of GHS. GHS is an artificial product, with an artificial price, that cannot form the foundation of an acceptable relevant product market for the purpose of the Tribunal’s analysis. On the facts of this case, the relevant products are the purchase of wheat and canola and the relevant price is the Cash Price.
(b)
The “value-added” approach
[289]
The Commissioner submits that the law acknowledges it can be appropriate to define the relevant product market around a component of the price of the final product which represents the specific value provided at an intermediary level by the merging firms. In this case, the Commissioner maintains that an Elevator’s specific contribution of value, namely, providing GHS as part of the Cash Price farmers receive for their grain, can be identified with reasonable clarity. According to the Commissioner, a product market definition focused on the competition affected by the merger and on the merging firms’ contribution to value is an approach supported by the jurisprudence from Canada, the U.S., and the European Union.
[290]
The Tribunal disagrees. Further to its analysis, the Tribunal instead concludes that there are no precedents, in Canada or in any other jurisdiction, where the “value-added” approach referred to in the US HMEGs has actually been used and applied by an adjudicating court or tribunal to define a relevant product market in a merger case. Moreover, the “value-added” approach advocated by the Commissioner cannot be reconciled with the history of the MEGs and the current 2011 MEGs issued by the Commissioner himself.
(i)
The US HMEGs
[291]
In support of his “valued-added” approach to the product market definition, the Commissioner relies heavily on the US HMEGs. These guidelines describe the approach to merger review by the U.S. antitrust agencies and were adopted in 2010. In his reply expert report, Dr. Miller referred specifically to them, drawing an analogy between the present case and the examples used in the US HMEGs (Exhibits P-A-172, CA-A-173 and CB-A-174, Reply Expert Report of Dr. Nathan Miller (“Dr. Miller Reply Report”), at paras 34–36 and fn 33).
[292]
In its section 4 on “Market Definition”, the US HMEGs contemplate that in certain situations, the benchmark price used for analyzing a product market — i.e., the base price — can be different from the price effectively charged for a final product. The US HMEGs indicate that, in a situation where explicit or implicit prices for the merging firms’ contribution to the value of the final product can be identified with reasonable clarity, the price used to assess the merger and define the relevant market can be the component of the price that the firms compete on. Similarly, the SSNIP threshold can also be adjusted. In paragraph 4.1.2 on “Benchmark Prices and SSNIP Size,” the US HMEGs state as follows:
The Agencies most often use a SSNIP of five percent of the price paid by customers for the products or services to which the merging firms contribute value. However, what constitutes a “small but significant” increase in price, commensurate with a significant loss of competition caused by the merger, depends upon the nature of the industry and the merging firms’ positions in it, and the Agencies may accordingly use a price increase that is larger or smaller than five percent. Where explicit or implicit prices for the firms’ specific contribution to value can be identified with reasonable clarity, the Agencies may base the SSNIP on those prices.
[Emphasis added.]
[293]
The US HMEGs then refer to three specific examples, illustrating situations where implicit prices for the firms’ specific contribution to value can be identified. These examples read as follows:
Example 8: In a merger between two oil pipelines, the SSNIP would be based on the price charged for transporting the oil, not on the price of the oil itself. If pipelines buy the oil at one end and sell it at the other, the price charged for transporting the oil is implicit, equal to the difference between the price paid for oil at the input end and the price charged for oil at the output end. The relevant product sold by the pipelines is better described as “pipeline transportation of oil from point A to point B” than as “oil at point B.”
Example 9: In a merger between two firms that install computers purchased from third parties, the SSNIP would be based on their fees, not on the price of installed computers. If these firms purchase the computers and charge their customers one package price, the implicit installation fee is equal to the package charge to customers less the price of the computers.
Example 10: In Example 9, suppose that the prices paid by the merging firms to purchase computers are opaque, but account for at least ninety-five percent of the prices they charge for installed computers, with profits or implicit fees making up five percent of those prices at most. A five percent SSNIP on the total price paid by customers would at least double those fees or profits. Even if that would be unprofitable for a hypothetical monopolist, a significant increase in fees might well be profitable. If the SSNIP is based on the total price paid by customers, a lower percentage will be used.
[294]
The US HMEGs therefore contemplate two situations where the usual base price and/or SSNIP threshold could be modified. The first one looks at the explicit or implicit price of the value added by the merging firms and, in cases where the value-added component can be identified with reasonable certainty, a base price other than the usual total price of the product may be used. The second refers to situations where the SSNIP remains based on the total price paid for the product, but the usual 5% SSNIP threshold is adjusted to deal with the realities of an industry.
[295]
However, the US HMEGs provide no guidance as to when the contribution to value, as opposed to the total price of a product, would be appropriate for the purposes of the relevant market definition. Nor do they contain any guidance on how and when the usual 5% SSNIP threshold should be adjusted. The Tribunal finally notes that the US HMEGs simply mention that the U.S. antitrust agencies “may base” the SSNIP and the HMT analysis on the price for the value added by the merging firms.
(ii)
No court or tribunal has ever applied the “value-added” approach
[296]
As correctly pointed out by P&H in its submissions, a review of the existing jurisprudence reveals that, contrary to the Commissioner’s submissions, there are no legal precedents, in Canada, in the U.S., or in any other jurisdiction, where a court or a tribunal has effectively applied or recognized the “value-added” approach set out in the US HMEGs. While the concept advanced by the Commissioner has been argued in a few cases in the U.S., the European Union and Australia, it has either been rejected by the courts or applied by competition agencies to facts which are distinguishable from the facts of this case. The Tribunal further observes that the “value-added” approach has never been considered or applied in respect of an “imputed” price. In other words, the Tribunal has not found any situation where a court or tribunal has accepted and retained an implicit price for the merging firms’ specific contribution to the value of a product to determine the relevant product, the relevant price, the SSNIP, and the relevant market in a merger case.
[297]
In essence, on this issue of the “value-added” approach, the Commissioner is asking the Tribunal to go where no other court or tribunal has yet agreed to go.
·
Canada
[298] As far as Canadian cases are concerned, the Commissioner relies on Hillsdown and Xerox. He claims that these cases both demonstrate that the value added by the merging firms can form the basis of a product market definition and that, to define the relevant product market, the Tribunal can focus on the portion of the final price that is impacted by the merger.
[299] The Tribunal is not persuaded by the Commissioner’s arguments.
[300] In Hillsdown, the first contested merger review proceeding under section 92, the Tribunal analyzed the merger of two companies that operated rendering businesses for the by-products from slaughterhouses and other entities. On the product market definition, the Tribunal had to decide whether to characterize the relevant market as the supply of “renderable material” from slaughterhouses to the renderers, or as the provision of the specific services that contribute to the end product’s value, namely, “rendering services” offered by the renderer to the slaughterhouses, meat processing plants, grocery stores, etc. Since the Tribunal decided to characterize the market as the provision of rendering services, claims the Commissioner, it recognized that the product market can be limited to the services capturing the value provided by an intermediary.
[301] The Tribunal considers that Hillsdown is of no assistance to the Commissioner on the “value-added” approach since in that case, the Tribunal expressly said that there was no difference between the two contemplated approaches to the product market, even though it decided to characterize the market as the provision of rendering services. If the first characterization was used, then the analysis for competition purposes would have focused on the possible monopsony power of the renderers as buyers of the raw materials. If the second characterization was used, then the analysis would have focused on the possible market power of the renderers as sellers of the rendering services. But, said the Tribunal, no significant difference resulted from the two characterizations. Moreover, there was clearly a price paid for the renderable materials or for the services provided (Hillsdown at pp 293 d-h, 299). In light of the foregoing, the panel finds that the Hillsdown decision is at best inconclusive on the issue of the “value-added” approach.
[302] In Xerox, a non-merger case, the Tribunal found that the relevant product market was the provision of intermediary services, namely, servicing copier parts that were not constrained by the sale of copiers to end customers. It is true that the Tribunal then recognized that a subset of a final product (i.e., the servicing of parts for copiers) can constitute a relevant product. However, in that case, there was a specific price charged for the specific service offered by the parties. Again, this precedent therefore offers at best weak support to the “value-added” approach contemplated by the Commissioner in this case, as no specific price is charged for GHS.
·
U.S.
[303] Turning to the U.S., even though the US HMEGs were adopted more than 12 years ago in August 2010, the Commissioner has not referred the Tribunal to any decision of a U.S. court where the “value-added” approach set out in the US HMEGs was accepted and actually applied.
[304] In his submissions to the Tribunal, the Commissioner referred to a matter involving Conagra Foods (United States of America v Conagra Foods, Inc, et al, 14-CV-000823 “Conagra”)) and to FTC v Whole Foods Market, Inc, 502 F.Supp. 2d 1 (2007) (“Whole Foods”). The Tribunal concludes that neither of these matters constitutes a convincing precedent to assist the Commissioner on the “valued-added” approach.
[305] The Conagra example was a case in the flour business, where a fee was charged by the millers for converting wheat into flour. The Commissioner claims that this was a situation where the U.S. Department of Justice applied the SSNIP to a component of the final price for the flour, namely, the converting fee, because this component was the subject of the effective competition at issue between the merging flour mills. However, this precedent is of no value to the Tribunal as the Commissioner was strictly relying on the argument presented to the court by the U.S. antitrust agency itself, and not on a decision issued by a U.S. court. This case is therefore not a legal authority but instead solely reflects the position taken by another competition authority on the “value-added” approach.
[306] With respect to Whole Foods, it was not a case where a “value-added” approach was applied or even considered by the U.S. court. It was instead a matter where the court mentioned that in the context of the HMT framework, lower SSNIP levels may be more appropriate for mergers in markets or industries characterized by high-volume sales but low-profit margins (Whole Foods at pp 9–11). This case therefore did not involve the application of a “value-added” approach to a component of a product but turned instead on the possible use of a smaller SSNIP threshold on the final product sold by the grocers. The Tribunal points out that in Whole Foods, the U.S. court did not analyze nor provide any guidance on the factors to take into account or the evidence required in order to determine the appropriate level of such a lower SSNIP level.
·
European Union
[307] The Commissioner also referred to two cases issued by the European Commission (“EC”) in the European Union, where the US HMEGs were approved and followed by the EC. These decisions dealt with extruded metals and, according to the Commissioner, recognized the value-added by an intermediary as a relevant and separate product market.
[308] In Norsk Hydro/Orkla/JV (Case No COMP/M.6756), 13 May 2013 (“Norsk Hydro”), two companies had operations in the aluminum sector where extrusion premia were charged. The extrusion premia represent the price paid by customers for the value added by companies that extrude the aluminum. In Norsk Hydro, the EC noted that there was a significant and persistent difference in the extrusion premia charged by soft-alloy extrusion suppliers in two different geographic markets. Therefore, in analyzing the merger between extruders, the EC considered that the relevant benchmark price was the actual price for the premium charged for the extruding process, and not the full price of the aluminum eventually sold to customers. The EC concluded that “in the presence of a similar price structure, it seems appropriate to take as a relevant benchmark price the extrusion premia rather than the full price” — citing with approval the US HMEGs (Norsk Hydro at paras 66–67). The EC found that the price negotiations were only on the extrusion premium, as other factors such as the price of aluminum or the billet conversion costs were typically fixed.
[309] The second European Union case relied on by the Commissioner is Inco/Falconbridge, (Case No COMP/M.4000), 4 July 2006 (“Inco”). In that case, the EC considered the market for high-purity nickel, stating that “the price increase must be seen in relation to the added value provided by the firms in the relevant market” (Inco at para 379). The EC observed that a price increase of an input good may have only a minor effect on the price of the final product (depending on the share of total input cost represented by the input good), but can nonetheless be considered significant from a competition perspective. Since the premium charged was directly negotiated with the customers, the EC determined that it was a separate product, with a separate price.
[310] The Tribunal pauses to note that P&H also referred to a third similar EC matter, also dealing with extruded metals, namely, Glencore/Xstrata (Case No COMP/M.6541), 22 November 2012. In that case, a premium applicable to the extruded metal differentiated the extruded product from the raw metal price, and was also the main element of the price negotiated between buyers and sellers.
[311] The Tribunal finds that these European Union cases are of no real precedential value. In any event, they are clearly distinguishable from the present case.
[312] Regarding their precedential value, the Tribunal underlines that these three matters are decisions issued by the EC, which is the European competition agency and the equivalent of the Commissioner in the European Union. These EC decisions are not decisions issued by an independent adjudicative court or tribunal. Under the European Union competition law regime, and in merger matters in particular, the EC is not strictly an investigator and law enforcer as is the Commissioner under the Act. The EC has the dual role of being not only the investigative authority but also the first-instance decision-maker. The EC decisions relied on by the Commissioner thus represent the position of the competition authority itself, as opposed to a decision by an independent judicial body like the courts or the Tribunal. The Tribunal is thus of the view that such decisions of the EC, while informative, carry less persuasive weight. They can certainly not be qualified as legal precedents on the issue of the “value-added” approach.
[313] The Tribunal further concludes that the Norsk Hydro and Inco precedents are of limited assistance to the Commissioner as they are distinguishable from the present case: according to the evidence in those cases, the premia were a separate price charged for a separate product, and were openly negotiated between the suppliers and the customers. These were situations where there was a specific price for the premia at issue. In other words, the price of the value-added product was not constructed or implicit. It was explicit and transacted.
·
Australia
[314] In its final submissions, P&H referred to an Australian case, Australian Competition and Consumer Commission v Metcash Trading Limited, [2011] FCA 967 (“Metcash”), aff’d [2011] FCAFC 151. This was in response to a Direction the Tribunal had issued prior to final oral argument, where it notably invited counsel for both parties to identify legal precedents in which a court or tribunal considered situations where it was proposed that an implicit price for the merging firms’ specific contribution to the value of a product should be used to determine the SSNIP and the relevant market, as well as legal precedents which considered situations where in applying the HMT and in defining the relevant market, a SSNIP smaller or larger than 5% should be used, and/or reviewed the factors to take into account in determining the appropriate level of the SSNIP.
[315] In the Metcash case, the Australian Competition and Consumer Commission (“ACCC”) challenged a merger of wholesale grocery suppliers. The ACCC sought to use an imputed price purporting to represent the value added by grocery wholesalers at their stage in the supply chain as the basis for considering the impact of a SSNIP in defining markets. The ACCC’s approach, which echoed the US HMEGs, was soundly rejected by the Federal Court of Australia (and affirmed on appeal). The Australian court found that the purported value-added services for which the ACCC sought to use an imputed price were not the extent of what the wholesaler actually provided, and that an imputed price solely for these services could therefore not be used as the basis for defining the relevant market. The court noted that “the associated services provided by [the wholesaler] are not available in the absence of the acquisition by a retailer of packaged groceries from [the wholesaler]” (Metcash at para 196).
[316] The Commissioner argues that this case is distinguishable as it involves a considerably more complex market structure, in an industry characterized by a mixture of self-supplying retailers and independent retailers, and involving different categories and types of products across multiple levels of the supply chain.
[317] The Tribunal does not share the Commissioner’s view. As admitted by counsel for the Commissioner at the hearing, Metcash is the sole judicial precedent identified by the parties that actually dealt directly with the so-called “value-added” approach exposed in the US HMEGs and advocated by the Commissioner in the present case. In Metcash, the Australian courts clearly rejected the “value-added” approach to product market definition and the use of an imputed price covering only a subset of the product effectively sold by the merging parties, as it did not reflect the commercial activity of the merging firms. Moreover, the alleged value added by the grocery wholesalers was not independently transacted.
[318] In the Tribunal’s view, this Australian case bears a number of striking similarities with the present case: like the situation in Metcash, the so-called value-added for GHS does not have a commercial life of its own, and the “imputed” price for GHS — or the Basis — is not the price ordinarily used in the grain industry for the product being transacted between the farmers and the Elevators. The Tribunal considers that the reasoning in this Australian decision, while not binding, is persuasive and is generally consistent with the Tribunal’s analysis in the present case.
(iii)
The “value-added” approach is not supported by the Commissioner’s own MEGs
[319] The Commissioner finally submits that the “value-added” approach he is proposing finds support in his own MEGs.
[320] With respect, the Tribunal again does not agree. The Tribunal instead finds that the Commissioner’s proposed approach to the application of the HMT and the choice of the price of GHS for market definition purposes cannot be reconciled with the removal, in 2004, of the “value-added” language that existed before then and the continued absence of such language in the current 2011 MEGs.
[321] It is worth reminding that the MEGs articulate the analytical framework that the Competition Bureau and the Commissioner apply in determining whether a merger is likely to substantially lessen or prevent competition. The MEGs were first introduced in 1991 to provide guidance to the Competition Bureau’s enforcement approach to the new merger provisions enacted in 1986 (“1991 MEGs”). The Competition Bureau’s 1991 MEGs were superseded in September 2004 with the release of the 2004 MEGs, which were themselves replaced by new, revised guidelines in October 2011, when the 2011 MEGs were issued. The Tribunal pauses to note that the 2011 MEGs followed the issuance of the revised US HMEGs in August 2010.
[322] Since the first adoption of the MEGs in 1991, several important changes have been made to the product market approach presented by the Commissioner. More specifically, the “value-added” approach, which the Commissioner is apparently attempting to resuscitate in this case, was expressly abandoned in the most recent iterations of the MEGs.
[323] In the 1991 MEGs, the Commissioner discussed the conceptual framework for market definition. With respect to the base price, the 1991 MEGs stated the following (1991 MEGs at p 9):
In general, the base price that is employed in postulating a significant and non-transitory price increase is whatever is ordinarily considered to be the price of the product at the stage of the industry (e.g., manufacturing, wholesale, retail) being examined. This is typically the cumulative value of the product, inclusive of the value added (mark-up) at the industry level in question. However, in certain industries, the value added is billed as a separate fee, and no mark-up is applied to the product in relation to which the service (or other value added) is performed. In such cases, the price increase will usually be postulated in relation to the fee.
[Emphasis added.]
[324] The 1991 MEGs therefore made an express distinction between the “cumulative price” and the “value added” price. They specifically identified a “value added” approach as an exception to the typical approach of using the cumulative price, and specified that this exception would apply if two conditions were met: “the value added is billed as a separate fee, and no mark-up is applied to the product in relation to which the service (or other value added) is performed.” The 1991 MEGs thus established that the value-added price could “be employed where it is billed as a separate fee and no mark-up is applied to the product in relation to which the value-added is applied” (Paul S. Crampton, Canada’s New Enforcement Guidelines: a “Nuts and Bolts” Review, 36 Antitrust Bulletin 883, 1991) (“Crampton 1991”) at p 914), The author, who was the drafter of the 1991 MEGs, used a pipeline example — similar to what the US HMEGs would use some 20 years later — to illustrate a situation where the value-added price could be used as the base price for market definition purposes: when two pipeline operators simply charge a tariff for transporting the oil, such billed fee can be used as the base price for the HMT analysis (Crampton 1991 at p 914).
[325] However, it is clear from the 1991 MEGs that the mainstream approach was to use the “cumulative value of the product” as the base price to postulate a SSNIP: the 1991 MEGs adopted the “common sense approach of using cumulative prices except where it is industry practice to bill the value added as a separate fee” (A. Neil Campbell, Merger Law and Practice: The Regulation of Mergers under the Competition Act, (Scarborough, Ontario: Carswell, 1997) (“Campbell”) at p 61).
[326] In the 2004 and 2011 MEGs, this reference to the alternative value-added price approach was taken out of the MEGs, and the sole reference to the base price remained the price “ordinarily considered to be the price of the product at the stage of the industry (e.g., manufacturing, wholesale, retail) being examined.” The Tribunal underlines that, even though the 2011 MEGs were adopted 14 months after the US HMEGs, they did not revive the value-added price abandoned in 2004 or echo the “implicit or explicit price” exceptions described in the US HMEGs.
[327] The Tribunal further understands that, before the present case, not a single merger was judicially contested by the Commissioner on the basis of a value-added price or the “value-added” approach.
[328] On March 10, 2021, the Tribunal sent a Direction to counsel inviting the parties to provide submissions regarding the specific reference that was contained in the 1991 MEGs to a “value-added” price and its absence from the subsequent iterations of the MEGs published by the Competition Bureau in 2004 and 2011.
[329] In his submissions to the Tribunal, the Commissioner did not identify a particular reason why the paragraph discussing the value-added approach was not explicitly retained in subsequent iterations of the MEGs. He noted that most commentators remained silent on the issue when the new version of the 2004 MEGs was discussed in draft form. The Commissioner submitted that even though the explicit reference to the value-added price was removed, the concept remains embedded in the subsequent versions of the MEGs. He added that a significant change in the 2004 MEGs was to explain that market definition is based on substitutability and focuses on demand responses to changes in relative prices, and that the focus of the market definition exercise is on those dimensions of competition that purchasers of the product value.
[330] The Commissioner further argued that, despite their silence on the “value-added” approach, the 2011 MEGs define the notion of price in a way that encompasses such a “value-added” approach, and that the language of the MEGs provides “latitude on what price is analyzed in a merger.” According to the Commissioner, the 2011 MEGs contemplate flexibility on the SSNIP test to be applied and on the 5% threshold. He added that in paragraph 4.2 of the 2011 MEGs, the reference to price is intended to capture any market that may be anti-competitive, and that the guidelines are agnostic as to how the price to supply the product is defined. The Commissioner also relied on the fact that the MEGs are also clear that in terms of the SSNIP, 5% is generally appropriate but “market characteristics may support using a different price increase” (2011 MEGS at para 4.3).
[331] The Tribunal is not convinced by the Commissioner’s arguments.
[332] The Tribunal first observes that the Commissioner’s references to the apparent flexibility in the 2011 MEGs language strictly relate to the level of the SSNIP threshold, not to the definition of the base price. In the 2011 MEGs, the Competition Bureau refers to the notion of “base price” at paragraphs 4.6 and 4.7. It expressly states that the base price used to postulate a price increase is “typically the prevailing price in the relevant market” (2011 MEGs at para 4.6). The Competition Bureau may elect not to use the prevailing price when market conditions (absent the merger) would likely result in a lower or higher price in the future. It then states that “[i]n general, the base price used to postulate a price increase is whatever is ordinarily considered to be the price of the product in the sector of the industry (e.g., manufacturing, wholesale, retail) being examined” (2011 MEGs at para 4.7). The Tribunal is unable to read in those provisions any direct, oblique, or implied reference to a value-added price or to the situations alluded to in the US HMEGs.
[333] Furthermore, the Tribunal does not accept the Commissioner’s disconcerting contention that the 2011 MEGs could or should somehow continue to be read as implicitly containing the express language of the 1991 MEGs that the Commissioner explicitly removed and abandoned in 2004, and did not re-insert when he issued his revised MEGs in 2011.
[334] In fact, as the panel indicated at the hearing, the Tribunal is left with the distinct impression that the Commissioner is urging the Tribunal to follow the US HMEGs and to prefer them to his own 2011 MEGs. This is not a path that the Tribunal will follow.
[335] The Tribunal instead agrees with P&H that the evolution of the MEGs since 1991 reinforces its position that the base price to be used for purposes of market definition is “whatever is ordinarily considered to be the price of the product at the stage of the industry […] being examined,” which is “typically the cumulative value of the product, inclusive of the value added (mark-up) at the industry level in question” (1991 MEGs at p 9; see also 2004 MEGs at para 3.8 and 2011 MEGs at para 4.7). In the present case, what is ordinarily considered to be the price of the product in the sector of the industry being examined, and what is the cumulative value of the product, is the Cash Price received by the farmers for their grain, be it wheat or canola.
[336] What is more, the conditions of the “value added” price once recognized by the Commissioner in the 1991 MEGs — namely, that the value added be “billed as a separate fee” with “no mark-up […] applied to the product in relation to which the […] value added is performed” — would not be met in this case. These conditions speak of a separate price attached to the value-added product and imply that the value-added element must be more than a simple cost component and must be transacted. Here, the Commissioner’s proposed price for GHS is founded on an imputed price equal to only a fraction of the total value-added provided by grain companies, there is no separate fee (either explicit or implicit) for the alleged value-added provided by grain companies for GHS, and P&H does not charge and bill a separate fee for GHS, or even for the Basis.
[337] The Tribunal is mindful of the fact that the MEGs are neither sacrosanct nor legally binding (Southam FCA at p 41). The MEGs do not restate or revise the law, nor do they substitute for professional advice (Canada (Commissioner of Competition) v Superior Propane Inc, 2001 FCA 104 (“Superior Propane II”) at paras 144–146). However, published guidelines such as the MEGs do provide guidance and notice to the public of how the Commissioner interprets the Act (Canada (Commissioner of Competition) v Canada Pipe Co, 2006 FCA 233, at paras 33, 39; JD Irving Ltd v General Longshore Workers, Checkers and Shipliners of the Port of Saint John, 2003 FCA 266 (“JD Irving”) at para 37). The Tribunal agrees with P&H that the MEGs serve as an important tool for the public and the business community to understand the application of the Act. While not legally binding, they serve as a meaningful element to delineate legal and economic principles that are not fully reflected in the Act itself, and they may be considered as an aid to the Act’s interpretation (Eli Lilly and Co v Apotex Inc, 2005 FCA 361 at para 33). The Tribunal has indeed often noted that the MEGs provide “important enforcement guidelines reflecting the Commissioner’s view on how the Act should be interpreted” (Superior Propane I at para 393).
[338] It is worth repeating the comments made by the Tribunal at paragraph 397 of Superior Propane I:
It must not be forgotten that the point of view put forward in the MEGs represents the considered opinion of the Commissioner, the official appointed by the Governor in Council to administer and enforce the Act. That view, it goes without saying, is the view arrived at by the Commissioner following careful advice given to him by his legal and economic advisers regarding the meaning of the various provisions of the Act. Although the Commissioner is not bound by the MEGs nor are they binding upon this Tribunal, the MEGs should be given very serious consideration by this Tribunal. Needless to say, the Tribunal can disagree and in fact should disagree if it is of the opinion that the interpretation proposed in the MEGs is wrong. However, when referring and considering the MEGs, one should bear in mind the comments in the preface to the MEGs made by Howard Wetston, then Director of Investigation and Research. He stated that the MEGs were published to promote a better understanding of the Director’s merger enforcement policy and to facilitate business planning. He also noted the extensive consultation process which was followed in their preparation.
[339] It cannot be said that the Commissioner was unaware of the fact that the “value-added” approach was expressly removed from the 2004 MEGs and is absent from 2011 MEGs, even if this most recent iteration was issued after the US HMEGs had been adopted in the U.S.
[340] The Commissioner provided no satisfactory explanation or compelling argument to convince the Tribunal that it should now depart from the 2011 MEGs, revive a “value-added” approach the Commissioner abandoned and removed from the MEGs in 2004, and embrace a new standard for product market definition that no longer forms part of the MEGs and the Commissioner has not publicly endorsed since at least 2004.
[341] The Tribunal is of the view that parties to transactions in Canada should be able to rely reasonably on statements of principle made by the Commissioner in published enforcement guidelines, including the MEGs, in order to know the rules applicable to their future activities and planned transactions. As the official responsible for the administration and enforcement of the Act, the Commissioner has a particular responsibility to provide guidance to parties through the publication and consistent application of relevant principles and approaches. Although the MEGs are not legally binding on the Tribunal or the Commissioner, if the Commissioner proposes to depart materially from them in litigated proceedings, the departure should be recognized, explained and justified (for example, by noting an amendment to the Act, or a recent Tribunal or court decision on point, or an advance in economic thinking or methodology) (see also JD Irving at para 37). The Commissioner did not do so with respect to the proposed new “value-added” approach to product market definition that he is inviting the Tribunal to adopt.
[342] In sum, from a legal standpoint, the Tribunal finds no support in adjudicated precedents or in the MEGs for the “value-added” approach to product market definition advocated by the Commissioner in this case. If anything, both the existing precedents and the MEGs instead reinforce the conclusion that, in this case, the relevant products are the purchase of wheat and canola and the relevant price is the Cash Price.
(iv)
The US HMEGs do not apply to this case
[343] The Tribunal makes one last observation on the US HMEGs. In his submissions, the Commissioner tries to draw a parallel between the present case and the hypothetical situations described in those guidelines. He argues that, just as in the US HMEGs, the Elevators’ specific contribution to value can be “identified with reasonable clarity,” through the Basis.
[344] The Tribunal is not persuaded that, even if they were retained, the US HMEGs’ requirements would be met in this case.
[345] The US HMEGs expressly require that in order to base a SSNIP on the value added by the merging firms, the firms’ specific contribution to value must be identifiable with reasonable clarity. In the Tribunal’s view, this is not a situation where an Elevator’s specific contribution of value (by providing GHS) to the Cash Price farmers receive for their wheat or canola can be identified with reasonable clarity. As discussed above, the evidence indicates that the price of GHS, as “imputed” by Dr. Miller, cannot be identified explicitly by the grain companies or the farmers, nor is it implicitly observable by industry participants, as it is not the actual Basis. And if the price for GHS is neither observed nor observable, the “value-added” approach must fail.
[346] The Commissioner’s price for GHS is neither implicit nor explicit; it is a constructed price. In the present case, the two observed or observable prices are the Futures Price and the Cash Price. Not the price for GHS. Moreover, the US HMEGs’ “value-added” approach requires a transacted price, not only a cost component.
(c)
The HMT framework and the SSNIP test
[347] The “value-added” approach proposed by the Commissioner also raises concerns from a conceptual perspective. As proposed, applying a 5% price increase as part of the SSNIP test only to the value-added portion of the price of a product would effectively alter the price change that a hypothetical monopolist must be able to sustain. That alteration to the HMT, without more, would have the effect of seriously modifying the current well-accepted economic analysis underlying market definition and the HMT framework. Since no consideration was given by the Commissioner to whether or how the applicable SSNIP threshold would have to be modified in a value-added scenario, adopting his proposed approach to the product market definition would imprint a profound change to the review of mergers and would significantly recalibrate the current HMT framework governing the market definition exercise.
[348] The Tribunal points out that such impact can be significant. Where the value-added component of a product accounts for 10% of the final price, applying a 5% SSNIP threshold to the value-added component is equivalent to applying a de minimis 0.5% SSNIP to the final price of the product. Conversely, in order to keep the usual well-accepted 5% SSNIP benchmark applied to the final transacted product, it would require applying a 50% SSNIP threshold when considering a component representing a 10% value-added to the final transacted product.
[349] As demonstrated by Ms. Sanderson in her testimony and by P&H in its final argument, the 5% SSNIP threshold used by Dr. Miller in his HMT analysis based on the “imputed” price of GHS (as a proxy for the Basis) translates into an unprecedentedly low SSNIP level when transposed to the total price of grain: the “value-added” approach of the Commissioner would mean that the equivalent SSNIP percentage calculated by Dr. Miller would vary between 0.6% and 0.8% for the purchase of wheat when expressed in terms of the Cash Price, and between 0.1% and 0.2% for the purchase of canola.
[350] The Tribunal is not convinced that in the circumstances of this case, the Commissioner has provided clear and convincing evidence, or arguments, supporting such a fundamental change to the market definition exercise. The Commissioner provided no submissions nor evidence on what the appropriate SSNIP threshold should be in the context of his “value-added” approach, or why it should nonetheless be kept at 5% when a smaller component of the final price is used as the base price. The Commissioner’s position simply assumed that the usual 5% SSNIP threshold should remain, and he applied it in the HMT analysis. These gaps in the conceptual framework undermine the approach he is proposing in this case.
[351] Antitrust economists in the U.S. have voiced concerns about the revised US HMEGs, pointing out that the “value-added” approach would lead to fewer potential substitutes and to lower effective SSNIP thresholds. In response to the draft version of the US HMEGs, some economists identified a fundamental flaw in the “value-added” approach to product market definition proposed in the guidelines, indicating that in many cases, “the value-added service is not actually purchased by customers on a standalone basis” and “customers are not able to substitute to sellers of just the value-added service” (E.M. Bailey, G.K. Leonard and L. Wu, “Comments on the 2010 Proposed Horizontal Merger Guidelines”, HMG Revision Project – Comment, Project No. P092900, June 3, 2010 at p 5). This is precisely the case here with GHS or the Basis. Other economists observed that “one implication of applying the [HMT] using a value-added approach is that it will tend to produce more narrowly defined markets whenever the threshold used for the value added test is not sufficiently increased to account for the ratio of value added to prices” [emphasis added] (P Davis and U Haegler, “Should competition agencies focus on ‘value added’ instead of final prices?”, March 1, 2016 http://ssrn.com/abstract=2740706 at p 16). Again, in the present case, neither the Commissioner nor Dr. Miller turned their mind to the impact of their proposed “value-added” approach on the effective SSNIP threshold.
[352] In Metcash, the Federal Court of Australia noted that on the facts of that case, applying a 5% SSNIP to the imputed value-added price (i.e., the wholesaler profit margin) would reflect approximately a 0.26% increase in the final retail price. The Australian court did not accept that such a small price increase could be used to define a product market, and refused the proposed “value-added” approach.
[353] The Tribunal is unaware of any precedent — and the Commissioner has not mentioned any — where a price increase of less than 5% has been utilized as the SSNIP threshold in applying the HMT analysis. The Tribunal finds that the “value-added” approach as proposed would profoundly change the current HMT framework, and it is not ready to accept that minimal price increases of less than 1% can become the yardstick to justify an intervention in the market. Lower effective SSNIP thresholds lead to narrower product markets, and to a higher likelihood of intervention in mergers.
[354] The Tribunal notes that in response to a question from the panel, Dr. Miller acknowledged that the “value-added” approach, whereby a component of the final price of a product is used as the base or benchmark price for a HMT analysis, is not really addressed in the broader empirical industrial organization literature (Consolidated Transcript, Public, at pp 1431–1432).
[355] In the Tribunal’s view, espousing the “value-added” approach cannot simply be a question of modifying the base price that will be used for the product market definition. Given its impact on the effective price increase it entails for the final product, it also implies, at a minimum, a consideration of the appropriate SSNIP threshold that should be used and an explanation or justification for the selected SSNIP threshold. One cannot dissociate the issue of the “value-added” approach from the issue of the SSNIP threshold. Here, the Commissioner has not presented any evidence nor any economic analysis or authority that would support keeping the 5% SSNIP threshold in the context of his “value-added” approach. Even though the scientific or economic foundation for adopting and using a 5% level remains unclear, the Tribunal underlines that this 5% SSNIP threshold was developed in a context where the base or benchmark price was the cumulative price for the final product sold or purchased. In this case, Dr. Miller and the Commissioner simply transposed this 5% threshold to a value-added price and to their HMT analysis, without explaining or justifying why such threshold could be imported as is in this different context.
[356] In light of this shortcoming in the Commissioner’s economic analysis and evidence, the Tribunal is not persuaded that the Commissioner’s proposed “value-added” approach can be sustained in the circumstances.
[357] The Tribunal makes one other observation.
[358] As acknowledged by Dr. Miller at the hearing, the US HMEGs refer to two possibilities for dealing with market definition in situations where the value added by the supplier of a product allegedly relates to a small portion of the total price of the product. The first option is resorting to a smaller component of the final price corresponding to the value-added to determine the base price, when the explicit or implicit price of the value-added can be identified with reasonable clarity, with an appropriate SSNIP threshold. The second option, when the merging firms’ specific contribution to value is not an implicit or explicit price, is to keep the overall final price of the product as the base price, but use a lower SSNIP threshold in the HMT analysis, adapted to the realities of the industry being examined.
[359] Since the constructed price for GHS, or the Basis, is a small component of the total price of grain that is not transacted in itself, the Commissioner could therefore have argued — as was alluded to in Whole Foods — that a lower SSNIP should be used in a HMT analysis based on the final price of the grain. However, this second option was not considered by the Commissioner nor by Dr. Miller in this case. At the hearing, the Tribunal asked Dr. Miller and Ms. Sanderson about the issue of using a smaller SSNIP threshold on the Cash Price, but did not receive clear answers from either expert.
[360] Resorting to this second option would have required evidence and some economic analysis supporting the different SSNIP threshold to be used. The Tribunal notes that no legal precedent has been identified where, in applying the HMT and in defining the relevant market, a court or a tribunal has discussed the factors or the evidence to be taken into account in order to adjust the appropriate level of the SSNIP threshold in light of the realities of a particular industry. Similarly, the Commissioner has not provided economic or antitrust literature pointing to analyses that could have been done to determine the appropriate SSNIP threshold to be used in the context of a small value added by the supplier or purchaser of a product.
[361] Given the profound change that adopting the “value-added” approach would entail for the current well-accepted HMT analysis underlying market definition, the Tribunal is of the view that no change can be adopted without addressing the SSNIP threshold, in one form or another. The failure to address the SSNIP threshold leaves the Tribunal with no clear and convincing evidence to assess whether either of the two options mentioned in the US HMEGs (i.e., the “value-added price component” or the “reduced SSNIP”) should or could be retained and applied in this case.
(4)
Conclusion on relevant product market(s)
[362] For all the reasons detailed above, the Tribunal concludes that the relevant product markets for the purpose of these proceedings are the purchase of wheat and the purchase of canola. Considering the evidence on the record in this proceeding, the Commissioner has not established, on a balance of probabilities, that there is a distinct relevant market for the supply of GHS for each of wheat and canola. When it comes to the value added by P&H and other Elevators further to their purchase of grain from farmers, there is no separate relevant “market” associated with it in which to conduct the necessary quantitative analysis.
[363] The main considerations weighing in favour of a conclusion that there are distinct relevant markets for the purchase of wheat and canola include the factual evidence as well as the absence of legal foundation or SSNIP threshold analysis supporting the “value-added” approach argued by the Commissioner.
[364] For greater clarity, the Tribunal is not saying that a “value-added” approach to product market definition could never be contemplated or applied. But if the Commissioner intends to resort to such an approach in future cases, he should first clarify the MEGs in that respect. Furthermore, for the Tribunal to be in a position to assess the merits of a “value-added” approach in any given case, the Commissioner would need to present clear and convincing evidence and submissions showing that the contemplated component of a final product is transacted, that it has a price attached to it or a measurable one, and that consideration is given to the SSNIP threshold to be used.
[365] As will be discussed below, this product market issue significantly influences many elements in the remainder of the Tribunal’s analysis.
B.
What is or are the relevant geographic market(s) for the purposes of this proceeding?
[366] The parties took different approaches to the geographic scope of their respective proposed competition markets. The Commissioner initially proposed a local geographic market in which the Virden, Moosomin, and Fairlight Elevators provided GHS to farmers for their wheat and canola. In its pleadings, P&H submitted that in the competition markets for the purchase of wheat and/or canola, the proper geographic area was at least Southeastern Saskatchewan and Southwestern Manitoba, comprising more than 20 Elevators and Crushers.
(1)
Analytical framework
[367] When identifying the geographic dimension of a competition market, the Tribunal typically applies the HMT, as it does for the product dimension. The HMT is designed to assist the Tribunal in identifying the smallest geographic area in which the merged entity, acting as a hypothetical monopolist or monopsonist, may profitably impose a SSNIP or SSNDP respectively — that is, the smallest geographic area over which it could exercise market power (Tervita CT at para 94; Canadian Waste at paras 61, 68, 69, 73; Superior Propane I at paras 84–85; see also VAA CT at paras 300–301; TREB CT at paras 121–124; Facey and Brown at pp 226–230; John S. Tyhurst, Canadian Competition Law and Policy, (Toronto: Irwin Law Inc., 2021) at pp 172–180).
[368] Given the practical challenges associated with determining the base price in respect of which the SSNIP or SSNDP assessment must be conducted, the market definition exercise will sometimes need to go beyond the analysis of prices through the HMT framework and to consider other evidence of substitutability or customer switching. Geographic market definition may therefore involve assessing indirect evidence of substitutability, including factors such as: switching costs; transportation costs; the views, strategies, behaviour and identity of buyers; trade views, strategies, and behaviours of other market participants; price relationships and relative price levels; and shipment patterns (2011 MEGs at paras 4.17–4.24).
[369] In defining the geographic scope of the relevant competition market, the Tribunal has previously concluded that it may be neither possible nor necessary to establish geographic boundaries with precision. The boundaries may well overlap with adjacent markets and be indistinct from those adjacent markets at many geographic points (VAA CT at para 305; Hillsdown at pp 301–302, 310; Canada (Director of Investigation and Research) v Laidlaw Waste Systems Ltd (1992), 40 CPR (3d) 289 (Comp Trib) at p 324). The Tribunal has also held, in particular, that there may be restraints on a merged firm’s market power that come from both inside and outside the defined geographic market (VAA CT at para 305; Hillsdown at p 310).
[370] It should once again be emphasized that business markets, service or trade areas, or operational areas used by company management, are not necessarily the same as a geographic market for the purposes of a competition analysis (Canadian Waste at para 72; Superior Propane I at paras 85, 101, 106).
(2)
Parties’ positions
(a)
The Commissioner
[371] Supported by expert testimony from Dr. Miller, the Commissioner submitted that the relevant geographic market is local in nature. The Commissioner’s principal position was that the Moosomin, Virden, and Fairlight Elevators constituted a relevant market. More precisely, the Commissioner proposed a geographic market around these three Elevators in which a hypothetical monopolist could impose a SSNIP for GHS. Dr. Miller’s expert evidence included a HMT analysis (for the purposes of the Commissioner’s proposed product market for the delivery of GHS) and concluded that the Moosomin, Virden, and Fairlight Elevators, acting as a hypothetical monopolist, would have the ability and incentive to impose a SSNIP for GHS in the geographic area served by these three Elevators. Dr. Miller used a 5% SSNIP threshold in his analysis.
[372] In addition, at the hearing and in response to Ms. Sanderson’s critique on the product market definition, Dr. Miller testified that the geographic area around the Virden, Moosomin, Fairlight, and Whitewood Elevators could serve as the relevant geographic dimension of a competition market for the cash sale of wheat, again based on his HMT analysis (Exhibits CA-A-192 and P-A-193, Relevant Results from Ms. Sanderson’s HMT Calculations (“Dr. Miller Revised HMT”).
[373] To support the argument for a local market, the Commissioner emphasized the importance to farmers of the distance between their farm and the point of delivery, and the associated transportation costs, when deciding on an Elevator or Crusher for the sale of their grain. The Commissioner submitted that most farms analysed by Dr. Miller deliver their grain to Elevators located less than 100 kilometers away. The Commissioner submitted that the size of the service areas from which the Moosomin and Virden Elevators draw at least 90% of the wheat or canola they handle demonstrated that most of the volumes are drawn from farms located near each of those Elevators.
[374] Dr. Miller developed a model of demand to understand how farms make decisions as between Elevators and Crushers to sell their grain. In his expert report, Dr. Miller relied on four qualitative elements, namely, a review of case documents, the distances that farms tend to send their grain, the distances between the Elevators, and the profit margins. He also relied on his HMT analysis (Dr. Miller Report at para 4).
[375] Dr. Miller testified that in his opinion, proximity is an important factor in a farm’s choice of a “primary” Elevator for the sale of its grain. Dr. Miller considered the driving distance between the Moosomin and Virden Elevators, and calculated the drive time and drive distance from farms to Elevators using various data, including data from the Elevator operators.
[376] During his analysis, Dr. Miller identified farms that were customers of the three Elevators (i.e., Moosomin, Virden, and Fairlight) in his proposed geographic market for the delivery of GHS. In developing a model for demand, Dr. Miller considered the “draw areas” for the Elevators and used available data to determine each Elevator’s “service area” — described as the set of closest Census Consolidated Subdivisions from which each Elevator draws at least 90% of its total wheat or canola intake. Dr. Miller used the service areas for these three Elevators to define a “Farmer Region,”namely, a collection of farms useful for understanding the pricing incentives at the three Elevators and how their current and prospective customers would respond to price changes (“Farmer Region”)
[6]
.
[377] In Dr. Miller’s analysis, the respective service areas for the Moosomin and Virden Elevators were contiguous with each Elevator. In addition, the service areas for the two Elevators substantially overlapped, suggesting that both Elevators expect to purchase grain from similar or geographically clustered farms. After determining the service areas for other Elevators and for more distant Crushers, Dr. Miller’s analysis of service areas also found that the median farm selling canola may be more willing to travel longer distances to sell to Crushers.
[378] Dr. Miller also quantified the role of distance in his demand model and considered internal documents from the merging parties to support a conclusion that the Moosomin, Virden, and Fairlight Elevators were close competitors. He also observed the margin (or mark-up) earned by the Virden Elevator, both for wheat and for canola, which supported a conclusion that the Virden Elevator faced a relatively small set of competitors.
[379] Turning to his HMT analysis, Dr. Miller determined that a HMT using a merger simulation model showed that the Moosomin, Virden, and Fairlight Elevators comprised a relevant geographic competition market where a hypothetical monopolist would find it profitable to impose a SSNIP on the price of GHS. For GHS for wheat, Dr. Miller’s predicted price increase of a hypothetical monopolist was CAD $9.03 per MT at the Moosomin Elevator and CAD $5.88 per MT at the Virden Elevator, representing changes in price (compared to his constructed pre-Acquisition price for GHS at each Elevator) of 26.0% at Moosomin and 21.6% at Virden. With respect to GHS for canola including Crushers, his predicted price increase of a hypothetical monopolist was CAD $2.76 per MT at the Moosomin Elevator and CAD $1.51 per MT at the Virden Elevator, representing changes in price (compared to his constructed pre-Acquisition price for GHS) of 22.2% at Moosomin and 7.6% at Virden (Dr. Miller Report at paras 74–79 and Exhibit 9). According to Dr. Miller, his simulation and the resulting projected price changes demonstrated that a hypothetical monopolist of the Moosomin, Virden and Fairlight Elevators would “increase price by far more than a typical SSNIP” (i.e., 5%) (Dr. Miller Report at para 79).
(b)
P&H
[380] P&H submitted that the relevant geographic area for competition market purposes was much larger. P&H submitted that all farmer witnesses can and do haul their wheat and canola significant distances to Elevators and (for canola) to Crushers because it is financially worthwhile to do so. P&H submitted that farms closer to a particular Elevator receive a premium for their products, because they do not have high costs associated with hauling their grain to that Elevator. In P&H’s submission, transportation costs do not provide any kind of constraint on competition in the relevant geographic market that shields the Virden and Moosomin Elevators from competition.
[381] P&H first relied on Mr. Heimbecker’s evidence. Mr. Heimbecker testified that in the area surrounding the Virden and Moosomin Elevators, P&H competes with numerous Elevators and Crushers to purchase wheat and canola from farms. For wheat in that area, Mr. Heimbecker identified over 20 rival Elevators owned by six competing grain companies. For canola in that area, he identified at least 27 rival purchasing locations owned by nine grain companies. Mr. Heimbecker also referred to P&H’s internal documents, including business plans, internal emails, and the competitors identified in “Draw Analysis Reports” prepared by third party consultants for P&H.
[382] P&H also relied on the expert evidence provided by Ms. Sanderson. Ms. Sanderson initially proposed a geographic market of “(at least) southeastern Saskatchewan and southwestern Manitoba” (Exhibits P-R-180 and CA-R-181, Expert report of Ms. Margaret Sanderson (“Ms. Sanderson Report”), at para 14). She adopted a monopsony framework, analysing the market for the purchase of wheat or canola by Elevators and Crushers. Ms. Sanderson’s view was that the Virden and Moosomin Elevators were small buyers in an unconcentrated market or industry and, specifically, that the geographic market is wider than merely the area around the Moosomin, Virden, and Fairlight Elevators. She did not conduct a formal HMT analysis to support her broad proposed geographic market but relied on other quantitative and qualitative evidence and draw areas.
[383] However, Ms. Sanderson did comment on hypothetical monopolist issues in responding to Dr. Miller’s HMT analysis in her hearing slides and during her testimony (Exhibits P-R-182, CA-R-183 and CB-R-184, Slides of Ms. Sanderson (“Ms. Sanderson Slides”), at pp 72, 74–75). While she did not do her own HMT analysis, Ms. Sanderson did a recalculation of Dr. Miller’s model to generate new values based on a different denominator — namely, the Cash Price instead of the constructed price for GHS —, using Dr. Miller’s diversion ratios and his Virden margin. Ms. Sanderson’s revised HMT analysis resulted in the same absolute price changes calculated by Dr. Miller, but with different relative price variations in light of the different denominator she used.
[384] Ms. Sanderson’s recalculation of Dr. Miller’s HMT analysis suggested that when the Cash Price is used as the denominator, a market made up of the Moosomin, Virden, Fairlight, Oakner, Whitewood, Elva, Shoal Lake, and Carnduff Elevators would be a relevant geographic competition market for wheat. The relevant geographic market for canola would be larger and would include other Elevators as well as several Crushers such as Bunge Harrowby and LDC Yorkton (Ms. Sanderson Slides at pp 74-75). Ms. Sanderson based her HMT results on the same diversion ratios as Dr. Miller and on the weighted average price changes for the Moosomin and Virden Elevators (as calculated by Dr. Miller), but she used the wheat prices prevailing at the Virden Elevator as the reference price to calculate her relative price decreases. Ms. Sanderson indicated in her testimony that she used the Virden prices because Virden was the Elevator from which Dr. Miller had calculated the margin he used in his economic model (Consolidated Transcript, Confidential A, at p 1782).
[385] In addition to her HMT recalculations, Ms. Sanderson studied the evidence in the transaction data and witness testimonies to provide an opinion on the relevant geographic market based on the number of competing buyers (Elevators and Crushers) P&H faced in purchasing wheat or canola, farms’ switching alternatives amongst the Elevators and Crushers, and the distance farms were prepared to travel to sell their products.
[386] Ms. Sanderson prepared “draw area” maps for the Elevators in Southeastern Saskatchewan and Southwestern Manitoba, and so-called “heat maps” derived from the overlapping draw areas. The draw areas for the Moosomin and Virden Elevators substantially overlapped.
[387] Ms. Sanderson also studied the Farmer Region identified by Dr. Miller as well as the “corridor of concern” identified by the Commissioner prior to the commencement of this proceeding. As described by Ms. Sanderson, the corridor of concern was a geographic polygon focused on approximately 80 farms located between the Virden and Moosomin Elevators on either side of the Trans-Canada Highway. (By definition, the “corridor of concern” included farms within a one-hour driving distance from both Elevators using commercial trucking roads. It included two small non-contiguous areas south of the main polygon.) The Moosomin and Virden Elevators’ draw areas both covered this “corridor of concern.”
[388] Ms. Sanderson observed that given the range of options available to farms for the sale of their crops, the Cash Prices set by P&H to purchase wheat and canola must be competitive with the Cash Prices set by numerous Elevators operated by many competitors, because they are all buying from the same farms.
[389] Ms. Sanderson also referred to internal documents of P&H and LDC. Ms. Sanderson’s report noted that P&H and its customer service representatives at the Moosomin Elevator referred to and tracked prices of more than a dozen other Elevators and Crushers, in addition to the Virden Elevator. LDC’s documents also showed that the Virden Elevator competed with almost a dozen other competitor purchase locations in addition to the Moosomin and Fairlight Elevators. Ms. Sanderson made specific reference to Elevators and Crushers identified by P&H before the Virden Acquisition in its fiscal 2019 and fiscal 2020 business plans for the Moosomin Elevator.
[390] Ms. Sanderson concluded that the transaction data, testimonies and documentary evidence demonstrated that there was a large set of relevant competing Elevators buying as much or more volume than the Moosomin or Virden Elevators from farms within Dr. Miller’s Farmer Region, that the geographic market had many participants, and that P&H had a small share of that market.
[391] In his reply expert report, Dr. Miller addressed the geographic market analysis of Ms. Sanderson. He testified that farms in the towns close to the Moosomin, Virden, and Fairlight Elevators were particularly likely to rely on those Elevators for the sale of their grain. Dr. Miller concluded that the desirability of travelling to a particular Elevator differs for farms located at different points within each Elevator’s draw area. Looking at the percentage of MT sold and the quantity sold on a town-by-town basis, Dr. Miller found that farms close to the centre of the geographic area served by the three Elevators had a “distinct preference” for working with the Moosomin, Virden, and Fairlight Elevators. He also found that while it was rare for a farm located directly between the Moosomin and Virden Elevators to choose a more distant Elevator, the frequency increases for farms farther from this centralized location — both for wheat and for canola, including Crushers.
[392] During the experts’ concurrent evidence session at the hearing, Ms. Sanderson agreed that distance matters to an individual farm and that farms preferred to sell closer and to travel a shorter distance to sell their wheat and canola. She also agreed that the transaction data, supported by the testimony of the farmer witnesses, were helpful in understanding what makes an Elevator attractive to a farm for the sale of wheat or canola. Dr. Miller and Ms. Sanderson agreed that the diversion ratios summarized the information about farmers’ sale behaviour in relation to the Elevators available and the relative distance to be travelled for delivery.
[393] As mentioned above, at the hearing and in response to Ms. Sanderson’s recalculation of his HMT analysis using the Cash Price, Dr. Miller also testified that the geographic area around the Virden, Moosomin, Fairlight, and Whitewood Elevators could serve as the geographic dimension of a competition market for the “cash sale of wheat.” In response, Ms. Sanderson observed that her analysis of the purchases from farms within Dr. Miller’s Farmer Region concluded that several other Elevators had higher wheat purchases than the Whitewood Elevator from the farms in that Farmer Region, and that two other Elevators (in addition to Whitewood) were located within it. Ms. Sanderson testified that an application of the HMT in Dr. Miller’s alternative geographic market would engage a larger set of Elevators than just the Moosomin, Virden, and Fairlight Elevators, but that the question was how many more.
[394] Both experts agreed that neither the Moosomin nor the Virden Elevator have any special or unusual competitive significance in the marketplace.
(3)
Tribunal’s assessment
[395] After considering all the evidence in this case, the Tribunal is able to describe the geographic dimension of the relevant competition markets based on the factual and expert evidence.
[396] Although Dr. Miller’s proposed product market (i.e., the delivery of GHS to farms by Elevators and Crushers) has not been accepted by the Tribunal, the panel nonetheless found his geographic market analysis to be helpful and persuasive in understanding certain aspects of the behaviour of farms in selecting an Elevator or Crusher to sell their grain. The panel also found Ms. Sanderson’s evidence useful, including her recalculation of Dr. Miller’s HMT analysis on the basis of the Cash Price, and incorporated it into its assessment.
[397] In addition to the evidence on the HMT analyses conducted by the two experts, the Tribunal also assessed a number of factors in determining the geographic scope of the relevant competition markets. The salient evidence concerned the purchases of wheat and canola by the Moosomin and Virden Elevators and other Elevators and Crushers in the area, and the corresponding selling behaviour of farms. It included the following: the experts’ analysis of transaction data in relation to the purchase and sale of wheat and canola to those Elevators and Crushers, and the oral testimony of the farmer witnesses; the evidence related to the distance that must be travelled to deliver grain to an Elevator or Crusher (and relatedly, the transportation costs and time it takes to travel that distance) and the volume and frequency of those purchases by Elevators from farms; expert evidence as to draw areas, heat maps, and diversion ratios; evidence as to prices paid to farms at the Elevators, including price-setting and prices that are negotiated and therefore depart from the Cash Price or Basis offered for each Elevator; and internal documents from the merging parties suggesting the perceived scope of the geographic market, including communications with farms about the purchase of their crops.
[398] In the result, the Tribunal’s analysis of the geographic dimension of the competition markets is not restricted to the geographic area identified in Dr. Miller’s analysis. However, the Tribunal does not agree with the much wider region initially advocated by P&H and described in Ms. Sanderson’s testimony (i.e., Southeastern Saskatchewan and Southwestern Manitoba). Like certain aspects of both Dr. Miller’s and Ms. Sanderson’s analyses, the Tribunal finds that the geographic area relevant to a competition analysis for wheat is different from the relevant geographic area for canola.
[399] The Tribunal concludes that, in general, Elevators that are closer to a farmer’s crop are more attractive to farms for the purchase of their wheat. Considering in particular the farmers’ testimonies concerning their selection of purchaser Elevators and the role of transportation costs, the setting and negotiation of prices, and the expert evidence, the Tribunal finds that the key competitors to the Moosomin and Virden Elevators are rival Elevators nearby to them, particularly the Fairlight Elevator. By contrast, Elevators that are farther away are not part of the relevant geographic market for competition purposes. Although more distant Elevators may purchase some quantity of grain and may provide some degree of competitive discipline on the Cash Price (including specifically, on the Basis) offered to farms by the Moosomin and Virden Elevators, that does not necessarily lead to the conclusion that those Elevators are all within the relevant competition market.
(a)
HMT analyses
[400] Three different HMT analyses have been presented to the Tribunal by Dr. Miller and Ms. Sanderson.
[401] Dr. Miller’s initial HMT analysis, summarized at Exhibit 9 of his expert report, concluded that the Moosomin, Virden, and Fairlight Elevators formed a relevant geographic competition market where a hypothetical monopolist would find it profitable to impose a SSNIP on the price of GHS. For GHS for wheat, Dr. Miller’s model predicted that a hypothetical monopolist could impose an increase of 26.0% on the constructed pre-Acquisition price for GHS for wheat at the Moosomin Elevator, and of 21.6% on the price for GHS for wheat at Virden. With respect to GHS for canola including Crushers, Dr. Miller’s predicted price increase of a hypothetical monopolist was 22.2% at the Moosomin Elevator and 7.6% at the Virden Elevator. In each case, Dr. Miller compared his projected price increase to his computed price for GHS prevailing at each of the Moosomin and Virden Elevators. Since all of his projected price increases clearly exceeded the typical 5% SSNIP threshold — they ranged from 7.6% to 26.0% —, Dr. Miller arguably did not have to calculate weighted average reference prices for GHS for wheat or canola, or weighted average price increases representing the combined average price increase for the Moosomin and Virden Elevators (which, in the Tribunal’s view, would be a more accurate basis for price change analysis under the HMT framework).
[402] Given the Tribunal’s finding on the relevant product market, and its conclusion that the relevant products are the purchase of wheat and canola and that the relevant base prices are the Cash Prices of wheat and canola paid to the farms by P&H, the Tribunal cannot entirely retain Dr. Miller’s initial HMT analysis: his projected price increases, expressed in terms of percentages, are based on the wrong base price, namely, the computed prices for GHS for wheat and canola.
[403] The second HMT analysis is the recalculation of Dr. Miller’s model done by Ms. Sanderson as part of her presentation at the hearing (Ms. Sanderson Slides at pp 72–75). This recalculation replicated Dr. Miller’s model, including the diversion ratios and the Virden Elevator margin calculated by Dr. Miller, but determined the relative values of the projected price changes based on a different denominator, namely, the Cash Prices instead of the constructed prices for GHS. The Tribunal observes that both experts agree on the absolute figures of the predicted price variations for each of wheat and canola (expressed in dollars per MT) coming from Dr. Miller’s HMT analysis (Consolidated Transcript, Confidential A, at p 1795).
[404] Ms. Sanderson’s revised HMT analysis used the same absolute price changes as those calculated by Dr. Miller but resulted in different relative price variations in light of the different denominator she used. Ms. Sanderson’s recalculation concluded that a hypothetical monopsonist controlling each of the Moosomin, Virden, and Fairlight Elevators could impose a price change of only 3.9% on the pre-Acquisition Cash Price for wheat at the Moosomin Elevator (i.e., CAD $9.03 on CAD $229.73 per MT), and of 2.5% on the Cash Price for wheat at the Virden Elevator (i.e., CAD $5.88 on CAD $239.11 per MT) (Ms. Sanderson Slides at p 72; Ms. Sanderson Report at Figure 49). As the price variations would be in the purchase of grain, these would be price decreases. With respect to canola including Crushers, Ms. Sanderson’s predicted price decrease was 0.6% on the pre-Acquisition Cash Price at the Moosomin Elevator (i.e., CAD $2.76 on CAD $461.46 per MT), and 0.3% to the Cash Price at the Virden Elevator (i.e., CAD $1.51 on CAD $452.80 per MT). None of the resulting price decreases exceeded the usual 5% SSNIP or SSNDP threshold, which led Ms. Sanderson to conclude that, based on her revised HMT analysis using Cash Prices, the relevant geographic market had to be larger than the Moosomin, Virden, and Fairlight Elevators.
[405] In her presentation at the hearing, Ms. Sanderson provided results for more Elevators than in Dr. Miller’s HMT analysis, using the Cash Prices (Ms. Sanderson Slides at pp 74–75). She concluded that, when the Cash Price is used as the denominator, a market made up of the Moosomin, Virden, Fairlight, Oakner, Whitewood, Elva, Shoal Lake, and Carnduff Elevators would be a relevant geographic competition market for the purchase of wheat. The relevant geographic market for the purchase of canola would be larger and would include the Moosomin, Virden, Fairlight, Oakner, Whitewood, Brandon (Richardson), Melville, Souris East, Shoal Lake, and Elva Elevators as well as the Crushers at Harrowby (Bunge), Yorkton (LDC), Velva (ADM), and Yorkton (Richardson). Ms. Sanderson based her HMT results on the same diversion ratios as Dr. Miller and on the weighted average absolute price changes for each of wheat and canola. In terms of average relative price changes, she expressed the price variations as a percentage of the weighted average wheat and canola prices calculated for the Virden Elevator (i.e., CAD $239.11 per MT and CAD $452.80 per MT, respectively).
[406] The third iteration of the HMT analyses prepared by the experts is Dr. Miller’s response to Ms. Sanderson’s recalculation, presented at the concurrent evidence session and summarized in Dr. Miller Revised HMT. This Revised HMT analysis only looked at the geographic market for the purchase of wheat, and did not consider canola. Since the HMT framework dictates that additional candidates for market definition purposes must be ordered by their diversion ratios, Dr. Miller determined that, for wheat, the next closest competitor Elevator to the Moosomin and Virden Elevators (other than the Fairlight Elevator) was the Whitewood Elevator. Dr. Miller testified that, based on his analysis, the diversion ratios for wheat were higher at the Whitewood Elevator than at the Oakner Elevator, when these ratios are weighted by sales from the Virden and Moosomin Elevators, or weighted by sales from the Virden, Moosomin, and Fairlight Elevators. Dr. Miller concluded that, even using the Cash Price as a denominator in his Revised HMT analysis, the relevant geographic market for wheat would be a smaller market comprising only the Moosomin, Virden, Fairlight, and Whitewood Elevators, as such relevant market would reach the 5% SSNIP threshold.
[407] The Tribunal pauses to note that Dr. Miller did not discard the recalculated HMT analysis made by Ms. Sanderson based on his data, model, and framework but using the Cash Prices — as opposed to his constructed price for GHS — as a denominator. On the contrary, in his Revised HMT analysis, Dr. Miller simply redid Ms. Sanderson’s recalculation by looking at additional Elevators using their respective weighted averaged diversion ratios (Consolidated Transcript, Confidential A, at pp 1787 ff). He also used a different “reference price” as the basis to calculate the relative price changes he observed: instead of solely using the prevailing pre-Acquisition price for wheat at the Virden Elevator as Ms. Sanderson did, he developed the weighted average HMT price change relative to the pre-Acquisition price for each of the Moosomin Elevator (CAD $229.73 per MT) and the Virden Elevator (CAD $239.11 per MT). He then selected the Moosomin Elevator price change and this led him to conclude that a geographic market comprised of the Moosomin, Virden, Fairlight, and Whitewood Elevators would meet the 5% SSNIP threshold, since his relative price change based on the lower Moosomin Elevator price was 5.06%. Dr. Miller’s table reproduced in his Revised HMT analysis indicates that, had he used the higher prevailing wheat price at the Virden Elevator, the relative price change would have been 4.86% and would thus have been below the 5% SSNIP threshold.
[408] It thus appears that the two experts disagree on the “reference price” to use for the calculation of the relative HMT price variations: Dr. Miller relied on the lower Moosomin Elevator price in his Revised HMT analysis, whereas Ms. Sanderson used the higher Virden Elevator price in her recalculation. Neither Dr. Miller nor Ms. Sanderson apparently considered using a weighted average of the Moosomin and Virden Elevators prices, which would be a more accurate base for price change analysis under the HMT framework. This is especially true in a case like the present one, where the results of adding candidate markets in the HMT analysis are each very close to the 5% benchmark. The purpose of the HMT analytical framework is to assess how a “hypothetical monopolist” (or “hypothetical monopsonist”) controlling a hypothetical group of entities would behave; as such, the proper relevant “reference price” of a product has to be the weighted average price of all entities controlled by such hypothetical monopolist or monopsonist that are supplying or purchasing the product. The Tribunal notes that in this particular case, changing the “reference price” has a direct impact on the conclusions to be drawn from the HMT analyses based on the Cash Price, as it significantly modifies the group of Elevators needed to meet the 5% SSNIP/SSNDP threshold. The Tribunal further observes that, even though Dr. Miller opted to calculate a weighted average of the Moosomin and Virden Elevators to determine the appropriate diversion ratios to the rival Elevators and used weighted average dollar price changes in his Revised HMT analysis, he relied solely on the lower Moosomin price for wheat as a denominator to determine his estimated price variations. During the concurrent evidence session, Dr. Miller provided no explanation for not also using a weighted average reference price in his calculations.
[409] At the hearing, Dr. Miller relied on his reading of paragraph 4.4 of the 2011 MEGs to state that the HMT is satisfied “if any price of the merging firms goes up by five per cent” (Consolidated Transcript, Confidential A, at p 1784). With respect, the Tribunal does not agree with Dr. Miller’s interpretation of the MEGs. What the 2011 MEGs say is that the HMT is met when “the price increase of at least one product of the merging parties” [emphasis added] exceeds the SSNIP threshold (2011 MEGs at para 4.4). It does not say the price of a product of one of the merging firms. Here, all the Elevators purchase the same product, namely, wheat or canola. And what the Tribunal needs to assess is the predicted price variation for wheat or canola for the selected Elevators acting as a hypothetical monopolist. In the Tribunal’s view, this has to be measured in relation to the weighted average price variation of wheat and canola for all Elevators involved — not to the price of just one of the merging firms.
[410] The Tribunal has not found in the evidence what the weighted pre-Acquisition average price for wheat would be for the Moosomin and Virden Elevators taken together, or for any larger group of Elevators. However, the Tribunal underlines that the simple arithmetical average between the weighted average base price for wheat at the Moosomin and Virden Elevators would be CAD $234.42 per MT (i.e., ($229.73 + $239.11) ÷ 2). Moreover, since the evidence indicates that, pre-Acquisition, the Virden Elevator handled relatively more wheat than the Moosomin Elevator (see, for example, Ms. Sanderson Report at Figure 62 and Consolidated Transcript, Confidential A, at p 1789), it is a mathematical certainty that the weighed average price for wheat for the Moosomin and Virden Elevators would be higher than the straight arithmetical average mentioned above, since the Virden Elevator has both the higher weighted average price and the higher volume quantity. Had Dr. Miller used a weighted pre-Acquisition average price for wheat, or even the more conservative arithmetical average of CAD $234.42 per MT, in his Revised HMT analysis, it is clear from his evidence that the results of his Revised HMT analysis for wheat using the Cash Prices would have concluded that even a geographic market made up of the Moosomin, Virden, Fairlight, Whitewood, and Oakner Elevators would not be large enough to satisfy the 5% SSNIP/SSNDP threshold: using the more conservative arithmetical average, the predicted price changes would not have exceeded 4.96% in a geographic market including the Whitewood Elevator, and 4.90% in a market including both the Whitewood and Oakner Elevators.
[411] Turning to Ms. Sanderson’s HMT analysis, had she used a weighted pre-Acquisition average price for wheat as a “reference price,” instead of the higher Virden Elevator price, the results of her recalculated HMT analysis for wheat using the Cash Prices would have yielded slightly higher relative changes. For example, for the group of seven Elevators ending with the Shoal Lake Elevator, her estimated price change of 4.86% would have been 4.96% (i.e. CAD $11.63 / CAD $234.42) and suggest that the 5% SSNIP/SSNDP threshold would have been close to be met with one less Elevator (i.e., without the need of adding the Carnduff Elevator).
[412] The Tribunal is mindful of the fact that Dr. Miller’s (and Ms. Sanderson’s) calculations twirl around the 5% threshold as soon as four or five Elevators are included in the geographic competition market for wheat, and that there are margins of error in these calculations.
[413] In light of the foregoing, based on the evidence before it regarding the HMT analyses using the Cash Prices, and considering possible margins of error regarding the median of the Basis, the Tribunal concludes that a geographic market for wheat including only the Moosomin, Virden, Fairlight, and Whitewood Elevators, or even those four Elevators plus the Oakner Elevator, has not been established on the evidence. In sum, in his HMT analyses, the Commissioner has not adduced clear and convincing evidence that the relevant geographic market for wheat could be limited to those four or five Elevators. Thus, further to its review of the HMT analyses, the Tribunal is of the view that, on a balance of probabilities, the relevant geographic market for wheat is more likely than not to include at least the following Elevators listed in Ms. Sanderson’s recalculated HMT analysis: the Moosomin, Virden, Fairlight, Oakner, Whitewood, Elva, and Shoal Lake Elevators. The Tribunal says “at least” because determining whether the Carnduff Elevator identified by Ms. Sanderson should also be included in the relevant geographic market is too close to call in light of the above discussion on the weighted pre-Acquisition average price for wheat.
[414] In the competition market for the purchase of canola, the Tribunal accepts that Crushers play a more significant role in the competitive process and reduce the likelihood that the Moosomin and Virden Elevators, acting as a hypothetical monopsonist, could impose a SSNDP in the purchase of canola. The evidence suggests that Crushers are able to attract some canola purchases from a longer distance than Elevators and that some canola deliveries to Crushers by-pass Elevators that are closer in distance to a farm. A Crusher may therefore have greater competitive impact on the merged entity’s prices and limit its ability to impose a SSNDP than would have an Elevator that buys canola. Furthermore, the Tribunal notes that in the concurrent evidence session, Dr. Miller did not respond to Ms. Sanderson’s recalculation HMT analysis for canola using the Cash Prices, found at page 75 of Ms. Sanderson Slides. In her analysis, Ms. Sanderson concluded that the relevant geographic market for the purchase of canola would include at least the Moosomin, Virden, Fairlight, Oakner, Whitewood, Brandon (Richardson), Melville, Souris East, Shoal Lake, and Elva Elevators as well as the Crushers at Harrowby (Bunge), Yorkton (LDC), Velva (ADM), and Yorkton (Richardson). Even for this larger group of Elevators and Crushers, Ms. Sanderson’s predicted price decrease of a hypothetical monopsonist would only be 3.08% (i.e., a weighted average price change of CAD $13.94 per MT on the pre-Acquisition price of CAD $452.80 at the Virden Elevator), significantly below the usual 5% threshold for market definition purposes. Here, the pre-Acquisition price for canola at the Moosomin Elevator (i.e., CAD $461.46) is higher than at Virden, and Virden’s purchases of canola are much higher than Moosomin’s (Ms. Sanderson Report at Figure 62). In light of the foregoing, and based on the evidence before it regarding the HMT analyses using the Cash Prices, the Tribunal concludes that, on a balance of probabilities, the relevant geographic market for canola is more likely than not to include at least all the Elevators and Crushers listed in Ms. Sanderson’s recalculated HMT analysis. Again, the Tribunal says “at least” because Ms. Sanderson’s HMT analysis stopped at a 3.08% SSNIP level, and it is uncertain how many other Elevators and/or Crushers would need to be added to reach the typical 5% SSNIP threshold.
(b)
Distance, transportation costs, and farms’ preferences
[415] Turning to other factors, the Tribunal accepts that most farms deliver grain to Elevators located less than 100 kilometers from the location of their crops. Like both Dr. Miller and Ms. Sanderson, the Tribunal finds that farms prefer to travel shorter distances to sell their grain. Most of the volume of grain purchased by the Moosomin and Virden Elevators comes from farms located in proximity to them. The Tribunal also finds, based on Ms. Sanderson’s data, that in the “corridor of concern,” a significant number of farms sell their wheat exclusively or substantially to the Moosomin, Virden, and Fairlight Elevators.
[416] At the hearing, several farmers testified about the sales of their grain to Elevators and Crushers. In general, they expressed a preference to sell to Elevators closer to their farms unless a more distant Elevator or Crusher made it worthwhile financially to travel the extra distance. The three farmers called to testify by the Commissioner generally stated their preference to sell most of their grain to the Virden and Moosomin Elevators, and sometimes to the Fairlight Elevator, rather than to more distant Elevators and Crushers. In several cases, the witnesses acknowledged in cross-examination that they have also delivered some proportion of their wheat and canola crops to purchasers more distant than the Moosomin, Virden, and Fairlight Elevators. When selling canola to Crushers, their canola crops would also travel past a number of possible Elevators to which they could have sold while en route to a Crusher location.
[417] When selecting an Elevator or Crusher for a shipment of their grain, the farmers’ principal considerations were price, timing (i.e., when grain could be accepted by the Elevator), the grade of the grain to be sold, and the travel distance to the purchase point. Additional factors included the farmers’ business relationship with each Elevator, each farmer’s ability to store grain at their own farm, road conditions including restrictions on the use of some highways in the springtime, and whether an Elevator was located on a main or a secondary highway (which affects the number of trips, speed of the truck en route, and the wear and tear on the vehicle).
[418] The testifying farmers called by the Commissioner and P&H generally indicated that a higher price is necessary to cause them to sell to a more distant location — to make it worthwhile to travel the extra distance to deliver the crop. Mr. Lincoln advised that his transportation cost was approximately CAD $8 per MT, with each additional 15 minutes drive costing CAD $1 per MT. Mr. Paull testified that he would not leave a local Elevator for a few pennies per bushel, but would do so for CAD $0.10, $0.15 or $0.20 per bushel. Sometimes, he could even get CAD $0.30 or $0.40 more per bushel to go a longer distance. He would sell to a more distant Elevator if there was “enough profit” in it. Mr. Wagstaff testified that his decision to do so also depended on his own time and the efficiency of deliveries. From his farm, he could deliver four loads per day with his own trailer to the Virden Elevator, but could only make one trip per day to the Crusher located in Bloom, Manitoba, which is much farther away from his farm. Mr. Duncan, Mr. Paull, and Mr. Hebert all testified that Crushers made it financially worthwhile to sell canola at a longer distance away from their respective farms.
[419] The Tribunal found Dr. Miller’s evidence concerning farms’ behaviour in selecting a purchaser to be helpful in assessing the geographic dimension of the markets for the purchase of wheat and canola by Elevators and Crushers. Dr. Miller’s analysis found that farms select anElevator for the sale of their grain based on proximity to the farm, because it decreases delivery costs and because they may have a relationship with personnel responsible for the Elevator. The farms’ travel time and cost to deliver the grain to the purchase point were important factors in the choice, together with price. Farms incur delivery costs by crop weight and by kilometre, whether they deliver themselves or hire commercial trucks to do so.
[420] Dr. Miller’s analysis considered travel distance and time between Elevators and from farms to the Moosomin and Virden Elevators, to identify close competitors to the merged entity. He used data collected from the merging parties and several other grain companies comprising over 20 Elevators, as well as census data and other source data. Dr. Miller found that the Fairlight Elevator was the most proximate competitor to the Moosomin and Virden Elevators.
[421] Dr. Miller also calculated the range of travel time for the delivery of each of wheat and canola, representing a percentage of deliveries. In general, the range of travel time for 50% of deliveries was longer for canola than for wheat. Based on travel distance and time, and weighted by quantity of grain sold, Dr. Miller analysed the range of time taken by 90% of farms to travel to each of the Moosomin, Virden, and Fairlight Elevators, all other Elevators, and Crushers. Dr. Miller’s analysis showed that the median, representing 50% of farms, was approximately half an hour drive time for both wheat and canola. Dr. Miller’s analysis also showed that 90% of farms travelled from four to 79 minutes to deliver wheat to all chosen Elevators in the model and from four to 75 minutes to deliver canola to Elevators. The time range to canola Crushers increased to 44 minutes up to a maximum of 147 minutes. Recognizing that this analysis occurred over one crop year, the Tribunal accepts Dr. Miller’s evidence to show that close proximity of a delivery point is important to farms when selling wheat and canola.
(c)
Draw areas and heat maps
[422] As noted, both Dr. Miller and Ms. Sanderson used draw areas (or service areas) for Elevators in their analyses. Both indicated that Elevators and Crushers draw the grain they purchase from areas surrounding them, creating a geographic cluster of supplier farms.
[423] Dr. Miller’s review found that proximity was an important factor in the farms’ choice of an Elevator for the sale of their grain. He also noted that some Crushers attract supply from greater distances than Elevators. He found, for example, that the median farm selling to the Moosomin Elevator is just five kilometers from its location and the median farm selling to the Virden Elevator is about 20 kilometers from it, whereas the median farm selling to the Yorkton Crusher is over 100 kilometers away from it. To Dr. Miller, this suggested that farms might be more willing to travel farther distances to sell to Crushers.
[424] Ms. Sanderson’s draw area maps displayed the geographic area for 95% of an Elevator’s purchases from farms, using the address of each farm. In these maps, the draw areas for the Moosomin and Virden Elevators could be compared to the draw areas for other Elevators and for Crushers that could be competitors for a farm’s sales. In this way, Ms. Sanderson’s analysis could indicate the number of farms that switch between buyers and the distance travelled to deliver their crops.
[425] Ms. Sanderson’s expert report found that for the crop year 2018-2019, the Moosomin Elevator purchased canola from XX farms and the Virden Elevator purchased from XX farms. The Moosomin Elevator drew XXX of its canola from a distance as far as XX kilometers while the Virden Elevator drew XXX of its canola from farms located as far as Xx kilometers. Both were based on commercial trucking travelling those distances. As for wheat, the draw area for the Moosomin Elevator included XXx farms, while the draw area for the Virden Elevator comprised XXx farms. The Moosomin Elevator drew XXx of its wheat from a distance as far as Xx kilometers while the Virden Elevator drew XXx of its wheat from farms located as far as XXx kilometers.
[426] Ms. Sanderson found that approximately two-thirds of the farms closest to the Moosomin and Virden Elevators (i.e., in the “corridor of concern”) sold crops to other rival Elevators and Crushers. Only four farms sold exclusively to the Moosomin and Virden Elevators in the last three crop years combined. Seven farms sold only to the Moosomin Elevator, while 17 farms sold only to the Virden Elevator. In addition, farms in the corridor of concern sold to more distant Elevators and Crushers. Ms. Sanderson’s written testimony included details for each of the 82 farms within the corridor of concern for both canola and wheat. Ms. Sanderson noted significant deliveries from those farms to Elevators and Crushers outside of Dr. Miller’s proposed geographic market.
[427] Ms. Sanderson prepared heat maps by overlapping all of the draw area maps cumulatively. The heat maps showed areas in darker colours where a higher number of Elevators and Crushers buy grain. Ms. Sanderson concluded that all farm locations within the draw areas for the Moosomin and Virden Elevators had more than six Elevators/Crushers bidding for their wheat and for their canola.
[428] Dr. Miller observed in his reply expert report that Ms. Sanderson’s use of overlapping draw areas did not present any evident consideration of what factors affect a farm’s choice of Elevator or, most importantly, how the farms would likely respond to a price change. Dr. Miller agreed with Ms. Sanderson that some farms scattered throughout the geographic region may elect to work with a more distant Elevator. Dr. Miller observed, however, that the overlap analysis masked that the desirability of travelling to a particular Elevator will differ for farms located at different points. In Dr. Miller’s view, Ms. Sanderson’s draw area maps assumed that every farm customer inside the boundary of the Elevator’s draw area is equally willing to choose that Elevator, which does not address the question posed by geographic market definition, i.e., where those farms would likely turn in reaction to a price increase.
[429] Specifically, Ms. Sanderson submits that what she referred to as “heat maps” provide a count of farms and their locations supplying canola and wheat to Elevators within the overlapping draw areas. However, the Tribunal observes that the overlapping draw areas do not incorporate the volume density supplied, frequency, and the respective geographic locations of those volumes. Such presentation does not assist the Tribunal in visually understanding the concentration of grain supply and corresponding distance to each Elevator as provided by the transaction data.
[430] The Tribunal recognizes that overlapping draw area maps may be useful in initially identifying the possible range of geographic scope of one or more candidate competition markets. They also may identify suppliers or customers who could be affected by a proposed transaction. However, they may be of less help in more precisely defining the scope of the relevant geographic competition market, unless they are coupled with evidence about the reaction of affected suppliers or customers to a change in price or another dimension of competition.
[431] In this case, the panel agrees with Dr. Miller’s concerns that the draw area maps and heat maps have less value on their own, insofar as they treat farms at different distances from an Elevator as likely to react the same way to a change in price. Draw area maps and heat maps may, together, suggest that farms have more ability to switch in the overlapping areas but must be taken as a non-definitive factor in the assessment of a geographic area for competition purposes. In this case, Ms. Sanderson’s draw area maps, on their face, give equal weight to each farm having sold grain to an Elevator regardless of frequency of sales and volumes sold. In sum, the heat maps provided additional information but, on their own, they are of limited assistance because they do not account for volume and the location of the farms from which those volumes are drawn. Both must be weighed with the evidence concerning farms’ behaviour, including from the farmers’ testimonies, the diversion ratio evidence, and other data and information as to preferences and switching.
[432] On this issue, the Tribunal finds Dr. Miller’s observations in his reply expert report to be compelling. Dr. Miller analyzed the percentage of MT of wheat and canola sold to the Moosomin, Virden, and Fairlight Elevators by town. The analysis showed that a farm close to an Elevator is more likely to rely on that Elevator and that more distant farms are less likely to do so. He found that farms close to the centre of the area around the Moosomin, Virden, and Fairlight Elevators had a distinct preference to work with those Elevators. This analysis is consistent with the testimonies of farmers. It also provides some more nuanced insight to assist in understanding the likely behaviour of customers in response to a price change — specifically, whether they are likely to switch Elevators.
[433] The Tribunal further finds that farms, as sellers of grain to Elevators, are less likely to switch to more distant rivals if the farm is near the centre of the geographic market, and more likely to switch as the location moves away from the centre and, the Tribunal infers, away from each of the Moosomin, Virden, and Fairlight Elevators. Consistent with this conclusion and with Dr. Miller’s report, the Tribunal finds that the closer farms are to the Moosomin and Virden Elevators, the more volumes they sell to those Elevators. Conversely, farms located farther away are delivering less to these two Elevators and more to other Elevators.
[434] The Tribunal has considered the “corridor of concern.” Ms. Sanderson’s analysis of the deliveries of wheat and canola from those farms to the Moosomin and Virden Elevators over three crop years showed that many farms sell all, or a very substantial proportion, of those crops to the Moosomin and Virden Elevators. In her report, Ms. Sanderson noted that 54 of the 82 farms in the “corridor of concern” (i.e., about two-thirds) sold wheat or canola to Elevators or Crushers that compete with the Moosomin and Virden Elevators at different points in the three successive crop years used in her analysis (Ms. Sanderson Report at Figures 19–21). She also noted that farms located in proximity adopted different approaches during that period (i.e., decide to sell at least once to a rival, or not). The Tribunal finds this analysis to be highly sensitive to change — a single sale (above the minimum volume Ms. Sanderson used) of either wheat or grain to a rival over the three crop years would change the classification of a farm. Ms. Sanderson’s observation also implies that a substantial number of the farms in the corridor of concern (i.e., 28 of 82, or a third) did not sell to a rival and therefore sold their wheat and canola only to the Moosomin or Virden Elevators (or to both) over three successive crop years. Ms. Sanderson also noted that the Fairlight Elevator is frequently listed as a purchaser from the corridor farms.
[435] As the Commissioner observed during final argument, Ms. Sanderson’s data effectively showed that 75% of the farms in the corridor of concern (i.e., 60 farms out of 80
[7]
) sold all of their wheat exclusively to the Moosomin, Virden, and Fairlight Elevators in any given year. For canola, the corresponding percentage was 55% of the farms selling canola only to those Elevators. The Tribunal notes that for the vast majority of the farms in the corridor, the Whitewood Elevator was not a purchaser over that time period. Finally, for canola, the corridor farms taken as a group sometimes sold to Crushers (e.g., to Bunge at Harrowby).
[436] There are some examples of switching behaviour by farms in the corridor of concern, through the sale of wheat or canola to Elevators or Crushers other than the Moosomin, Virden, and Fairlight Elevators. However, the weight of the evidence is that these farms rely acutely on the Moosomin and Virden Elevators, as well as the Fairlight Elevator, for the sale of their grain, particularly wheat.
(d)
Diversion ratios
[437] Both Dr. Miller and Ms. Sanderson provided diversion ratio calculations. Diversion ratios are calculated to estimate the proportion of a competitor’s customers lost to one or more rivals if that competitor raises its price. Higher diversion ratios imply more substitution between two competitors. That is, in a HMT analysis, if competitor A raises its price for the supply of a product, the diversion ratios calculation shows the resulting switching (i.e., diversion) of customers to the product offered by competitors B, C, and D. Diversion ratios are expressed as a percentage, namely, the proportion of customers diverted to the products of each of competitors B, C, and D, as a percentage of the overall number of diverted customers. The same type of calculation may be done for diversions caused by a monopsonist that lowers its price.
[438] Diversion ratios assist in the assessment of how close two or more competitors may be. As part of a larger model, diversion and information about profit margins may be used to understand the dollar value of diverted sales and specifically, the dollar value of customer purchases that may be recaptured by a merged entity (for example, after the merger of competitors A and B in the example above). The results assist to understand the incentive of a merged entity to raise prices and the predicted price effects of a proposed merger (2011 MEGs at para 6.15).
[439] Dr. Miller’s estimated diversion ratios indicated that many farms viewed the Moosomin and Virden Elevators as substitutes. For wheat, Dr. Miller calculated the diversion ratios from the Moosomin Elevator to the Virden Elevator to be 23.8%, and 16.8% from Virden to Moosomin. In Dr. Miller’s view, the diversion ratios for wheat between the Moosomin and Virden Elevators indicated that they were “relatively close competitors” (Dr. Miller Report at para 114 and Exhibit 11). For canola, the diversion ratios between the Moosomin and Virden Elevators were smaller, at 13.1% and 5.3% respectively. However, the Fairlight Elevator had large diversion ratios with both Elevators, suggesting to Dr. Miller that there was likely indirect competition between the Moosomin and Virden Elevators, through the Fairlight Elevator, for both wheat and canola.
[440] Ms. Sanderson also presented diversion ratios. Methodologically, Ms. Sanderson reran diversion ratios using the formulas in Dr. Miller’s expert report based on farm data within the union of the 90% service areas for the Moosomin, Virden, and Fairlight Elevators.
[441] In that context, Ms. Sanderson did not dispute Dr. Miller’s estimates of the diversion from the Moosomin Elevator to the Virden Elevator for each of wheat, canola including Crushers, and canola excluding Crushers, and the diversion from Virden to Moosomin in the same categories. Ms. Sanderson also presented diversion ratios from each of the Moosomin, Virden, and Fairlight Elevators to many other Elevators and Crushers, in several figures attached to her expert report (Ms. Sanderson Report at Figures 47, 48, 50).
[442] The Tribunal finds that Ms. Sanderson’s diversion ratios indicated, generally, that the Moosomin, Virden, and Fairlight Elevators were relatively close competitors, more so for wheat than for canola. As Ms. Sanderson observed in her expert report, there were smaller diversion ratios for canola from the Virden Elevator to the Moosomin Elevator than from Virden to several other Elevators and Crushers. The same was true for the canola diversion ratios from the Moosomin Elevator to the Virden Elevator. Ms. Sanderson’s diversion ratios to rival Elevators beyond the Moosomin, Virden, and Fairlight Elevators also showed, as she testified, that estimated diversions from both the Moosomin Elevator and the Virden Elevator to all other rivals (in the aggregate) were high for both wheat and canola. For wheat, the diversion ratios found by Dr. Miller were 23.8% from the Moosomin Elevator to the Virden Elevator, and 36.3% to the Fairlight Elevator, for a total of 60.1%. This means that 40% of the sales of wheat diverted from the Moosomin Elevator would go to rival Elevators other than Virden or Fairlight. Conversely, for the Virden Elevator, the diversion ratios for wheat were 16.8% to the Moosomin Elevator and 20.3% to the Fairlight Elevator, for a total of 37.1% of the sales to these two Elevators. This, observed Ms. Sanderson, means that 63% of the diverted sales of wheat from the Virden Elevator would go to Elevators other than Moosomin or Fairlight. In the case of canola, 65% of the sales diverted from the Moosomin Elevator would go to rival Elevators other than Virden or Fairlight, and 77% of the sales diverted from the Virden Elevator would be absorbed by Elevators other than Moosomin or Fairlight. In light of these figures, Ms. Sanderson opined that the diversion ratios to rival Elevators other than Moosomin, Virden, or Fairlight are significant, with many other Elevators and Crushers having diversion ratios similar or higher than those calculated for the Moosomin and Virden Elevators.
[443] In his reply to Ms. Sanderson’s opinion that the geographic market should include more Elevators given their large diversion ratios, Dr Miller referred to the extract from the 2011 MEGs stating that a relevant market is defined as the “smallest group of products” including at least one product of the merging parties, and the “smallest geographic area,” in which a sole profit-maximizing seller would impose and sustain a SSNIP above levels that would likely exist in the absence of the merger. Dr. Miller further explained that defining a relevant market is important because it is impractical to consider all sources of competition. Indeed, doing so would significantly increase the burden of antitrust inquiry, while shedding very little light on the competitive effects of the Transaction.
[444] The Tribunal observes that, in aggregate, 60.1% of switched volumes of wheat are diverted from Moosomin to Virden and Fairlight, and 37.1% are diverted from Virden to Moosomin and Fairlight. The Tribunal finds that the magnitude of these diversions signals a meaningful potential impact of the Acquisition on reducing alternatives for local wheat farms residing in the corridor of concern. But the diversion ratios also reflect the fact that Elevators other than Moosomin, Virden, and Fairlight represent alternatives for the purchase of wheat.
[445] The Tribunal takes less certain direction concerning the purchase of canola, as the diversion ratio data are much less convincing. The Tribunal does not agree with the position, advanced by Ms. Sanderson in her expert report, that higher comparative diversion ratios necessarily leads to the inclusion of all additional Elevators or Crushers on the periphery of the service areas for the Moosomin and Virden Elevators. However, the diversion ratios for canola suggest that it is less likely that the Moosomin and Virden Elevators, acting as a hypothetical monopsonist, would be able to exercise market power in a market defined geographically around the Moosomin, Virden, and Fairlight Elevators. The panel notes that the diversion data suggest that the Whitewood Elevator and the Harrowby Crusher are the next closest competitors to the Moosomin, Virden, and Fairlight Elevators for the purchase of canola, and that other Crushers and Elevators offer alternatives to the farmers for their canola.
(e)
Evidence related to prices and price negotiations
[446] P&H emphasized that its posted prices were set on a centralized basis, and that the vast majority of its sales were at the Cash Price offered each day based on the Basis set each morning and the fluctuating Futures Price.
[447] Dr. Miller testified that he found evidence that farms may sometimes individually negotiate prices with Elevators. Such negotiations may depend on long-standing relationships and revenue dependence, as well as subjective assessments of whether a farm could credibly purchase GHS from another, competing Elevator. Dr. Miller also found documentary evidence that farms negotiate prices that deviate from posted prices in a number of ways, including price matching by Elevators of their competitors’ prices, not charging for certain services, and purchasing grain on the basis of a higher grade price with the intent to blend the grain for later sale. Dr. Miller also noted that a farm’s commitment to purchase crop input products from the Elevator could also affect price.
[448] While the Tribunal agrees that uniformity of prices may be indicative of a geographic competition market, the evidence disclosed that there was a material proportion of transactions that involved a negotiated price. Specifically, P&H confirmed that approximately XX of its transactions occurred at a price set as a result of a successful negotiation of a Cash Price between a farm and the company through its representatives at a particular Elevator. (There was no percentage provided in respect of unsuccessful or attempted price negotiations.)
[449] There was also some evidence that, while negotiating a price to be offered to a farm, P&H representatives were aware of the specific distance from specific farms to the company’s Elevators and to competitor Elevators, and that this information affected the assessment of whether a rival Elevator’s offered price would be matched or not. In determining a potential purchase price for grain, representatives of the Elevator were able to closely analyze circumstances affecting price that were material to an individual farm. Knowledge of the location of the farm, and thus the distance from the farm to each Elevator bidding for the crop, was a key factor in deciding whether to raise the price to be offered for the crop. The internal correspondence recognized that a farm located closer to a rival Elevator would find the rival more attractive as a purchaser at the same price; a higher price would be required to attract volumes of grain away from the rival in order to make up for the time and cost of transportation.
[450] This evidence is indicative of an ability by P&H to discriminate on price when departing from the Basis determined by using its Workback Algorithm. It also further demonstrates the salience of distance (and associated transportation costs) and shows the sophistication of the buy-side analysis of price.
(f)
Mr. Heimbecker’s testimony and P&H business records identifying competitors
[451] The Tribunal appreciates that Mr. Heimbecker identified many competitors to the Moosomin and Virden Elevators during his testimony, and that there are documents to support the view that those Elevators and Crushers are competitors in a business sense. Those competitors were principally those within the draw areas of the Elevators. Like Mr. Heimbecker’s testimony, the parties’ submissions both referred to documents from the files of the merging parties that identified various competitors. Those documents included reports for planning purposes, emails that identified a rival Elevator to which a farm’s sale was lost, and emails for bidding or individual price negotiation purposes (i.e., to obtain supply from a farm).
[452] The Tribunal has considered this documentary evidence and Mr. Heimbecker’s testimony. For the purposes of a competition analysis and geographic market definition, not all business competitors are equal: different competitors may have different abilities to affect the competitive process. Some may have considerable ability to constrain a price increase by the merged entity (or otherwise discipline key dimensions of competition in a market), while others have little or no ability to do so. In this case, applying an HMT approach, the panel finds that the competitive rivals that can constrain a SSNDP by the merged entity do not correspond with competitors from a business perspective. The evidence of internal documents and from Mr. Heimbecker identifying business rivals is relevant but, overall, is a less weighty factor in the Tribunal’s assessment of the geographic scope of the competition markets.
(4)
Conclusion on relevant geographic market(s)
[453] Having considered all the quantitative and qualitative factors described above, the Tribunal concludes that the relevant geographic markets for the purchase of each of wheat and canola are not likely to be smaller or larger than those resulting from the HMT analyses. The Tribunal acknowledges that factors such as the price negotiations on a non-negligible proportion of grain purchases, the diversion ratios between the Moosomin, Virden, and Fairlight Elevators on wheat, the purchases of farmers in the “corridor of concern,” and the evidence on distance travelled by farmers suggest a geographic market definition that would be more localized. Conversely, the Tribunal is not persuaded that the evidence flowing from heat maps or business records identifying numerous competitors is sufficient to justify an expansion of the relevant geographic markets resulting from the HMT analyses. In the end, the Tribunal finds that the evidence related to the geographic markets does not amount to clear and convincing evidence allowing the Tribunal to move away from the results coming from the HMT analyses.
[454] Therefore, on a balance of probabilities, the Tribunal is of the view that the relevant geographic market for the purchase of wheat is more likely than not to include at least the Moosomin, Virden, Fairlight, Oakner (Cargill), Whitewood (Richardson), Elva (Cargill), and Shoal Lake (Richardson) Elevators. With respect to the relevant geographic market for the purchase of canola, it includes at least the Moosomin, Virden, Fairlight, Oakner (Cargill), Whitewood (Richardson), Brandon (Richardson), Melville (G3), Souris East (Viterra), Shoal Lake (Richardson), and Elva (Cargill) Elevators, as well as the Crushers at Harrowby (Bunge), Yorkton (LDC), Velva (ADM), and Yorkton (Richardson).
C.
Has the Commissioner established, on a balance of probabilities, that the Virden Acquisition lessens, or is likely to lessen, competition substantially?
[455] The Tribunal now turns to the main element of the merger provisions, namely, whether the Virden Acquisition lessens competition substantially, or is likely to have that effect.
(1)
Analytical framework
(a)
The statutory language
[456] Subsection 92(1) of the Act provides that the Tribunal may make a remedial order if it finds that a merger or proposed merger prevents or lessens, or is likely to prevent or lessen, competition substantially.
[457] The anti-competitive threshold is directly linked to the concept of market power. As discussed above, market power is the ability to profitably influence the price or non-price dimensions of competition in the market for the supply or purchase of a product. The price and non-price dimensions of competition show the intensity of rivalry between or among competitors in a market (Tervita SCC at para 44; Tervita CT at paras 371–373). A merger will only be found to lessen or prevent competition substantially if it is likely to create, maintain, or enhance the ability of the merged entity to exercise market power, whether unilaterally or in coordination with other firms. The market power analysis in respect of a merger centres on the question of whether the merged entity is able, or is likely to be able, to exercise more market power than it could have exercised in the absence of the merger. When a merger is not likely to have market power effects, “it is generally not possible to demonstrate that the transaction will likely prevent or lessen competition substantially” (2011 MEGs at para 2.8). Without market power effects, section 92 will not generally be engaged (Tervita SCC at para 44).
[458] If there are no market power implications of a merger, there can be no anti-competitive implications. If there are market power implications of a merger, competition can be taken to be lessened or prevented to some extent (see Facey and Brown at p 181; Campbell at p 100). However, it is only where the prevention or lessening of competition is substantial that the Tribunal can intervene under section 92. There can therefore be situations where market power is created, maintained or increased without necessarily resulting in a substantial lessening or prevention of competition.
[459] Subsection 92(2) expressly provides that the Tribunal “shall not find” that a merger or proposed merger prevents or lessens, or is likely to prevent or lessen, competition substantially “solely on the basis of evidence of concentration or market share.” However, depending on the circumstances, post-merger market share may be a useful or reliable indicator of market power (Hillsdown at p 318). In sum, evidence of changes in market shares and concentration levels are relevant and often influential, but not determinative (The Commissioner of Competition v Parkland Industries Ltd, 2015 Comp Trib 4 at para 89; Tervita CT at para 360; Canadian Waste at para 108, 193–195, 204–205, 224; Superior Propane I at paras 126, 304–313).
[460] Section 93 of the Act provides a non-exhaustive list of factors that the Tribunal may consider when assessing whether a merger substantially lessens or prevents competition or is likely to do so. These factors include whether a party is a failing business, the availability of acceptable substitutes, barriers to entry into the relevant market, the extent to which effective competition remains or would remain after a merger, whether the merger would result in the removal of a vigorous and effective competitor, and the nature and extent of change and innovation in a relevant market.
[461] The Tribunal points out that none of the section 93 factors specifically refers to exports or to the pro-competitive dimension or business rationale of a merger. The Tribunal further reaffirms that the intent of the parties is irrelevant in determining whether a merger will likely reduce competition (Canadian Waste at para 118).
(b)
The “substantial lessening” analysis
[462] As the present case solely concerns an alleged substantial lessening of competition, the Tribunal’s analysis will focus on that branch of the assessment of anti-competitive effects. In Tervita SCC, the SCC confirmed that the language in section 92 concerning anti-competitive effects is very close to the corresponding words in paragraph 79(1)(c) of the Act dealing with abuse of dominance (Tervita SCC at para 50). The legal framework applicable to analysis of effects under the two provisions has common features, so court and Tribunal decisions under both provisions provide guidance in relation to the assessment of a substantial lessening of competition.
[463] As the Tribunal discussed in VAA CT at paragraphs 632–644 and in TREB CT at paragraphs 456–483, there are two dimensions in the Tribunal’s substantial lessening of competition analysis. The first considers a forward-looking, counterfactual comparison. The second considers whether the alleged anti-competitive effects are substantial.
[464] First, the Tribunal’s review under section 92 examines whether the merger will give the merged entity the ability to lessen competition, compared with the pre-merger benchmark or “but for” world. The analysis involves a forward-looking counterfactual scenario where the Tribunal compares the state of competition that exists or would likely exist in the presence of the merger with the state of competition that would have likely existed in the absence of the merger (Tervita SCC at paras 51, 54; Tervita FCA at para 108). The focus is on whether the merged entity is likely to be able to exercise materially greater market power than in the absence of the merger, through either materially higher prices or materially lower non-price aspects of competition in the market (Tervita SCC at paras 15, 50–51, 54, 80–81; Tervita FCA at para 108; VAA CT at paras 636, 642; Tervita CT at paras 123, 229(iv), 377)
[8]
. The Tribunal’s approach thus contemplates an assessment that emphasizes the comparative and relative state of competition before and after the merger, as opposed to the absolute state of competition at those two points in time. In a case involving an alleged likely substantial lessening of competition, the Tribunal will assess whether the merger is likely to enable the merged entity to exercise new or enhanced market power (Tervita SCC at para 55, citing Tervita CT at para 368). That is, the Tribunal will consider whether the merger has likely created a new ability to exercise market power, or enhanced the merged entity’s existing ability to exercise market power.
[465] In the second part of its analysis, the Tribunal determines whether the difference between the level of competition in the presence of the merger, and the level that would have existed “but for” the merger, is substantial. The extent of a merger’s likely effect on market power is what determines whether its effect on competition is likely to be “substantial” (Tervita SCC at para 45; TREB FCA at paras 82, 86–92). The issue is whether competition would likely be substantially greater, “but for” the implementation of the merger or proposed merger, through the merged entity’s ability to profitably influence price, quality, service, advertising, innovation, or other dimensions of competition (Canadian Waste at paras 7, 108; Di Domenico at p 554). For a merger to be subject to a remedial order by the Tribunal, it is not enough to demonstrate that an actual or likely lessening of competition will result, or the mere creation or enhancement of market power. In a merger review, the Tribunal’s assessment focuses on “whether the merged entity is likely to be able to exercise materially greater market power than in the absence of the merger” [emphasis added] (Tervita SCC at para 54, citing Tervita CT at para 367).
[466] Again, the test is relative and requires an assessment of the difference between the level of competition in the actual world and in the “but for” world (TREB FCA at para 90). What is substantial is not defined in the Act. The Tribunal may consider evidence of market shares and concentration levels, together with the factors listed in paragraphs 93(a) to (g.3) of the Act and, under paragraph 93(h), “any other factor” relevant to competition in a market that is or would be affected by the merger or proposed merger. In each given case, all relevant indicators of market power need to be considered, but the relevance and weight to be assigned to each indicator will vary with the factual context. There is no precise scale by which to measure what is substantial, and this determination will be “highly contextual” (Facey and Brown at p 184).
[467] In conducting its assessment of substantiality, the Tribunal will look at three key components, namely, the degree, scope, and duration of the lessening of competition (Tervita SCC at para 45; VAA CT at para 640).
[468] With respect to degree, or magnitude, the Tribunal assesses whether the impugned merger is enabling or is likely to enable the merged entity respondent to exercise materially greater market power than in the absence of the merger (Tervita SCC at paras 50–51, 54). When assessing whether competition with respect to prices is or is likely to be lessened substantially, the test applied by the Tribunal is to determine whether prices are or likely would be materially higher than in the absence of the merger. With respect to non-price dimensions of competition, such as quality, variety, service, or innovation, the test applied is to determine whether the level of one or more of those dimensions of competition is or likely would be materially lower than in the absence of the merger (Tervita SCC at para 80; TREB FCA at paras 89–92; Tervita CT at paras 123–125, 376–377; VAA CT at para 642).
[469] In assessing whether the degree, or magnitude, of lessening of competition is sufficient to be considered “substantial,” the Tribunal will consider the overall economic impact of a merger in the relevant market. Proof of a likely post-merger price increase must be assessed in relation to its materiality in the specific market at issue, the nature and extent of pre- and post-merger competition, and the rest of all the quantitative and qualitative evidence related to the affected dimensions of competition.
[470] On the price dimension of competition, the Tribunal has not found it useful to apply rigid numerical criteria in conducting this assessment. In short, there is no specific quantum of price variation implying that a merger lessens competition substantially. The Tribunal agrees with the 2011 MEGs that there is no rigid “numerical threshold” for a material price increase (2011 MEGs at para 2.14; see also Hillsdown at p 329). The Tribunal pauses to underline that the use of a 5% increase in price for the purposes of the HMT analysis must not be confused with the materiality of a price increase under the substantial lessening of competition analysis. The conceptual SSNIP threshold of 5% in the HMT analysis for market definition purposes is distinct from the assessment of substantiality of anti-competitive effects. It is therefore incorrect to state that the Commissioner must adduce quantitative evidence showing a 5% variation in post-merger prices in order to establish a lessening of competition that is “substantial.” The required magnitude of a “substantial” price increase will instead vary from case to case and will depend on the facts of each case (Tervita SCC at para 46; TREB FCA at para 88; Hillsdown at pp 328–329). A substantial price variation can be less than 5%.
[471] In fact, as Chief Justice Crampton explained in his concurring opinion in Tervita CT, the degree of market power used in assessing whether competition is likely to be prevented or lessened substantially must be recalibrated downward when a 5% price increase is used to assess the degree of market power held by a hypothetical monopolist for the purposes of the HMT analysis and the SSNIP threshold. At paragraphs 376–377 of Tervita CT, he said:
[376] […] However, given that the Tribunal has now embraced the hypothetical monopolist framework and the SSNIP test for market definition, it is necessary to revisit this definition of substantiality. This is because if the degree of market power used to define relevant markets is the same as the degree of market power used to assess competitive effects, a merger would not be found to be likely to prevent or lessen competition substantially unless the degree of new, enhanced or maintained market power of the merged entity is the same degree of market power held by as [sic] the hypothetical monopolist that was conceptualized for the purposes of market definition.
[377] Accordingly, the degree of market power used in assessing whether competition is likely to be prevented or lessened substantially must be recalibrated downwards. That recalibrated degree of market power is a level of market power required to maintain prices materially higher, or to depress one or more forms of non-price competition to a level that is materially lower, than they likely would be in the absence of the merger. […]
[Emphasis in original.]
[472] In sum, the substantiality level contemplated by the “substantial lessening of competition” analysis can be lower than the level under the HMT analysis and the SSNIP threshold.
[473] It must also be emphasized that there is no requirement for the Tribunal to find a likely increase in price; it is sufficient for the Tribunal to conclude that the merged entity has the ability to increase price or to reduce quality, service, or product choice.
[474] Turning to scope, the assessment involves determining whether the lessening of competition affects the entire relevant market or a material part of it. If the alleged anti-competitive effects do not extend throughout the totality of the relevant market, the Tribunal will assess their scope and whether they extend throughout a “material” part of the market, or in respect to a material volume of sales / business (Tervita FCA at para 108; Tervita CT at paras 375, 378).
[475] With respect to duration, the test applied by the Tribunal is whether a material increase in price or material reduction in non-price dimensions of competition resulting from a merger is likely to be maintained for approximately two years (Tervita SCC at para 80; Tervita CT at para 123).
[476] In assessing substantiality and its various components, the Tribunal considers quantitative evidence, qualitative evidence, or both, related to the price and non-price dimensions of competition (TREB FCA at para 16; VAA CT at paras 124, 639; TREB CT at paras 469–471). In Tervita SCC, the SCC held that the Commissioner was not, in law, required to quantify any anti-competitive effects under section 92 (Tervita SCC at paras 121–122, 166; TREB FCA at paras 99–100; TREB CT at para 469). That said, in all situations, the Commissioner must always adduce sufficiently clear and convincing evidence, and he bears the burden to demonstrate, on a balance of probabilities, that the merger lessens or is likely to lessen competition substantially, as well as the basic facts of the “but for” scenario that are required to make that demonstration (Tervita SCC at paras 65–66; TREB FCA at paras 87; Tervita FCA at paras 107–108; VAA CT at para 644).
(2)
Parties’ positions
(a)
The Commissioner
[477] The Commissioner submits that the Virden Acquisition is likely to cause a substantial lessening of competition in the relevant markets owing to the elimination of a vigorous and effective competitor, namely, the Virden Elevator. The Commissioner claims that both the quantitative and qualitative evidence demonstrates that farmers in the relevant markets will pay materially more for GHS for wheat and canola over the next two years and will lose other impactful aspects of competition. With the control of the Virden Elevator, says the Commissioner, P&H has the ability and incentive to unilaterally exercise market power in the relevant markets. The Commissioner contends that the lessening of competition is substantial in terms of magnitude, duration, and scope: it adversely impacts competition to a degree that is material, the duration of the anti-competitive effects is substantial, and the anti-competitive effects extend to a substantial part of the relevant markets.
[478] In his final submissions, the Commissioner argued that the substantial lessening of competition is demonstrated by the following elements, which echo many of the factors listed in section 93 of the Act: 1) the high margins at the Virden Elevator, which provide direct evidence of P&H’s existing market power; 2) P&H’s ability to engage in price discrimination; 3) P&H’s high market shares in the relevant markets; 4) the removal of the vigorous and effective competition to the Moosomin Elevator that the Virden Elevator provided prior to the Acquisition; 5) the material impact of the Virden Acquisition on the price for GHS for wheat and canola; 6) the postponement of the planned expansion of the Moosomin Elevator that would have made P&H a more effective competitor to the Virden Elevator in the absence of the Acquisition; 7) the inability of Viterra’s Fairlight Elevator to constrain an increased exercise of market power by P&H; and 8) the existence of high barriers to entry and expansion.
[479] The Commissioner further submitted that other more distant Elevators, canola Crushers, and direct purchasers of wheat or canola are unable to constrain an exercise of market power by P&H as they do not have sufficient capacity and farmers would have to incur higher transportation costs to deliver their wheat and canola to these locations. While the Commissioner’s submissions mostly focused on the anti-competitive price effects of the Virden Acquisition, the Commissioner maintains that quantified price effects are only one element of the substantial lessening of competition caused by the Acquisition. According to the Commissioner, there is also significant other evidence demonstrating that the contemplated price effects are material to the farmers.
[480] The Commissioner submits that the Tribunal should adopt a framework that allows for an economic analysis that can credibly assess the impact of local competition between Elevators that was lost when P&H acquired the Virden Elevator from LDC. The Commissioner considers that such local competition is expressed through the Basis. He states that the main issue to be determined by the Tribunal can be summarized as follows: when competition effectively takes place on, and affects one component of, the overall final price of a product, as he suggests is the case here, how should the Tribunal assess and measure the magnitude of harm and the materiality required for the lessening of competition to be substantial?
[481] In support of his arguments on the substantial lessening of competition, the Commissioner relies on three pillars of evidence: the expert evidence of Dr. Miller (including his merger simulation model), the fact witnesses (notably, the farmers who testified at the hearing), and the documentary evidence.
[482] The Commissioner does not dispute that he has the burden to adduce sufficiently clear and convincing evidence to demonstrate, on a balance of probabilities, that competition is or is likely to be lessened substantially as a result of the Virden Acquisition.
(b)
P&H
[483] P&H responds that the Virden Acquisition does not, and is not likely to, lessen competition substantially in any relevant market. More specifically, P&H submits that the Commissioner has failed to meet his burden to prove, on a balance of probabilities, his alleged substantial lessening of competition. According to P&H, the evidence on the record does not establish that the markets at issue would be substantially more competitive, “but for” the Virden Acquisition.
[484] In its Response, P&H denied that the Virden Acquisition creates, enhances, or maintains monopsony power in any properly defined market for the purchase of wheat or canola. P&H argued that it will continue to face vigorous and effective competition from numerous competing Elevators and Crushers located in Manitoba and Saskatchewan. P&H further claimed that barriers to entry and expansion are low and that rival Elevators have excess capacity, allowing them to expand their purchases of wheat and canola and to constrain any attempt by P&H to exercise monopsony power. P&H also submitted that the predicted price variations determined by Dr. Miller are immaterial and unlikely.
[485] In its closing submissions, P&H elaborated by focusing on the fact that: 1) barriers to entry and expansion are low; 2) the Virden Elevator has become and will remain a vigorous and effective competitor further to the Transaction; and 3) the Transaction enhances non-price competition.
[486] In support of its arguments on the absence of any substantial lessening of competition, P&H relies on Ms. Sanderson’s expert evidence, more particularly her critique of the price effects alleged by Dr. Miller and her post-Acquisition price analysis, and on the evidence provided by the farmer witnesses.
(3)
Tribunal’s assessment
[487] The Tribunal notes at the outset that the evidence adduced by the Commissioner on the substantial lessening of competition primarily focused on the quantification of the alleged price effects of the Virden Acquisition. As the Commissioner said in his oral submissions, his demonstration that the lessening of competition is substantial was mostly done through Dr. Miller’s quantification work including evidence such as market shares and margins. As part of its assessment, the Tribunal has therefore considered whether the Cash Prices paid by P&H to the farmers for their wheat or canola are or would likely be materially lower, “but for” the Virden Acquisition. The Tribunal also assessed other evaluative factors raised by the Commissioner and covered by section 93 of the Act. These factors notably included likely entry and expansion, excess capacity, and the extent of any remaining vigorous and effective competitors.
[488] For the reasons discussed below, the Tribunal concludes that the Commissioner has not demonstrated, on a balance of probabilities, that the Virden Acquisition lessens, or is likely to lessen, competition substantially in the relevant markets. The Tribunal accepts that the joint control of the Virden and Moosomin Elevators by P&H has and will continue to have some limited adverse effects on competition in the purchase of wheat. However, on the evidence before it, the Tribunal is not persuaded that such lessening of competition reaches or is likely to reach the substantiality required by section 92 of the Act.
[489] The Tribunal also acknowledges that the materiality level to assess the substantial lessening of competition varies from case to case, and that a lower materiality level could apply in cases, such as this one, where competition between rivals takes place, at least in part, on one more specific component of the overall final price of a product. However, the Tribunal observes that, even though the Commissioner insisted that competition between Elevators and Crushers revolved around the Basis, he has not provided any compelling submissions, nor any clear and convincing evidence, supporting a particular materiality level that the Tribunal should apply in the current circumstances. Moreover, even considering that some competition between Elevators effectively takes place on one component of the overall price of grain, namely, the Basis, the Tribunal finds that the low magnitude of harm revealed by the evidence is not enough to meet the materiality required for the lessening of competition to be substantial, whether in relation to the Basis or the Cash Price.
(a)
P&H’s alleged pre-existing market power
[490] The Commissioner argues that P&H already had existing market power prior to the Acquisition and now has the ability to increase this market power by virtue of its ownership of the Moosomin and Virden Elevators. He claims that two pieces of evidence demonstrate P&H’s pre-existing market power: the high margins prevailing at the Virden Elevator and P&H’s ability to price discriminate.
[491] In his expert report, Dr. Miller calculated that the Virden Elevator earned a 55.2% margin on GHS for wheat and a 39.3% margin on GHS for canola. In Dr. Miller’s opinion and experience, those are relatively high margins “consistent with localized competition rather than significant competition from many distant competitors” (Dr. Miller Report at para 72). Dr. Miller’s margins are economic margins. Ms. Sanderson did not provide any specific margin estimates of her own. Apart from Dr. Miller’s estimates, no other evidence was provided to the Tribunal with respect to Elevators’ margins on the purchase of grain.
[492] To calculate his margins for the Virden Elevator, Dr. Miller identified those Virden Elevator costs which are marginal or incremental, and he excluded fixed costs. Ms. Sanderson did not dispute Dr. Miller’s categorization of the fixed and marginal costs, but she criticized his estimated margins on the ground that they were overstated and failed to include certain freight costs and other costs relating to export terminal operations.
[493] The Tribunal accepts Dr. Miller’s estimated margins for the Virden Elevator, based on his allocation of both revenue and costs to Elevators for the purposes of estimating marginal costs. The Tribunal agrees that the marginal costs related to GHS are a satisfactory proxy for the marginal costs associated with the purchase of grain. Although Dr. Miller’s allocations were based on variable costs associated with the delivery of GHS, the same operating activities are also closely associated with the purchase of grain. The Tribunal finds that the variable costs allocated by Dr. Miller are properly allocable to the purchase of grain (and respective revenue generated) at an Elevator, in contrast with freight and other costs that are properly attributable to the marginal cost for (and revenue generated by) the sale and distribution of grain downstream at export terminals and other destinations.
[494] Turning to price discrimination, the evidence from discovery is clear that P&H knows the location of its customers and has the ability to use that information to engage in price discrimination. To the extent that Elevators sometimes negotiate individual prices with farmers, a price-discrimination framework may thus be more descriptive of the grain industry. However, the evidence on the record indicates that price negotiations between farms and P&H only occur for about XX of P&H’s transactions with farmers, with the vast majority (i.e., the remaining Xxx) of transactions between farmers and the Elevator being done on the basis of posted prices. This XX percentage is arguably conservative, as it does not include those transactions where farms attempted price negotiations but were unsuccessful. The Tribunal also points out that, while there is some evidence of price discrimination, Dr. Miller stated that a posted price model was the appropriate framework to study how prices are set in the grain industry (Dr. Miller Report at paras 140–142). The Tribunal considers that the evidentiary record in this case demonstrates that P&H has some ability to price discriminate based on its customers’ locations, and that it can exercise that ability when it is in its interest to do so. However, the evidence shows that P&H’s actual use of this ability is limited.
[495] The Tribunal agrees that high margins and the ability to price discriminate can constitute direct evidence that P&H has some pre-existing market power. This was recognized by the Tribunal in Tele-Direct at paragraphs 286 and 297. In that case, the Tribunal looked at Tele-Direct’s behaviour towards consultants and whether it had abused its dominant position. The Tribunal found that “[w]here a firm with a high degree of market power is found to have engaged in anti-competitive conduct, smaller impacts on competition resulting from that conduct will meet the test of being “substantial” than where the market situation was less uncompetitive to begin with” (Tele-Direct at para 758). The Tribunal points out that in that case, Tele-Direct was found to have “overwhelming” market power.
[496] Based on the evidence before it, the Tribunal is satisfied that the high margins calculated by Dr. Miller at the Virden Elevator constitute clear and convincing evidence of P&H’s pre-existing market power. Similarly, while the evidence with respect to P&H’s ability to price discriminate is not as compelling given that the evidence showed a practice affecting only a limited portion of P&H’s purchases of wheat and canola, the Tribunal is satisfied that, on a balance of probabilities, this evidence also supports a finding of some pre-existing market power for P&H.
(b)
Price effects
[497] With respect to the price effects, the Commissioner relied on the expert evidence of Dr. Miller to support his position that prices for GHS are or will likely be materially higher than they would have been in the absence of the Virden Acquisition.
[498] In his analysis, Dr. Miller found that the diversion ratios between the Moosomin and Virden Elevators ranged between 15% and 25% for wheat and between 5% and 15% for canola. More detail about these diversion ratios were provided above in the Tribunal’s discussion of the geographic market definition. Dr. Miller used these diversion ratios to quantify the UPP created by the Acquisition. The UPP is a tool that is often used in merger review to approximate the incentive for the merging parties to unilaterally increase prices following a merger, and to measure such price effects. Dr. Miller computed several measures of UPP, all of which showed that prices for GHS would likely rise as a result of the Acquisition, for both wheat and canola. His results suggested that the Transaction generates impetus for price increases, with UPPs of over CAD $2.50 per MT for wheat and over CAD $0.30 per MT for canola. He estimated the gross UPP indices (“GUPPI”) at over 9% for wheat and over 1% for canola.
[499] As explained by both Dr. Miller and Ms. Sanderson in the concurrent evidence session, the magnitudes of UPP, GUPPI, and price effects from merger simulations depend on the amount of diversion between the merging firms and on the mark-up (or margin). Holding all else equal, greater diversion ratios between the merging firms and higher margins will increase UPP, GUPPI, and the simulated price effects.
[500] Using his diversion ratios and his estimate of the Virden Elevator’s margin, Dr. Miller constructed a merger simulation model to quantify the price impact of the Virden Acquisition on farmers. A merger simulation model is a widely accepted econometric method for calculating the predicted price effects from a merger, and to quantify changes to consumer surplus, profit, and the DWL. It is not disputed that the models constructed by Dr. Miller for his farm choice and his merger simulation are standard economic models, and that Dr. Miller’s analysis reflects the principles established in the economic literature. The Tribunal pauses to mention that Dr. Miller’s merger simulation model was used both for market definition purposes and for measuring the anti-competitive effects of the Acquisition.
[501] The Tribunal notes that there was no disagreement between Dr. Miller and Ms. Sanderson on the calculation of the diversion ratios that went into the merger simulation model; those diversion ratios came out of the transaction-level data.
[502] For his substantial lessening of competition analysis, Dr. Miller considered a large number of competing entities, namely, 15 Elevators in the case of wheat and, for canola, 15 Elevators and five Crushers. These were Elevators and Crushers to which there are positive deliveries of canola or wheat made by farms located within Dr. Miller’s Farmer Region. In sum, Dr. Miller’s merger simulation model included Elevators and Crushers that come from both inside and outside the defined geographic markets resulting from his HMT analysis.
[503] There was agreement by both experts that they were able to interpret the data provided and that Dr. Miller relied on a rich and robust data set for his merger simulation model. P&H had voiced concerns about the fact that the Commissioner did not collect data from two Paterson Elevators located at Carnduff and Binscarth, nor from the Cargill Elevator at Nesbitt. All six farmers who testified at the hearing said they do not sell to any of these three Elevators and did not produce receipts showing sales to these Elevators. In addition, no reference was made by Ms. Sanderson to these Elevators as a competitor in any of the Moosomin business documents. The Tribunal is therefore satisfied that, in the end, there were no real issues with alleged missing data in Dr. Miller’s model.
[504] Dr. Miller’s merger simulation model provided both relative and absolute values for his predicted price effects. Dr. Miller and Ms. Sanderson agreed on the magnitude, in absolute dollar terms, of the price effects predicted by Dr. Miller’s merger simulation model. They however disagreed on the percentage of the price variations, as the relative price changes varied significantly depending on the denominator being used (i.e., the price of GHS or the Cash Price).
[505] In light of the foregoing, the Tribunal is satisfied that it can accept Dr. Miller’s merger simulation model and its absolute results.
[506] Both Dr. Miller and Ms. Sanderson further acknowledged that, if the diversion ratios and the mark-ups are positive, a merger simulation model will always predict price increases whenever efficiencies are not directly modeled (Ms. Sanderson Report at para 78; Consolidated Transcript, Public, at pp 1525–1526, 1529; Consolidated Transcript, Confidential A, at pp 1711, 1718–1719, 1784–1785, 1871–1872). This is a reflection of economic theory, which says that, when there is competition, there will be lower prices. In other words, a merger that reduces competition in the sale of a product will raise prices to some degree. The Tribunal must, however, determine whether the predicted price increases — or, in the case of a monopsony, price decreases — are meaningful and substantial.
[507] Further to its review of the evidence, the Tribunal is not persuaded that the Virden Acquisition is decreasing or will likely decrease the Cash Prices for wheat or canola to a material degree in the relevant markets, relative to the prices that likely would have existed “but for” the Acquisition. Stated differently, the Commissioner has not demonstrated that, “but for” the Virden Acquisition, the prices received by farmers for their wheat and canola are or would likely be materially lower. This is the case for both the relative and absolute measures coming out of Dr. Miller’s evidence.
(i)
Relative measures
[508] The Tribunal first considers the relative price changes predicted by Dr. Miller’s model, as this is typically how expected price effects are measured and assessed by the Tribunal. The relativity of predicted price changes is an important benchmark, as price effects are not measured in a vacuum, but against a certain reference or base price. Price variations will be material when they represent a meaningful proportion of the reference price.
[509] Dr. Miller’s merger simulation model based on the farmers’ choices of Elevator predicts that the price of GHS for wheat will increase by CAD $2.49 per MT for the Moosomin Elevator, and by CAD $2.07 per MT for the Virden Elevator (Dr. Miller Report at Exhibit 14). This corresponds to 7.1% and 7.6% price increases, respectively, relative to the reference pre-Acquisition price of GHS for wheat at each Elevator. Turning to the price of GHS for canola, Dr. Miller’s projected price increases will be between CAD $0.91 per MT and CAD $1.21 per MT at the Moosomin Elevator, and between CAD $0.25 per MT and CAD $0.35 per MT at the Virden Elevator. The range reflects the different values for canola including or excluding Crushers. These observed variations amount to a 7.3%–9.7% and 1.3%–1.7% increase in the price of GHS for canola at the Moosomin and Virden Elevators, respectively.
[510] Dr. Miller’s expert opinion is that the Virden Acquisition allows P&H to charge farmers 7% to 8% more to handle their wheat, and between 1% to 7% more to handle their canola (using the data for canola including Crushers) (Dr. Miller Report at para 6). According to Dr. Miller and the Commissioner, these results from Dr. Miller’s merger simulation show a material increase in the price of GHS per MT, for both wheat and canola, when considered against the price farmers pay for GHS.
[511] The Tribunal notes that the highest predicted price increase of CAD $2.49 per MT for wheat represents a variation of 6.8 cents per bushel (i.e., CAD $2.49 / 36.7444 bushels). The corresponding highest predicted price increase of CAD $0.91 per MT for canola (including Crushers) equates to 2.1 cents per bushel (i.e., CAD $0.91 / 44.092 bushels). Both of these highest price variations were measured for the Moosomin Elevator.
[512] Ms. Sanderson further calculated that, for wheat, the average price increase predicted by Dr. Miller for the Moosomin, Virden, and Fairlight Elevators would be CAD $1.39 per MT (or 4 cents per bushel). The average price increase for canola would be CAD $0.23 per MT (or 1 cent per bushel). When expressed as a percentage of the weighted average imputed prices for GHS at the Moosomin, Virden, and Fairlight Elevators, the price effects calculated by Dr. Miller are, on average, 1.51% for canola and 4.62% for wheat (Ms. Sanderson Slides at pp 93–94).
[513] However, as Ms. Sanderson points out in her evidence, when the predicted price changes found by Dr. Miller are expressed in relation to the Cash Prices, the picture of the relative price variations is quite different.
[514] In the period of reference used by Dr. Miller, the weighted average price for wheat was CAD $229.73 per MT at the Moosomin Elevator, and CAD $239.11 per MT at the Virden Elevator. For its part, the weighted average price for canola was CAD $461.46 per MT at the Moosomin Elevator, and CAD $452.80 per MT at the Virden Elevator. Therefore, for wheat, Dr. Miller’s predicted price variations of CAD $2.49 per MT for the Moosomin Elevator and CAD $2.07 per MT for the Virden Elevator represented changes of only 1.1% and 0.9%, respectively, compared to the pre-Acquisition Cash Price for wheat at each Elevator. With respect to canola including Crushers, Dr. Miller’s projected price variations of CAD $0.91 per MT for the Moosomin Elevator, and CAD $0.25 per MT for the Virden Elevator amounted to relative changes of 0.2% and 0.05%, respectively, compared to the pre-Acquisition Cash Price for canola at each Elevator. When expressed as a percentage of the average Cash Prices at the Moosomin, Virden, and Fairlight Elevators, the average relative price effects calculated by Dr. Miller are 0.05% for canola and 0.60% for wheat (Ms. Sanderson Slides at p 94).
[515] The Commissioner therefore asks the Tribunal to find a substantial lessening of competition in a situation where his expert’s average predicted price increases represent 1.51% of his imputed price for GHS for canola, expressed as a percentage of the average price for GHS at the Moosomin, Virden, and Fairlight Elevators, and 4.62% of his imputed price for GHS for wheat at those same Elevators. When they are expressed in relation to the product market and the appropriate reference price identified by the Tribunal — namely, the Cash Price —, Dr. Miller’s predicted price changes represent between 0.05% and 0.2% of the Cash Price for canola, and between 0.60% and 1.1% of the Cash Price for wheat. The Commissioner claims that, even if the purchase of grain and Cash Prices should be the denominator, these price effects are still material when viewed against the qualitative evidence.
[516] With respect, the Tribunal disagrees. The Tribunal is of the view that price changes of this magnitude (i.e., at most 1.1% of the Cash Price for wheat and at most 0.2% of the Cash Price for canola) cannot be qualified as “material.” On the contrary, the Tribunal finds that predicted price variations representing such a small fraction of the pre-Acquisition price for wheat or canola at the Moosomin or Virden Elevators are immaterial, especially in light of the fact that a merger simulation model will always predict a price increase. For the purchase of canola, price variations of 0.2% or less (or between one and two cents a bushel) are de minimis. For the purchase of wheat, price variations reaching at most 1.1% (or a maximum of seven cents a bushel) are very minor and far from substantial in this market. Indeed, the Tribunal observes that, in his submissions at the hearing, the Commissioner admitted that such relative price changes were “small” when expressed in terms of percentage. Hence, the Commissioner’s focus on the absolute values of the predicted price changes, to which the Tribunal will turn below.
[517] The Tribunal accepts that the Basis plays a certain role in the competition between Elevators at the local level. The evidence indicates that there can be adjustments to the Basis or to the Cash Price after or in addition to changes in the Futures Price. In some cases, the Basis fluctuates for reasons other than a change in the Futures Price, such as negotiations between farms and Elevators or limited-tonne and limited-time specials offered by the Elevators. The Tribunal also accepts that the price variation threshold can certainly be lower than 5% (contrary to P&H’s argument) in order to meet the substantiality level. The Tribunal is further mindful of the fact that, when a firm has high pre-existing market power, smaller impacts on competition can be enough to meet the test of substantiality (Tele-Direct at para 758). The Tribunal pauses to note that, while it finds that P&H had “some pre-existing market power” in this case, the facts do not support a conclusion that P&H had “high” market power and certainly not “overwhelming” market power as in Tele-Direct.
[518] However, the Commissioner has not presented any compelling argument nor any clear and convincing evidence regarding the materiality level (in terms of percentage) that should apply to the substantial lessening of competition analysis in this case. More specifically, the Commissioner has not made submissions regarding the relative materiality level that should apply in a case where competition allegedly takes place on one component of the final price for wheat or canola, namely, the Basis. Similarly, the Commissioner has submitted no analysis nor any evidence to demonstrate that, in the particular circumstances of this case, the acceptable materiality level for a price decrease could be as low as around 1% or less.
[519] In fact, the Tribunal is not aware of any merger cases, in Canada or in any other jurisdiction, where a court or tribunal has recognized that a predicted price effect revolving around 1% could be enough to meet the test of substantiality. Indeed, since merger simulation models predict price increases (as discussed above), the Tribunal is of the view that, absent expert evidence allowing it to conclude differently, relative price variations predicted by a merger simulation model have to be more than 1% in order to have any significance or materiality.
[520] For all the above reasons, the Tribunal agrees with P&H and Ms. Sanderson that the relative effect of the Virden Acquisition on the Cash Prices paid by P&H for wheat or canola is not material.
(ii)
Absolute measures
[521] The Commissioner also takes the position that the absolute price variations observed by Dr. Miller are material. In his submissions, the Commissioner relied on the absolute magnitude of Dr. Miller’s predicted price increases and what he claimed was their resulting materiality. The Commissioner argued that, in this case, the Tribunal should prefer and adopt an absolute notion of materiality with respect to the price effects and consider the impact that the Acquisition will have on farmers, in terms of changes in “cents per bushel” they will pay for GHS or receive for their grain. The Commissioner submits that the absolute amount of the effects measured by Dr. Miller is evidence of a substantial lessening of competition. The price increases projected by Dr. Miller, says the Commissioner, are also well above 2 cents per bushel (or, equivalently, CAD $0.73 per MT for wheat and CAD $0.88 per MT for canola), and this is sufficient to demonstrate materiality.
[522] As mentioned above, Dr. Miller’s highest predicted price variations are 6.8 cents per bushel for wheat and 2.1 cents per bushel for canola. Dr. Miller and Ms. Sanderson agree that these absolute effects are the same regardless of whether the product market is GHS or the purchase of grain. In sum, these price effects are not dependent upon the definition of the relevant market or on the selection of the SSNIP. The relevant market only impacts the computation of the price effects in relative terms, and the experts indeed disagree on the percentage of the price variations (in terms of the imputed price of GHS or the Cash Price).
[523] The Commissioner further claims that price increases of 2 cents to 7 cents per bushel are material when viewed against the qualitative evidence and the pre-existing market power of P&H. The Commissioner submits that the materiality of the price effects is enhanced when all of the other evidence of a substantial lessening of competition is considered, namely, P&H’s pre-existing high margins, high market shares, and ability to price discriminate, the removal of the Virden Elevator as a vigorous and effective competitor, and the loss of competition between the Virden and Moosomin Elevators as a result of the Moosomin Elevator’s delayed expansion.
[524] The Tribunal is not persuaded by the Commissioner’s submissions and evidence. More specifically, the Tribunal is not convinced that there is evidence showing, on a balance of probabilities, that 2 cents a bushel “matter” to farmers. The evidence is only in respect of amounts much higher than that, and higher than Dr. Miller’s predicted price increases. The Tribunal agrees with P&H that arguing that a 2 cents per bushel price increase, or even a 7 cents price increase, is material is inconsistent with the farmers’ evidence and with the high volatility of prices in the grain industry, often vacillating by plus or minus 10 cents a bushel in a day.
[525] The Commissioner first relies heavily on the so-called “2 cents challenge.” The “2 cents challenge” refers to an email sent by X XX XX XX XXX XX XX XX XX XXXXXXXXXXXX XXXXXXXXXXXX to XXXXXXXXXX and XXXXXXXXXXXXXX XXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX in the regular course of business (Commissioner Read-In at p 643). In that email, XXXXXXX said that 2 cents per bushel translated into CAD $3.2 million in profitability for P&H. The Commissioner claims that this evidence is very important to assessing materiality because XXXXXXXx posted that if P&H bought each bushel 2 cents cheaper, it would add more than CAD $3 million in profitability to P&H. Moreover, when Mr. Heimbecker was asked on discovery what CAD $3.2 million in profit meant to him, he said that such a sum of money was “not insignificant” (Commissioner Read-In at p 178).
[526] The Commissioner adds that 2 cents per bushel is a reasonable threshold for materiality as it represents 2.1% of GHS for wheat and 7.1% of canola handling prices for the Moosomin Elevator, and 2.7% of GHS for wheat and 4.4% of canola handling prices for the Virden Elevator.
[527] The Commissioner also refers to another email where XXXXXXXXXXXXX from P&H said that he does not need to be XXXXX per bushel higher than his competitors on wheat, and that the damage of such a discount to P&H’s buying program would be XXXXXXXXXXXXX (Commissioner Read-In at p 800). On the basis of this evidence, the Commissioner maintains that P&H cares about pennies on the bushel.
[528] The Tribunal does not find this evidence convincing on the issue of materiality of price effects in the context of its substantial lessening of competition analysis. For the purpose of section 92 of the Act and the Tribunal’s assessment of anti-competitive effects, materiality is analyzed in relation to a market situation, to the competitive behaviour of market participants as a result of a merger, and to competition in the relevant market. It is not analyzed in relation to a firm’s overall profitability or bottom line. In this case, it is analyzed in relation to the situation in the relevant geographic markets for wheat and canola as a result of the Virden Acquisition. The Tribunal appreciates that 2 cents a bushel, when projected on the overall profitability of P&H, may have some significance for P&H. But the fact that 2 cents a bushel may be important to P&H’s overall bottom line does not constitute clear and convincing evidence that such an amount is or would likely be “material” with respect to a lessening of competition in the purchase of wheat or canola in the relevant markets defined above.
[529] The materiality of predicted price changes, whether relative or absolute, is not to be measured in relation to the general profitability of one specific producer or supplier, or one specific customer. This is even more so when the Commissioner’s Application only challenges one specific portion P&H’s business, namely, the acquisition of only one Elevator out of an overall network of 29. The issue to be determined by the Tribunal is whether the adverse price effects on competition are material in relation to the market and industry at stake. The “2 cents challenge” email offers no clear and convincing evidence to support the Commissioner’s position on the alleged price effects attributable to the Virden Acquisition.
[530] The Tribunal further observes that virtually any business person would say that any additional cent of profit per unit of product sold does matter and has some significance for his or her business. In the Tribunal’s view, the assessment of materiality was not meant to be reduced, and cannot be reduced, to a single comment such as those made by XXXXXXX.
[531] With respect to the email from XXXXXXXx, the Tribunal notes that the XXXXx XXXXXxxXX comment was not made in reference to a specific benchmark, let alone to a 2 cents per bushel benchmark. It was generally made in reaction to sales representatives saying that “farmers are getting better prices somewhere” with no information to support such claims. Moreover, when XXXXXXXX referred to a particular amount P&H should not contemplate, he said that P&H did not need to be XXXXX per bushel higher than its competitors. XXXXX per bushel is far higher than the 2 cents or even 7 cents a bushel measured by Dr. Miller.
[532] The Commissioner further submits that cents per bushel matter not only to P&H, but also to farmers. In that regard, the Commissioner relies on several extracts and comments made by the farmer witnesses regarding the price changes they face in their business.
[533] Again, the Tribunal is not persuaded by the Commissioner’s submissions. The Tribunal has not found clear and convincing evidence allowing it to conclude, on a balance of probabilities, that a few cents a bushel matter to farmers and to their behaviour in the market. On the contrary, the evidence from farmers is in respect of amounts much higher than 2 cents a bushel and higher than the predicted price increases of Dr. Miller.
[534] It is true, as the Commissioner argues, that Mr. Lincoln (a farmer called by the Commissioner) testified in his witness statement that that he needs “every penny to be able to hit the profitability levels” required to operate his farm (Lincoln Statement at para 16). However, immediately before he made that statement, Mr. Lincoln referred three times to an example of what a material amount meant to him: it was 10 cents a bushel. He indicated that the Moosomin Elevator would have to offer him 10 cents a bushel above the Fairlight Elevator for it to be worthwhile driving past this Elevator (Lincoln Statement at para 15). When he went on to quantify the adverse impact of a price decrease on his business, he again used the value of 10 cents per bushel (Lincoln Statement at para 16). Nothing in the Lincoln Statement supports the contention that an amount as low as 2 cents a bushel or in the range predicted by Dr. Miller would be material to him and to his competitive behaviour in the sale of his grain.
[535] The Commissioner also referred to the testimony of Mr. Hebert, one of P&H’s farmer witnesses, singling out his comments and podcasts about the farmers’ margins being squeezed. Further to its review of Mr. Hebert’s evidence, the Tribunal finds no support in that evidence for the proposition that 2 cents or a few cents a bushel matter to Mr. Hebert. The documents mentioned by Mr. Hebert in his testimony contained comments he made about his “5% rule” in the context of the farmers’ business: for Mr. Hebert, farmers should concentrate on trying to improve certain metrics of their business by 5%. The only references to cents per bushel were to an amount of 50 cents per bushel. Again, this is far above 2 cents or 7 cents per bushel.
[536] Mr. Paull, another farmer witness called by P&H, testified that he will only switch to a more distant Elevator if it means receiving more cents a bushel. Regarding the magnitude of the price differential, Mr. Paull testified that he would not leave a local Elevator for “a few pennies a bushel or a few cents a bushel,” but if he could make “10 or 15, 20 cents, sometimes 30 or 40 cents by going further,” he would (Consolidated Transcript, Confidential B, at p 1020). Mr. Paull added that he would not switch from a local Elevator if the difference in net price is 5 cents a bushel or lower (Consolidated Transcript, Confidential B, at p 1022). He also mentioned that he would not haul his grain 50 or 100 miles away for a few cents a bushel. More distant Elevators would need to offer higher prices than that for him to consider sending his grain farther. In fact, Mr. Paull sold all of his canola in the last four years to Bunge Altona, a Crusher located some 350 kilometers and three and a half hours away from his farm. He did it because it was profitable to do so.
[537] In sum, Mr. Paull testified that a few pennies are not material enough for him to change his selling behaviour and to switch Elevators. It starts to matter for him at 10 cents per bushel.
[538] As Mr. Duncan, Mr. Paull, and Mr. Hebert all said in their witness statements, the net price or Cash Price they receive from the Elevators is what matters to them as farmers.
[539] In his submissions, the Commissioner indicated that farmer witnesses were not meant to be a statistically representative sample and were not the way the Commissioner intended to demonstrate that the lessening of competition is substantial. He said that substantiality flows from Dr. Miller’s quantification work and other evidence on market shares and margins. However, even the farmers’ examples used by the Commissioner do not support his argument that 2 cents per bushel is a material price impact on the facts of this case.
[540] Dr. Miller also testified that the price effects he calculated would represent a loss of about CAD $2,000 per farm in the Elkhorn area, located between the Moosomin and Virden Elevators. However, the Commissioner did not point to any clear evidence allowing the Tribunal to determine how such an amount could be material or not to the farmers, or whether this would change their selling or competitive behaviour in the relevant markets for wheat or canola.
[541] The Tribunal makes one other observation in relation to the “2 cents per bushel” issue. P&H pointed out that volatility in the grain industry is often in excess of 10 cents a bushel on any given day. In her expert report, Ms. Sanderson calculated the average of the within-day price variations (Ms. Sanderson Report at para 104 and Figure 30). The average within-day variation is CAD $4.32 per MT for canola and CAD $3.82 per MT for wheat. When expressed in dollars per bushel, these values translate into approximately 10 cents a bushel for each of wheat and canola. Should a farm be successful in timing its grain sale within any given day, it can therefore achieve a purchase price that is 10 cents a bushel higher by selling grain at the right hour of the day. This evidence was not contradicted.
[542] In light of this evidence, said Ms. Sanderson, a 2 cents variation that is significantly lower than the typical within-day daily fluctuations in the purchase price of wheat or canola cannot be a material price variation in the grain industry. Furthermore, Ms. Sanderson stated that, during 2018-2019, the average cash purchase price paid at the Moosomin Elevator for canola was CAD $10.47 a bushel (or CAD $461.46 per MT), making 2 cents equal to 0.19% of the Cash Price of canola. During the same period, the average cash purchase price paid at the Moosomin Elevator for wheat was CAD $6.25 CAD a bushel (or CAD $229.73 per MT), making 2 cents equal to 0.32% of the Cash Price of wheat (7 cents would be equal to 1.12%).
[543] According to Ms. Sanderson, this evidence suggests that a material change in price cannot be less than 10 cents a bushel. The Tribunal agrees. The weighted average price increases predicted by Dr. Miller’s merger simulation model are only 1 cent per bushel in canola and 4 cents per bushel for wheat (in relation to the average price at the Moosomin, Virden, and Fairlight Elevators), which are both well below 10 cents per bushel.
[544] In sum, even looking at the absolute price effects measured by Dr. Miller, the Tribunal does not find evidence supporting the Commissioner’s position that the projected price variations are material. They are rather of a small magnitude and immaterial, consistent with the fact that P&H faces considerable competition from several rival Elevators and Crushers to constrain material price decreases after the Transaction.
[545] The Tribunal pauses to note the following. On this whole issue of P&H’s prices, the Tribunal agrees with the Commissioner that little weight should be given to Mr. Heimbecker’s evidence on this front, as he was obviously not very familiar with P&H’s day-to-day operations in relation to these pricing issues. However, even without taking into account Mr. Heimbecker’s evidence, there is no clear and convincing evidence of a material decrease in the price for wheat or canola, in absolute or relative terms.
[546] The Tribunal concludes that the evidence of absolute pricing variations does not constitute clear and convincing evidence supporting a conclusion that, “but for” the Virden Acquisition, the prices of wheat or canola paid to farmers by P&H are or would likely be “materially” lower.
(iii)
Difference-in-differences analysis
[547] As part of her expert report, Ms. Sanderson did a difference-in-differences regression analysis to verify whether, since the closing of the Transaction, P&H has effectively lowered the Cash Prices “it pays farms at Virden or Moosomin post-Acquisition in an economically significant way” (Ms. Sanderson Report at para 110). Ms. Sanderson’s difference-in-differences is a retrospective merger analysis based on what actually happened to P&H’s post-Acquisition posted prices since the merger.
[548] Difference-in-differences regressions can be informative to study the effect of events such as mergers if the facts are consistent with the empirical framework. These analyses are called difference-in-differences because they aim at determining whether the merger has created a difference going above and beyond the other changes that would have occurred in any event, and whether the relative difference will be the same through the review period.
[549] In this case, Ms. Sanderson used a difference-in-differences regression analysis of posted prices at the Moosomin and Virden Elevators to test if the Acquisition has effectively reduced P&H’s Cash Prices at Moosomin and Virden, relative to the prices prevailing at a benchmark P&H Elevator (namely, the Dutton Elevator) that is unaffected by the Acquisition. Ms. Sanderson considered that the Dutton Elevator was an appropriate comparator Elevator as it was unaffected by the Acquisition and was outside of any acquired LDC Elevator’s draw areas, it was subject to the same network dynamics as the Moosomin Elevator, and it had posted prices. The graphs provided in Ms. Sanderson’s expert report show that the Dutton Elevator had closely similar behaviour and common trends compared to the Virden and Moosomin Elevators, with no evidence of materially lower purchase prices than the Virden and Moosomin Elevators.
[550] Ms. Sanderson calculated a difference-in-differences regression comparing net price changes for the Moosomin and Virden Elevators using the post-Acquisition months from December 10, 2019 to June 30, 2020 and the same months for 2016 to 2019 in the pre-Acquisition period. By using similar months every year, Ms. Sanderson removed added seasonal variations that increase modeling noise, and she chose the Dutton Elevator outside the Transaction’s competitive markets to control for impact of market conditions external to the impact of the Transaction.
[551] Ms. Sanderson concluded to a minimal decline in the Cash Price of canola and to a posted Cash Price increase of 1.5% for wheat since the Acquisition (Ms. Sanderson Report at para 112 and Figure 33; Ms. Sanderson Slides at p 86). More specifically, for the Moosomin Elevator, her difference-in-differences analysis found that canola prices were 0.5% higher and that wheat prices were 0.6% lower. For the Virden Elevator, she found that canola prices were 0.6% lower and wheat prices 1.5% higher. On that basis, she concluded that her regression analysis constituted further evidence that, as far as price effects are concerned, the Virden Acquisition does not, and is not likely to, lessen competition substantially.
[552] Dr. Miller acknowledged that Ms. Sanderson’s difference-in-differences methodology was “about as good as you can do” and “as good as the economic literature can get you here” (Consolidated Transcript, Confidential A, at p 1889). However, Dr. Miller considered that Ms. Sanderson’s regression framework does not fit the facts of this case, and is not a good predictor of P&H’s pricing in the future.
[553] Three concerns were expressed by Dr. Miller about Ms. Sanderson’s difference-in-differences analysis. First, Dr. Miller’s main issue was about the usefulness of doing such an analysis while a merger review process is still on-going. In other words, he questioned the probative value of a retrospective merger analysis while the merger itself is still under review before the Tribunal, and had serious reservations about it. Second, he questioned the use of posted prices as opposed to realized transactions which could include results of negotiations (bearing in mind that, according to the evidence, XXx of P&H’s prices for wheat and canola are negotiated). These negotiations often impact the Basis, and not only the Cash Price. Third, he was concerned about the reliability of the data in the period surrounding the Transaction, which is influenced by an international trade war and the effects of the COVID-19 pandemic. Dr. Miller also submitted that Ms. Sanderson does not have enough data post-merger to do a robust analysis, nor does she have a sufficient control set against which to reliably interpret her results, as she only had one benchmark Elevator, the Dutton Elevator, in her control group.
[554] In her presentation at the hearing, Ms. Sanderson took into account the concerns raised by Dr. Miller (Ms. Sanderson Slides at p 87). She responded that she used the Dutton Elevator to counter the impact of factors such as an international trade war or COVID-19. She is comparing the changes at the Moosomin and Virden Elevators to the changes at the Dutton Elevator. All three Elevators were impacted by these exogenous factors, and the assumption was that the broader market effects would be similar on the three Elevators.
[555] With respect to the quality of her price data, Ms. Sanderson emphasized that her regression is a difference-in-differences analysis, and that using the posted price data would be suitable for picking up whether there is a difference due to the Acquisition. Approximately XXx of P&H’s transactions are at posted prices, and there was no evidence that the situation about negotiated prices is different at the Moosomin and Virden Elevators from what it is at other Elevators, including the Dutton Elevator. More generally, Ms. Sanderson advised that there is no indication that the Dutton Elevator was an outlier or had delivered results different from other Elevators in the P&H network. Ms. Sanderson’s model results also indicate that the R-squared variable, which is an indicator measuring how well the variables in the regression predict the dependent variable, is fairly high, thus showing a good fit for her model.
[556] The Tribunal notes that Dr. Miller identified a few reasons as to why Ms. Sanderson’s methodology could provide inadequate results, but no evidence was provided to show that this occurred with Ms. Sanderson’s model in this case.
[557] However, the Tribunal agrees with some of Dr. Miller’s concerns about Ms. Sanderson’s difference-in-differences regression, in that it did not account for autocorrelation in the modeling, which leads to overstating the precision of her results. The Tribunal also notes that the post-Acquisition time period for the data used by Ms. Sanderson, being less than a year, is fairly short and that there were inventory issues at the Virden Elevator towards the end of 2019.
[558] One member of the Tribunal, Ms. Samrout, has additional methodological concerns with Ms. Sanderson’s analysis. Ms. Samrout is of the view that Ms. Sanderson’s regression treated the Futures Price as a control variable that explains a portion of the variation in the observed Cash Price. Although this variable provides the average change in the Cash Price pre- and post-Transaction as the Futures Price changes, while holding all other variables constant, it is unclear whether the model fully or partially explains the magnitude of the variation in the Cash Price since the interaction between the timing of the transaction (pre- and post-Transaction) and the Futures Price was not tested. In other words, a variable should have been included to capture the difference in the magnitude of how well the Cash Price is affected by the Futures Price pre- and post-Transaction.
[559] As part of her difference-in-differences analysis, Ms. Sanderson also presented in her testimony (Ms. Sanderson Slides at pp 82–85) the line trends tracking the difference in Cash Price before and after the Acquisition in comparison with the Dutton Elevator’s Cash Price, showing that there is little difference between the Moosomin and Virden pricing behaviour in Cash Price in comparison to Dutton. Ms. Samrout notes that Ms. Sanderson compared the Cash Price before and after the Acquisition without any visual linkage to the fluctuation and trending of Futures Prices and their corresponding posted Basis to arrive at the Cash Prices.
[560] In the end, the Tribunal is satisfied that Dr. Miller’s critiques do not significantly alter Ms. Sanderson’s conclusions and agrees that, on a balance of probabilities, it is more likely than not that P&H has not lowered its purchase prices for wheat and canola since the Acquisition. In sum, the Tribunal is of the view that Ms. Sanderson’s difference-in-differences analysis is more consistent with P&H’s position than the Commissioner’s, and that it provides some support for the non-material nature of the price effects resulting or likely to result from the Virden Acquisition. However, the Tribunal finds the analysis to be of more limited assistance given the concerns raised by the limitation of the data and by the fact that it is looking at the behaviour of P&H while the Virden Acquisition was under review. The Tribunal thus concludes that it should be given little weight in its price effects analysis.
(iv)
Conclusion on price effects
[561] In light of the foregoing, the Tribunal is left with unpersuasive and insufficient evidence regarding the alleged substantiality of the price effects of the Virden Acquisition. The measured price effects are de minimis for the purchase of canola. For wheat, the Tribunal finds that the Acquisition is having or is likely to have some very minor price effects on the purchase of wheat but they are far from substantial. In sum, having regard to the evidence presented, the Tribunal finds that the likely price variations due to the Virden Acquisition, whether in absolute terms or in relative terms, are immaterial and are likely to remain immaterial for both the purchase of wheat and the purchase of canola.
[562] The Tribunal underlines that, even though farmers located in the “corridor of concern” are arguably the most affected by the Virden Acquisition as a vast majority of them sell their wheat and canola exclusively to the Virden, Moosomin, and Fairlight Elevators, the Commissioner presented no quantitative evidence regarding the predicted price changes in that specific part of the relevant geographic markets for wheat and canola. There is therefore insufficient evidence allowing the Tribunal to determine the relative or absolute magnitude of the predicted price variations for the farmers located in the corridor of concern.
(c)
Market concentration and market shares
[563] The Tribunal now turns to the effect of the Virden Acquisition on P&H’s post-merger market shares. As discussed above, Dr. Miller and Ms. Sanderson disagree as to the number of participants in the relevant geographic markets; hence, they disagree on whether or not the merged firm will have a market share that exceeds 35%, which is considered a “safe harbour” in the Commissioner’s 2011 MEGs.
[564] Relying on Tele-Direct at paragraph 226, the Commissioner submits that high market shares are an indirect indicator that the Virden Acquisition allows P&H to exercise increased market power: in this case, Dr. Miller’s calculations indicate that P&H’s post-Acquisition market shares are 59.6% for wheat and 53.9% for canola (Dr. Miller Report at Exhibit 10). These calculations are for Dr. Miller’s original relevant geographic market based on the “imputed” price for GHS, which was defined by Dr. Miller as including solely the Virden, Moosomin, and Fairlight Elevators.
[565] According to Ms. Sanderson, using the geographic markets based on the Cash Prices with a 5% SSNIP threshold, P&H’s post-Acquisition market shares would be much lower (Ms. Sanderson Slides at pp 74–75). It would be 31.6% in the relevant geographic market for the purchase of wheat, defined by Ms. Sanderson as including the Moosomin, Virden, Fairlight, Oakner, Whitewood, Elva, Shoal Lake, and Carnduff Elevators, using the average price at the Virden Elevator as the reference price. Based on Ms. Sanderson’s figures, this percentage would increase to 34.6% if the Carnduff Elevator is left out of the market. Post-Acquisition market shares would amount to only 16.1% in the relevant geographic market for the purchase of canola, defined by Ms. Sanderson as including the Moosomin, Virden, Fairlight, Oakner, Whitewood, Brandon (Richardson), Melville, Souris East, Shoal Lake, and Elva Elevators as well as the Crushers at Harrowby (Bunge), Yorkton (LDC), Velva (ADM), and Yorkton (Richardson).
[566] Ms. Sanderson further opined that what matters for the competitive effects analysis is P&H’s post-merger share of purchases, because this determines the competitive alternatives available to farms if P&H were to seek to reduce its purchase prices for wheat and canola post-Acquisition. Ms. Sanderson testified that, post-Acquisition, the Moosomin and Virden Elevators account for only 15% of total canola purchases and only 26% of total wheat purchases from farms within Dr. Miller’s Farmer Region.
[567] While subsection 92(2) of the Act expressly precludes the Tribunal from making a finding of substantial lessening of competition “solely on the basis of evidence of concentration or market share,” it is not disputed that evidence of changes in market shares and concentration levels are relevant and often influential in the Tribunal’s assessment. The Tribunal further agrees with the Commissioner that, in its market share and concentration calculations, it should not take into account the purchases of all Elevators located both inside and outside the relevant markets; the market shares should instead be computed based on the purchases made by those participants that are part of the relevant markets, as these markets are defined by the Tribunal.
[568] In this case, based on the evidence before it and considering the geographic market definition discussed above, the Tribunal finds that the post-Acquisition market share of P&H would be at most 16.1% in the relevant geographic market for the purchase of canola. This is a conservative measure, as the relevant geographic market for canola identified by Ms. Sanderson was based on SSNIP level well below 5% — it was 3.08% — and on the lower Virden Elevator price for canola as a reference price. This concentration level is significantly below the 35% safe harbour threshold identified in the MEGs, and cannot be considered indicative of a lessening of competition in the purchase of canola, let alone a substantial one.
[569] Turning to wheat, the Tribunal concludes that the post-Acquisition market share of P&H would likely fall within a range corresponding to the potential relevant geographic markets discussed above: these market shares would vary between 31.6% for a relevant geographic market including all Elevators retained by Ms. Sanderson in her analysis and approximately 34.6% if the market is limited to the Moosomin, Virden, Fairlight, Oakner, Whitewood, Elva, and Shoal Lake Elevators (and excludes the Carnduff Elevator). These market share figures approaching 35% provide, in the Tribunal’s view, no more than weak evidence of enhanced market power in the purchase of wheat.
(d)
Removal of a vigorous and effective competitor
[570] The Commissioner submits that the Virden Acquisition eliminates intense rivalry between the two main suppliers of GHS for wheat and canola in the corridor of concern. He asserts that the Moosomin and Virden Elevators competed head-to-head on price and service, and were each other’s closest competitors. The removal of a vigorous and effective competitor is one of the factors specifically contemplated by section 93, at paragraph (f).
[571] The Tribunal agrees with the Commissioner on this point.
[572] Three sources of evidence support the Commissioner’s arguments: the diversion ratios, the documentary evidence, and the farmers’ testimonies.
[573] One of the factors that can be relevant when considering the likely competitive effects of a merger is the diversion ratios between the products of the merging parties. This is because such ratios can provide information regarding the closeness of competition between those products. In this case, the diversion ratios calculated by Dr. Miller demonstrate that the Virden and Moosomin Elevators are close competitors. For wheat, the diversion ratios found by Dr. Miller were 23.8% from the Moosomin Elevator to the Virden Elevator, and 36.3% to the Fairlight Elevator. Conversely, for the Virden Elevator, the diversion ratios for wheat were 16.8% to the Moosomin Elevator and 20.3% to the Fairlight Elevator. True, the highest diversion ratios from each of the two Elevators were to the Fairlight Elevator. But the diversion ratios between the Moosomin and Virden Elevators were second, reflecting the fact that the Moosomin and Virden Elevators are relatively close competitors. As will be discussed below, the Tribunal is mindful of the fact that these diversion ratios — being around 20% and far below 50% — are not objectively high, and that there is significant diversion to rival Elevators other than Moosomin, Virden, or Fairlight. However, this does not diminish the fact that, in light of these observed diversion ratios, the Acquisition removes a vigorous and effective competitor.
[574] There is also ample documentary evidence of direct, head-to-head competition between the Moosomin and Virden Elevators.
[575] Finally, the farmer witnesses who testified on behalf of the Commissioner — Mr. Lincoln, Mr. Pethick, and Mr. Wagstaff — all referred to the fact that they were using the two Elevators and complained about the loss of one competitive option further to the Virden Acquisition.
[576] Based on all this evidence, the Tribunal finds that the Acquisition eliminates a vigorous competitor —the Virden Elevator — which was a close rival to the Moosomin Elevator. The Tribunal is satisfied that, in the absence of the Acquisition, the Virden and Moosomin Elevators would have continued to vigorously compete in the relevant markets for the foreseeable future. There is no evidence to indicate that the Virden Elevator would have ceased being a vigorous and effective competitor, but for the Acquisition.
(e)
Effective remaining competitors
[577] While the removal of a vigorous and effective competitor is a factor that the Tribunal may have regard to, paragraph 93(e) of the Act also directs the Tribunal to consider the “extent to which effective remaining competition remains or would remain” in a market affected by the merger. P&H maintains that there are several remaining Elevators and Crushers which will continue to discipline the Virden and Moosomin Elevators after the Acquisition, in the purchase of both wheat and canola, and that these rivals will be capable of constraining P&H’s ability to exercise increased market power after the Acquisition.
[578] The Tribunal agrees in part with P&H. The fact that the Moosomin and Virden Elevators went toe-to-toe on several occasions does not mean that they did not also have to go toe-to-toe with other rival Elevators located farther away. At the stage of the substantial lessening of competition, the Tribunal considers whether rivals’ purchase locations (including those close to the border of the relevant geographic market) provide competition and constrain the supply/purchase locations that are within the geographic market (VAA CT at para 305; Hillsdown at p 330).
[579] Dr. Miller and Ms. Sanderson agree that, for the purpose of analyzing anti-competitive effects and the substantial lessening of competition, the Tribunal needs to consider competitors located both inside and outside the defined relevant geographic markets in order to assess their constraining impact on the price and non-price dimensions of competition. In other words, the competitive constraints on a merged firm can come from rivals within and outside the relevant geographic markets (Consolidated Transcript, Public, at p 1831). Indeed, the number of competitors considered by Dr. Miller for his substantial lessening of competition analysis were not limited to those Elevators he had included in his narrow geographic market definition; they instead extended to 15 Elevators for wheat and, for canola, to these same 15 Elevators as well as five Crushers. In sum, Dr. Miller’s merger simulation model included Elevators and Crushers that come from outside the defined geographic markets.
[580] In her testimony before the Tribunal, Ms. Sanderson stated that, however the geographic market is defined, what matters for anti-competitive effects is P&H’s post-merger share of purchases because this determines the competitive alternatives available to farms if P&H were to seek to reduce its purchase prices post-Acquisition. Dr. Miller and Ms. Sanderson did not dispute P&H’s share of purchases. As mentioned above, post-Acquisition, the Moosomin and Virden Elevators account only for 15% of canola purchases and 26% of wheat purchases from farms within Dr. Miller’s Farmer Region.
[581] P&H’s evidence indicates that it competes with numerous Elevators and Crushers in the purchase of wheat and canola in Dr. Miller’s Farmer Region. According to P&H, its competitors for wheat include the following Elevators and grain companies: Viterra at Fairlight, Brandon, Souris East, Grenfell, Waldron, Binscarth, and Carnduff; Paterson at Binscarth and Carnduff; Richardson at Shoal Lake, Kemnay, Langenburg, Melville, Minnedosa, Estevan, and Whitewood; Ceres at Northgate; Cargill at Oakner, Nesbitt, and Elva; and G3 at Bloom and Melville (Heimbecker Statement at paras 119, 121–124). P&H adds that its competitors for canola include the following Elevators and Crushers, and grain companies: Viterra at Fairlight, Brandon, Souris East, Grenfell, Waldron, Binscarth, and Carnduff; Paterson at Binscarth and Carnduff; Richardson at Shoal Lake, Kemnay, Langenburg, Yorkton, Melville, Minnedosa, Estevan, and Whitewood; Ceres at Northgate; Cargill at Oakner, Nesbitt ,and Elva; G3 at Bloom and Melville; LDC at Yorkton; ADM at Velva; and Bunge at Harrowby and Altona (Heimbecker Statement at paras 120–125).
[582] According to Ms. Sanderson, there are many other Elevators and Crushers buying canola and wheat from farms in Dr. Miller’s Farmer Region. As indicated in Ms. Sanderson’s expert report, 11 rival Elevators and Crushers are buying more canola than does the Moosomin Elevator (Ms. Sanderson Report at Figure 24; Ms. Sanderson Slides at pp 77–78). And Viterra’s Fairlight Elevator and Bunge’s Harrowby Crusher also purchase more canola than the Virden Elevator. For wheat, while the Virden, Moosomin, and Fairlight Elevators are the largest purchasers of wheat, six rival Elevators each have at least a 5% share of total wheat purchases (Ms. Sanderson Report at Figure 25; Ms. Sanderson Slides at pp 79–80).
[583] P&H further submits that internal business documents from both its records and LDC’s records further show that each of the Moosomin and Virden Elevators compete with and track prices of many other Elevators beyond Moosomin, Virden, and Fairlight (Ms. Sanderson Report at paras 81–82). In her evidence, Ms. Sanderson referred to numerous contemporaneous business documents from P&H and LDC where they referred to competitor pricing, such as Viterra at Carnduff, or Richardson at Whitewood.
[584] Ms. Sanderson’s analysis of P&H’s documents and LDC’s Producer Reports showed that individual price negotiations are relatively infrequent, echoing the evidence demonstrating that about XXx of P&H’s transactions are at posted prices. In her analysis of those instances where price negotiations actually occurred, Ms. Sanderson found that 72 of the 213 reports referring to negotiations at the Virden Elevator identified farms using rival Elevator prices in a negotiation: only six of those 72 mentioned the Moosomin Elevator, and 27 mentioned Viterra’s Fairlight Elevator. Other rival grain companies identified in the reports included G3 (13 times), Cargill (11 times), and Richardson (eight times). Conversely, 17 reports referring to negotiations at the Moosomin Elevator identified farms using rival Elevator prices in a negotiation: five of those mentioned Virden, and the Fairlight Elevator appeared 12 times (Ms. Sanderson Report at paras 116–117, Figures 35a–35b; Ms. Sanderson Slides at p 88). This evidence, says P&H, reflects the presence of other remaining Elevators and Crushers constraining P&H’s pricing behaviour.
[585] Turning to the diversion ratios found by Dr. Miller, Ms. Sanderson advises that, for wheat, they show that 40% of the sales diverted from the Moosomin Elevator would go to rival Elevators other than Virden or Fairlight. For the Virden Elevator, 63% of the diverted sales of wheat from the Virden Elevator would go to Elevators other than Moosomin or Fairlight. In the case of canola, these percentages are higher: 65% of the sales diverted from the Moosomin Elevator would go to rival Elevators and Crushers other than the Virden or Fairlight Elevators, and 77% of the sales diverted from the Virden Elevator would be absorbed by Elevators and Crushers other than the Moosomin or Fairlight Elevators (Dr. Miller Report at Exhibit 11; Ms. Sanderson Report at Figure 50). The Tribunal agrees with Ms. Sanderson that these diversion ratios to rival Elevators other than Moosomin, Virden, or Fairlight are not insignificant, with several other Elevators and Crushers having diversion ratios for canola that are similar or higher than those calculated between the Moosomin and Virden Elevators (Ms. Sanderson Report at Figure 47). For wheat, diversion ratios from one of the Moosomin or Virden Elevators exceed 10% to the Fairlight (Viterra), Whitewood (Richardson), Souris East (Viterra), and Oakner (Cargill) Elevators (Ms. Sanderson Report at Figure 48).
[586] Distance (and hence trucking costs) between individual farms and Elevators was included in Dr. Miller’s farm choice model, which found that the Virden and Moosomin Elevators would lose a significant portion of their wheat or canola sales to rival Elevators other than Moosomin, Virden, and Fairlight.
[587] According to Ms. Sanderson, the relatively small price variations predicted by Dr. Miller’s merger simulation model reflect the relatively low diversion ratios between the Moosomin and Virden Elevators, compared to the diversion ratios to rival Elevators and Crushers (Ms. Sanderson Slides at p 5). P&H submits that the diversion ratios support a finding that P&H also faces competition from several rival Elevators and Crushers.
[588] The evidence from the farmer witnesses also suggests that farmers have numerous alternatives to which they sell their grain, and to which they could sell more, should P&H attempt to reduce its purchase prices for wheat or canola (Ms. Sanderson Report at paras 12–13). Any farm located within the draw areas of the Moosomin and Virden Elevators has at least six other Elevators or Crushers available to it, should it not want to sell its grain to P&H, if the price offered price at those other locations was attractive. According to P&H, the evidence establishes that all farmers in the Commissioner’s relevant geographic market and in the Commissioner’s corridor of concern within that geographic market, including each of the farmer witnesses, consistently sell to numerous Elevators and Crushers beyond the Virden, Moosomin, and Fairlight Elevators. P&H also argued that the evidence also establishes that the farmers could easily switch to other Elevators and Crushers without negative financial impact on them in the event that P&H were to attempt to pay them less for their wheat or canola (P&H Read-In at pp 26–27, 52–54, 61–67, 77, 551, 553–556, 558).
[589] The three farmer witnesses called by the Commissioner (i.e., Mr. Lincoln, Mr. Wagstaff, and Mr. Pethick) have sold their grain to Elevators located farther away when offered an acceptable price from more distant Elevators. Farmer witnesses further confirmed in their testimony that while transportation costs are important to farmers in deciding between competing Elevators and/or Crushers, they do not preclude sales of grain to competing alternatives available to farmers at a greater distance. This is so because rival Elevators and Crushers offer posted Cash Prices that are high enough to cover farmers’ hauling costs and make it worthwhile for them to sell to those Elevators and Crushers (Exhibit P-R-14, Witness Statement of Mr. Edward Paull, at para 25; Exhibit P-R-077, Witness Statement of Mr. Kristjan Hebert, at para 26; Exhibit P-R-095, Witness Statement of Mr. Timothy Duncan, at para 21).
[590] In light of the foregoing, the Tribunal agrees that there are some effective remaining competitors able to constrain P&H’s attempt to increase its market power further by the Virden Acquisition. The more difficult question is to identify how many there is and who they are. The Tribunal is not persuaded that all rival Elevators and Crushers singled out by P&H remain, and are likely to remain, effective competitors post-Acquisition. All competitors are not equal, even if they purchase some grain from the farms located in Dr. Miller’s Farmer Region.
[591] In the Tribunal’s view, the effective remaining competitors certainly include those Elevators and Crushers forming part of the relevant geographic markets accepted by the Tribunal. For wheat, these competitors are the following Elevators (and grain companies): Fairlight (Viterra), Oakner (Cargill), Whitewood (Richardson), Elva (Cargill), and Shoal Lake (Richardson). For canola, this effective remaining competition includes several Elevators — namely, Fairlight (Viterra), Oakner (Cargill), Whitewood (Richardson), Brandon (Richardson), Melville (G3), Souris East (Viterra), Shoal Lake (Richardson), and Elva (Cargill) — as well as four Crushers — namely, Harrowby (Bunge), Yorkton (LDC), Velva (ADM), and Yorkton (Richardson). The Tribunal observes that these competing Elevators and Crushers are all operated by major grain companies that purchase wheat and canola in competition with P&H throughout Western Canada, including Viterra and Richardson, the two largest. They also include most of the Elevators and Crushers having the higher diversion ratios discussed above, as well as all the rivals which buy more canola than the Virden or Moosomin Elevators and four out of the six Elevators having at least a 5% share of total purchases of wheat.
[592] One notable exception is the Viterra Elevator in Souris East, which is not part of the relevant geographic market for wheat despite its high diversion ratio from the Virden Elevator (i.e., 12.9%) and its 5.3% share of total wheat purchases. The Tribunal is of the view that, in light of this evidence, this rival Elevator can be considered as an effective remaining competitor in wheat even though it is not part of the relevant geographic market.
[593] With respect to other competing Elevators and Crushers located outside the relevant geographic markets identified in these Reasons, the Tribunal accepts that, from time to time, they are purchasing grain from farms located in Dr. Miller’s Farmer Region. However, the Tribunal does not find that, on a balance of probabilities, they can be qualified as “effective” remaining competitors able to constrain P&H post-Acquisition.
[594] In sum, the Tribunal agrees with P&H that there is and will be several remaining effective competitors in the relevant markets for the purchase of wheat and canola. On a balance of probabilities, there are competitors remaining that are able to discipline P&H’s attempts to exercise market power.
(f)
Fairlight Elevator
[595] The Commissioner claims that, while the Fairlight Elevator will remain as an effective competitor, it is not sufficient to constrain the exercise of market power by P&H post-Acquisition.
[596] According to the Commissioner, the simple way to demonstrate this is through Dr. Miller’s quantitative evidence. There is also qualitative evidence to support this quantitative evidence: the Virden and Moosomin Elevators are both on the Trans-Canada Highway, while the Fairlight Elevator is located approximately 35 kilometers away down a secondary road. This secondary road is subject to weight restrictions in the spring. Mr. Wagstaff testified that, contrary to the Trans-Canada Highway where one can drive at an average speed, going up highway 41 or down road no. 8 to get to the Fairlight Elevator is more difficult, as the road conditions are worse, and it takes longer to haul the grain.
[597] The Tribunal accepts that, in and of itself, the presence of the Fairlight Elevator may not be enough to constrain the exercise of increased market power by P&H post-Acquisition. However, the Tribunal nonetheless observes that the Fairlight Elevator is a very important competitive constraint that will remain after the merger given its market shares and the evidence of diversion ratios. In the geographic market for wheat, the Fairlight Elevator would have the highest market share after the Moosomin-Virden combination. In the geographic market for canola, it would have the second highest market share after the Crusher in Harrowby and Moosomin-Virden. Moreover, the Fairlight Elevator’s diversion ratios (at 20.3% and 36.3%) are respectively higher than the Moosomin Elevator (16.8%) and Virden Elevator (23.8%) for diversion of wheat from the Virden Elevator and the Moosomin Elevator. It is also true for canola: the Fairlight Elevator’s diversion ratios (at 17.2% and 22.2%) are respectively higher than those of the Moosomin Elevator (5.3%) and the Virden Elevator (13.1%) for diversion of wheat from the Virden Elevator and the Moosomin Elevator.
[598] For those reasons, the Tribunal is of the view that the Fairlight Elevator is a significant competitor that, on the evidence, cannot be qualified as being less effective than other remaining Elevators and Crushers.
(g)
Moosomin expansion
[599] The Commissioner contends that P&H’s decision not to expand the Moosomin Elevator’s rail capacity is another reflection of the anti-competitive effects caused by the Virden Acquisition. Prior to the Acquisition, P&H had planned to expand railcar access at the Moosomin Elevator, allowing it to load up 112 railcars at once, instead of its current limit of 56. This expansion would have allowed the Moosomin Elevator to access lower bulk rates on its freight, and to be more competitive.
[600] Considering the evidence provided by Mr. Heimbecker on the reasons for the delayed expansion at the Moosomin Elevator, the Tribunal is not persuaded by the Commissioner’s arguments. In sum, there is no clear and convincing evidence supporting the Commissioner’s assertion that the delay in expansion is or could likely be the result of the Virden Acquisition.
[601] The evidence before the Tribunal shows that in July 2019, P&H made the decision not to expand rail capacity at the Moosomin Elevator as initially planned before LDC solicited P&H to buy its 10 Elevators, including the Virden Elevator. Mr. Heimbecker further testified that later in 2019, P&H decided to postpone all capital expenditures, including the Moosomin Elevator expansion, for a period of one year, as a matter of prudent financial management, in light of the fact that P&H would be spending more than XXXXXXXXXX to purchase the LDC Elevators (Heimbecker Statement at para 138, Exhibit 34). Mr. Heimbecker further confirmed in his testimony that, subject to the outcome of this Application (as a result of which P&H could potentially be ordered to divest the Moosomin Elevator), the P&H Board is expected to approve the 112-car spot expansion at the Moosomin Elevator.
[602] The Tribunal accepts Mr. Heimbecker’s evidence on that point, which has not been contradicted. The Tribunal does not dispute that if the Moosomin Elevator’s rail capacity had been expanded, the Moosomin Elevator would have been a more vigorous competitor. But the evidence before the Tribunal does not allow it to conclude that the delayed Moosomin expansion can be attributed to the Virden Acquisition.
[603] This is therefore not a factor supporting the Commissioner’s claim of a substantial lessening of competition caused by the Virden Acquisition.
(h)
Barriers to entry and expansion
[604] In assessing whether competition is or is likely to be substantially lessened by a merger, an important factor to consider is whether entry or expansion into the relevant market likely is or likely would be substantially faster, more frequent, or more significant “but for” the merger (TREB CT at para 505). This factor is specifically mentioned at paragraph 93(d) of the Act. As previously noted by the Tribunal, “[t]he conditions of entry into a relevant market can be a decisive factor in the Tribunal’s assessment of whether a merger is likely to prevent or lessen competition substantially” (Tervita CT at para 216).
[605] In assessing whether new entry into, or expansion within, a relevant market can be relied upon to conclude that a substantial lessening of competition is likely to occur, the Tribunal will consider whether such entry or expansion will be timely, likely, and sufficient (Tervita CT at para 217). For a new entry to be timely, the assessment is not limited to the actual construction time. The period starts when firms begin considering sites and go through the regulatory approval process.
[606] According to the Commissioner, entry or expansion by competitors into the relevant markets is unlikely to occur in a timely and sufficient way because barriers to entry and expansion are high in the grain industry. They include capital costs of CAD $40-50 million to construct an Elevator, as well as having to find a location where there is sufficient demand and a suitable site permitting adequate rail and road access. A potential entrant would further take more than two years to build an Elevator, says the Commissioner.
[607] P&H responds that, in terms of entry, there is evidence that at least 20 new high throughput Elevators have been built in Western Canada since 2015, including 10 Elevators by G3, two Elevators by P&H, one by Ceres, and four by GrainsConnect. P&H adds that G3 also has three additional Elevators currently under construction.
[608] For the reasons detailed below, the Tribunal finds that the Commissioner has provided credible and persuasive evidence confirming that barriers to entry and expansion are high. This evidence came from G3 and from P&H itself.
[609] Mr. Malkoske of G3 testified that, based on his experience with G3’s Vermillion site, the total costs to build a new Elevator are approximately XXXXXXXXXXXXXXXXXXXXXXX (Exhibits P-A-047 and CB-A-048, Witness Statement of Mr. Brett Malkoske, at para 8). He also stated that it typically takes between XXXXXX months from deciding to construct an Elevator to commencing operations.
[610] The evidence of G3 was corroborated by P&H’s own evidence.
[611] When building a new Elevator in Dugald, Manitoba, P&H considered sites in June 2018 and, going through the permit process, regulatory approvals, and construction, it expected the Elevator to be completed in XXXXXX, approximately XXXXXX months after its initial consideration. P&H further estimated the cost of building a new Elevator to be in the range of XXXXXXXXXXX to XXXXXXXXXXX.
[612] The Tribunal notes P&H’s evidence on its experience that a motivated competitor wishing to construct greenfield Elevators could build a new Elevator in the Virden/Moosomin area in approximately 18 months. However, the Tribunal is not convinced by the general statement of Mr. Heimbecker on this point, in light of the other specific evidence on the record, referring to actual experiences by both G3 and P&H itself in building new Elevators over a much longer timeframe.
[613] As a result, the Tribunal is satisfied that there is clear and convincing evidence to support the conclusion that entry or expansion is not likely to occur on a sufficient scale or scope within the next two years and that new entrants are not sufficient to have a material impact on the price and non-price dimensions of competition in the purchase of wheat and canola. Having regard to the foregoing evidence, the Tribunal finds that there are significant barriers to entry into the purchase of wheat and canola. Entry and expansion within the relevant markets is not likely to be sufficient to ensure that the Virden Acquisition does not and will not likely lessen competition substantially, and to prevent P&H from imposing and sustaining decreased prices for its purchase of wheat and canola.
(i)
Excess capacity
[614] At the Elevator level, the general issue of entry and expansion entails considering the potential of adding capacity through new entry, the possibility of expanding existing capacity, and the existence of excess capacity that already lies dormant in the industry.
[615] In his submissions, the Commissioner addresses the first two elements — namely, entry and expansion — but not the last one, excess capacity.
[616] P&H claims that the presence of significant excess capacity is and will remain a constraining factor on P&H after the Virden Acquisition. In paragraphs 141 to 147 and 152 of his witness statement, Mr. Heimbecker discussed the maximum observed throughput and actual throughput at P&H and other Elevators in the grain industry. Using publicly available CGC data (Heimbecker Statement at Exhibit 35), Mr. Heimbecker prepared a table setting out the average amount of grain purchased and shipped annually by each of P&H’s rival Elevators over a five-year period between the 2014-2015 and 2018-2019 crop years. The table also provided information on each of the Elevators’ maximum annual throughput in that five-year period and their average and best turn rates (Heimbecker Statement at paras 143–145).
[617] Summing their individual maximum annual throughputs, Mr. Heimbecker testified that the aggregate maximum “capacity” of competing Elevators (i.e., Elevators other than Moosomin and Virden) is at least 4,562,400 MT. In comparison, the five-year average of total throughput of these Elevators was 3,570,700 MT (Heimbecker Statement at para 146). A comparison of these two figures led Mr. Heimbecker to conclude that these rival Elevators are capable of handling at least 991,700 MT more than their average throughput over the five-year period. In his witness statement, Mr. Heimbecker also addressed the rail shipping capacity, at paragraphs 148 to 151, and concluded that the unused car spot capacity added up to 4,765,200 MT.
[618] In a preliminary motion objecting to some paragraphs of the Heimbecker Statement, the Commissioner submitted that Mr. Heimbecker cannot opine as to what rival Elevators can do and took exception with Mr. Heimbecker’s statements that rival Elevators “could easily increase their purchases of wheat and canola from farms in the Virden/Moosomin area” or “add significant grain purchasing capacity,” made at paragraphs 141 and 152 of the Heimbecker Statement.
[619] In a decision issued in December 2020, the Tribunal agreed with P&H that those paragraphs on average throughput and capacity generally described Mr. Heimbecker’s observations and perceptions from data published by the CGC showing the volume of grain that P&H and rival Elevators purchased and shipped (i.e., their effective and maximum throughput) in the five-year period between the 2014-2015 and 2018-2019 crop years (Parrish & Heimbecker at paras 25–30). The Tribunal was satisfied that the statements on the actual measures of capacity and throughput generally reflected Mr. Heimbecker’s own observation of data and amounted to simple arithmetical calculations required to establish averages, totals, and differences regarding maximum throughput at Elevators within a five year span. Given his long experience in the grain industry, the Tribunal noted that Mr. Heimbecker was well positioned to assist the Tribunal in this regard. However, the Tribunal held that Mr. Heimbecker was not qualified to form conclusions as to what the rival Elevators would do with their alleged excess capacity. Extrapolating from the throughput data to what rival Elevators could do in their businesses and to their future conduct in terms of purchases of wheat and canola are inferences or conclusions that could be done by experts or argued by counsel, and which will ultimately be for the Tribunal to determine. The Tribunal therefore struck certain passages of the Heimbecker Statement as inadmissible lay opinion evidence.
[620] Mr. Heimbecker’s evidence on throughput and excess capacity calculations was not contradicted. The Tribunal, however, observes that no evidence has been provided regarding what other Elevators not owned by P&H could or would do with their excess throughput capacity or railcar capacity, and how it could or would translate in terms of their purchases of wheat and canola.
[621] In their testimonies, Mr. McQueen from Viterra and Mr. Wildeman from Ceres confirmed the nature of the data and information reported to the CGC regarding Elevators’ throughput, which was used by Mr. Heimbecker for his calculations.
[622] The Commissioner did not adduce (or obtain) any evidence as to excess capacity nor did he challenge the evidence put forward by P&H in this regard.
[623] The Tribunal is not satisfied that the arithmetic calculation presented by Mr. Heimbecker can qualify as a true measurement of excess capacity. Mr. Heimbecker’s calculations provide no reference to grain volumes produced during each crop year corresponding to the grain purchased and shipped for that year, as well as to the grain companies’ ability to ship and sell those volumes. Nor do they provide a clear understanding of the supply chain capacity and fluidity impacting Elevators’ throughput with respect to ships availability, and the capacities of export terminals, Elevators, ports, and rail networks.
[624] In previous decisions, the Tribunal has considered excess capacity in its substantial lessening of competition analysis and when considering market power at the market definition stage (see, for example, Hillsdown, at pp 318–321). In Hillsdown, the Tribunal noted that if rival firms in a market have excess capacity, they “can respond to a supra-competitive price rise by flooding the market at a lower price level” (Hillsdown at p 318). As a result, “the best question to ask when assessing market power, in some circumstances, is whether the respondents’ [merging parties’] current competitors have capacity available to service what otherwise would be the merged firm’s customers” (Hillsdown at p 318).
[625] The Tribunal in Hillsdown considered evidence about the excess capacity of several specific rivals and the merged firm’s own excess capacity. At page 320, the Tribunal concluded that in the industry, it was fairly easy for renderers to increase their capacity or, in the case of multi-plant firms, to shift renderable material among different plants to open up capacity at a given plant when it was needed. The Tribunal also found that there appeared to be significant excess capacity in the industry generally, and that the merged firm was not capacity-constrained. The excess capacity of firms both inside and outside the relevant market would provide a degree of competitive pressure on the merged firm and restrain to a considerable degree its ability to raise prices (Hillsdown at p 321).
[626] In the present case, the Tribunal accepts that, at an industry level, the Elevators are likely able to move more grain as suggested by Mr. Heimbecker’s evidence and to transport it by rail to ports. However, the Tribunal cannot place much weight on this general finding in its substantial lessening of competition assessment. An assessment of excess capacity in the grain industry would have had to consider many additional factors and evidence that is not before the Tribunal. For instance, in the Tribunal’s view, there is insufficient evidence about the excess capacity of specific rival Elevators close to the Moosomin and Virden Elevators — namely, how the owners of those rival Elevators could adjust their local grain inventories and rail capacities at different points in the crop year to respond to lower grain purchase prices offered to farmers at those two P&H Elevators. In the absence of additional factual evidence (e.g., from rival Elevator owners) or expert evidence about the industry that assessed capacity issues in more detail, the Tribunal can reach no firm conclusion on the impact of any excess throughput capacity on the intensity of rivalry after the Transaction, or on the ability of P&H to decrease prices for wheat and canola.
(j)
Other factors
[627] Paragraph 93(h) of the Act also allows the Tribunal to take any other relevant factor in consideration in its assessment of the substantial lessening of competition. The Tribunal is aware that, following the most recent amendments to the Act entered into force after the hearing of this Application, the Act specifically recognizes, at paragraph 93(g.3), that the Tribunal may have regard to any effect of a merger on “price and non-price competition, including quality, choice or consumer privacy.”
[628] In this case, the non-price effects alleged by the Commissioner are very limited. More specifically, there is no issue of change and innovation being affected by the Virden Acquisition (see paragraph 93(g)).
[629] In his final submissions, the Commissioner argued that the evidence also supports the existence of non-price competition. He referred to evidence from Mr. Pethick stating that the ability of an Elevator to accept grain during harvest is important to him (Pethick Statement at para 13). He also mentioned that the ability of an Elevator to quickly and efficiently receive grain, and its capacity to do so, is an aspect of non-price competition that the Tribunal should consider. The Commissioner however added that, while these aspects of competition are not necessarily reflected in the Elevator’s posted price, they “ultimately all are reflected in the price paid to the farmers by the Elevator” (Commissioner’s Closing Submissions at para 44).
[630] The non-price effects alleged by the Commissioner thus appear to be reflected in the transaction data and in the Cash Prices ultimately charged to the farmers by P&H and other Elevators for their wheat and canola. They are part of the price effects on which the Commissioner has focused in his submissions and in the evidence he presented.
[631] The Tribunal therefore determines that there is no clear and convincing evidence of separate non-price factors demonstrating, in and of themselves, a lessening of competition, let alone a substantial one.
(k)
Magnitude, duration, and scope of the anti-competitive effects
[632] Having regard to all of the foregoing, the Tribunal concludes that, “but for” the Virden Acquisition, there are and will likely be some fairly limited adverse effects on competition in the purchase of wheat in the relevant markets, and virtually no proven effects in the purchase of canola. More specifically, there is not, and would likely not be, any new entry or expansion; there is, or likely would be, some very minor price decreases in the amount paid to farmers for their wheat; the Acquisition removes a close competitor to P&H’s Moosomin Elevator, as the Virden Elevator provided vigorous and effective competition; and there is also evidence of some pre-existing market power held by P&H. However, those anti-competitive effects are far less than what the Commissioner alleged.
[633] The Tribunal finds that the price effects, whether they are measured in relative or absolute terms, are minimal and immaterial in the circumstances, for both wheat and canola. Moreover, there will be several remaining effective rival Elevators and Crushers to compete with the Virden and Moosomin Elevators, and this situation is unlikely to change for the foreseeable future. In addition, the post-merger market shares of P&H would be far below the safe harbour threshold in the 2011 MEGs for purchases of canola; for wheat, the market shares are more likely than not to be slightly lower than the 35% safe harbour. Finally, the postponement of the Moosomin expansion cannot be attributed to the Virden Acquisition, and there appears to be some varying capacity at rival Elevators though the evidence does not allow the Tribunal to assess its constraining impact on P&H.
[634] The Tribunal must now determine whether the limited anti-competitive effects attributable to the Virden Acquisition and identified above, taken together, rise to the level of substantiality required by section 92. Further to its assessment of all the evidence before it, and notably the immaterial price effects resulting from the Virden Acquisition, the Tribunal finds that this is not the case. In particular, the Tribunal puts significant weight on the evidence showing an absence of any material price effects resulting from the Virden Acquisition, in the purchase of both wheat and canola, and the continuing presence of several effective remaining competitors in the markets. In brief, the aggregate impact of the limited anti-competitive effects that have been demonstrated to result from the Virden Acquisition does not constitute an actual or likely substantial lessening of competition in the relevant markets.
[635] Stated differently, the Tribunal is not persuaded that the evidence regarding the absence of likelihood of additional entry and expansion, the minimal predicted price variations, and the loss of the Virden Elevator as a competitor are sufficient, cumulatively, to enable the Commissioner to discharge his burden under section 92, even in a context where P&H has some pre-existing market power. Without a link between, on the one hand, such evidence and, on the other hand, some material impact on the price or non-price dimensions of competition in a material part of the relevant markets (Tervita FCA at para 108), the Commissioner’s evidence falls short of the mark. In this regard, when measured against factors such as the immaterial price variations, the effective competitors remaining, and the post-merger market shares below the 35% safe harbour threshold, the Tribunal agrees with P&H that the Commissioner’s evidence does not provide clear and compelling evidence that there is, or would likely be, materially lower price or non-price competition “but for” the Acquisition.
[636] Regarding scope, the Tribunal typically considers whether the merged entity would likely have the ability to impose the anti-competitive effects in a material part of the relevant market, or in respect of a material volume of sales. The evidence relied on by the Commissioner does not establish that it does.
[637] The Tribunal looked more specifically at the evidence regarding farmers located in the corridor of concern, arguably the most affected by the Virden Acquisition as a vast majority of them sell their wheat and canola exclusively to the Virden, Moosomin, and Fairlight Elevators. However, as mentioned above, there was an absence of quantitative and qualitative evidence regarding the predicted price changes in that specific part of the relevant geographic markets for the purchase of wheat and canola. Similarly, the Commissioner has not presented compelling evidence regarding the particular effects that the removal of the Virden Elevator as a vigorous and effective competitor would have in the corridor of concern. There is therefore insufficient quantitative and qualitative evidence to allow the Tribunal to determine the degree to which the price and non-price effects resulting from the Virden Acquisition would be different for the farmers located in the corridor of concern.
(4)
Conclusion on substantial lessening of competition
[638] In light of all of the foregoing, the Tribunal is not satisfied that the anti-competitive effects that could be attributable to the Virden Acquisition are, individually or in the aggregate, “substantial” as required by section 92 of the Act. The evidence does not allow the Tribunal to conclude that the Virden Acquisition has adversely affected or is adversely affecting price or non-price competition in the relevant markets, to a degree that is material, or that it is likely to do so in the future. True, the Virden Acquisition allows and is likely to allow P&H to increase its market power in the purchase of wheat; but the evidence does not support a finding that, on a balance of probabilities, it allows P&H to be able to exercise materially greater market power than in the absence of the merger. The Tribunal therefore concludes that the Commissioner has not demonstrated, on a balance of probabilities and with clear and convincing evidence, that the requirements of section 92 are met and that “but for” the Virden Acquisition, purchase prices paid by P&H for wheat and canola are or would likely be materially lower in the relevant markets, or that there are or would likely be materially greater non-price competition in those markets.
D.
If the Commissioner has established that the Virden Acquisition lessens, or is likely to lessen, competition substantially, what is the remedy to be ordered?
[639] Given the Tribunal’s conclusion that the Virden Acquisition does not lessen, and is not likely to lessen, competition substantially in any relevant market, there is no remedy to be ordered.
E.
Has P&H established, on a balance of probabilities, that the gains in efficiency will be greater than, and will likely offset, the effects of any lessening of competition pursuant to section 96 of the Act?
[640] In light of the Tribunal’s findings on the substantial lessening of competition, and since there will be no remedial order under section 92 of the Act, the Tribunal does not need to determine the issue of efficiencies claimed by P&H under section 96. However, considering the extensive submissions made by the parties on the issue of efficiencies and the nature of the issues raised, the Tribunal will address the matter.
(1)
Analytical framework
[641] The section 96 efficiencies defence is an exception to the application of section 92. It “prohibits the Tribunal from making an order precluding a merger when it finds that the merger is likely to bring about gains in efficiency that would be greater than and would offset the anti-competitive effects of the merger” (Tervita SCC at para 17).
(a)
The statutory language
[642] In the current case, P&H relies on both the general language of section 96 and the more specific provisions of subsection 96(2) to support its efficiency claims. Each will be discussed in turn.
(i)
Section 96
[643]
Section 96 of the Act is reproduced in Schedule “A” to these Reasons. Subsection 96(1) provides in relevant part that the Tribunal shall not make an order under section 92 if it finds that the merger in respect of which the application is made has “brought about or is likely to bring about gains in efficiency that will be greater than, and will offset, the effects of any […] lessening of competition that will result or is likely to result from the merger […] and that the gains in efficiency would not likely be attained if the order were made.” Subsection 96(3) further instructs the Tribunal not to find that a merger or proposed merger has brought about or is likely to bring about gains in efficiency “by reason only of a redistribution of income between two or more persons.”
[644]
The analysis mandated by subsection 96(1) has three components: 1) an assessment of the anti-competitive effects of any lessening or prevention of competition resulting or likely to result from the merger; 2) an assessment of the gains in efficiency brought about or likely to be brought about by the merger, and that would not likely be attained if the contemplated remedial order is made by the Tribunal; and 3) a trade-off analysis (or balancing test) to determine whether the assessed gains in efficiency will be greater than, and will offset, the assessed anti-competitive effects. The analyses of anti-competitive effects and e
fficiencies analyses are both forward-looking estimations and are therefore associated with varying degrees of uncertainty.
[645] In Tervita SCC and Tervita CT, the SCC and the Tribunal provided important guidance on the three elements of section 96.
·
Anti-competitive effects
[646] For the purpose of the efficiencies defence and section 96, the anti-competitive effects include all effects of “any” lessening of competition, as long they result or are likely to result from the merger in respect of which the application is made.
[647] Anti-competitive effects include the likelihood of price increases, but they are not confined to such resource allocation effects, as the exercise of market power can manifest itself in ways other than an increase in price. Under both sections 92 and 96 of the Act, the anti-competitive effects encompass all relevant price and non-price effects that are likely to arise from a merger, including the following: negative effects on allocative, productive, and dynamic efficiency; redistributive effects; and effects on service, quality, and product choice (2011 MEGs at paras 12.21–12.22).
[648] The main anti-competitive effect considered under section 96 is the DWL associated with a likely increase in price. The DWL represents the loss of allocative efficiency and reflects the reduction in total consumer and producer surplus within Canada resulting from a merger. As the SCC stated in Tervita SCC, the “total surplus standard involves quantifying the DWL which will result from a merger — ‘the amount by which total surplus is reduced under certain market conditions that reduce the quantity of a good that is supplied’ (Facey and Brown at pp 256–57)” (Tervita SCC at para 94). The DWL “results from the fall in demand for the merged entities’ products following a post-merger increase in price, and the inefficient allocation of resources that occurs when, as prices rise, consumers purchase a less suitable substitute” (Canada (Commissioner of Competition) v Superior Propane Inc, 2003 FCA 53 (“Superior Propane IV”) at para 13). Estimates of the elasticity of demand (i.e., the degree to which demand for a product varies with its price) are necessary to calculate the DWL (Tervita CT at para 244). Put differently, a price increase or decrease is not enough to determine the extent of the DWL, if there is no evidence on the price elasticity of demand (Tervita SCC at paras 132–133; Tervita FCA at para 124).
[649] The focus of this DWL approach is on the magnitude of the total surplus. The degree to which total surplus is allocated between producers and consumers is not considered. In other words, the total surplus standard measures the total benefit flowing to the economy and is not concerned with whom the benefits flow to (Tervita SCC at para 95; Superior Propane IV at para 16).
[650] A lessening or prevention of competition resulting from a merger can also lead to non-price effects in the form of a reduction in service, quality, product choice, incentives to innovate, or other dimensions of competition that customers value. While some indicators of quality may be translatable into dollar terms by making use of available statistical or survey data, others may not be expressible in that way. As such, it is not always possible to translate anti-competitive effects related to non-price factors into consumer or producer welfare terms, as can be more easily done with price effects.
[651] Anti-competitive effects covered by both sections 92 and 96 are therefore not strictly limited to reductions in price and to the quantified or quantifiable DWL. They may also include unquantifiable non-price effects such as: reduction in service, quality, and product choice; loss of productive efficiency; and loss of dynamic efficiency (2011 MEGs at paras 12.29–12.31). If a non-price anti-competitive effect is not reasonably measurable, it may be assessed using qualitative evidence. The SCC recognized it in Tervita SCC, when it noted that “qualitative elements of a merger, including in some cases such things as better or worse service or lower or higher quality, may not be measurable as they are dependent on individual preferences in the market” (Tervita SCC at para 100, citing Superior Propane I at paras 459–460).
[652] This qualitative assessment of anti-competitive effects may only be resorted to where such effects are not quantifiable: “[q]ualitative anti-competitive effects, including lessening of service or quality reduction, are only assessed on a subjective basis because this analysis involves a weighing of considerations that cannot be quantified because they have no common unit of measure (that is, they are “incommensurable”)” (Tervita SCC at para 125). However, quantifiable effects which the Commissioner failed to quantify cannot be resurrected as qualitative effects (Tervita SCC at para 100). Anti-competitive effects should be quantified wherever reasonably possible, and the weight given to unquantifiable qualitative effects must be reasonable (Tervita FCA at para 148).
[653] In some circumstances, the anti-competitive effects to be assessed under section 96 may further include redistributive effects, namely, wealth transfers from buyers to sellers which amount to a social loss. These are commonly referred to as a “socially adverse wealth transfer.”
[654] It may happen that the Tribunal finds a substantial lessening of competition under section 92, but no anti-competitive effects under section 96, as was the case in the Tervita matter, a situation characterized by Justice Rothstein as a “paradoxical” result (Tervita SCC at para 166). Depending on the evidence available, the statutory scheme under sections 92 and 96 does not bar a finding of likely substantial lessening or prevention of competition where there has been a failure to quantify the DWL. But the test under section 96 does require that quantifiable anti-competitive effects be quantified in order to be considered under the efficiencies defence. In Tervita SCC, the SCC determined that the failure to quantify the DWL barred consideration, under section 96, of the quantifiable effects that supported a finding of likely substantial prevention of competition under section 92.
·
Cognizable efficiencies
[655]
With respect to efficiencies, section 96 provides that, to be recognized and be qualified as “cognizable,” efficiencies must meet three different requirements. First, the gains in efficiency must be brought, or be likely to be brought, by “the merger or proposed merger in respect of which the application is made” (subsection 96(1)). Second, the gains in efficiency would not likely be attained if the order contemplated by the Tribunal to remedy the substantial lessening or prevention of competition were made. Third, the gains in efficiency must not result “by reason only of a redistribution of income between two or more persons” (subsection 96(3)).
[656]
As is the case for anti-competitive effects, cognizable efficiencies include both efficiencies that can reasonably be quantified as well as those that cannot reasonably be quantified and are therefore qualitative. Since it must be “likely” that the claimed gains in efficiency are achieved because of the merger, there must be
evidence of the claimed savings and of the implementation process leading to the materialization of the claimed efficiencies (Superior Propane I at paras 347–348). As stated in the 2011 MEGs, “the parties must be able to validate efficiency claims to allow the Bureau to ascertain the nature, magnitude, likelihood and timeliness of the asserted gains, and to credit (or not) the basis on which the claims are being made” (2011 MEGs at para 12.3).
[657] In Tervita CT, the Tribunal adopted five screens to standardize the methodology used to eliminate efficiencies that are not cognizable under subsections 96(1) and 96(3) (Tervita CT at paras 261–264). Those screens may be summarized as follows:
1. The claimed gains in efficiency must involve a type of productive or dynamic efficiency or be likely to result in an increase in allocative efficiency (Tervita SCC at para 102; Tervita CT at para 262).
2. The claimed gains in efficiency must likely be brought about by the merger (Tervita SCC at para 113; Tervita CT at para 262).
The claimed gains in efficiency must not be brought about by reason only of a redistribution of income, or amount to a simple wealth transfer between organizations in Canada. This screen serves to discard savings that result solely from a reduction in output, service, quality or product choice, reductions in taxes, and savings from increased bargaining leverage (Tervita CT at para 262).
The claimed gains in efficiency must not be achieved outside Canada and must instead flow back to Canadian shareholders. Under this fourth screen, savings from operations in Canada that would flow through to foreign shareholders are eliminated (Tervita CT at para 262).
The claimed gains in efficiency must not (a) be attainable through alternative means even if the Tribunal were to make an order to eliminate the substantial prevention or lessening of competition, or (b) be achievable through the merger even if the remedial order were made (Tervita CT at para 264). In sum, the gains in efficiency are evaluated in light of the order contemplated by the Tribunal (Commissioner of Competition v Superior Propane Inc, 2002 Comp Trib 16 (“Superior Propane III”) at para 149), and gains that would have occurred irrespective of the merger are not cognizable (Hillsdown at p 83).
[658] Implicit in these five screens is that, in order to be cognizable under section 96, the claimed gains in efficiency must accrue to the benefit of society — that is, to the Canadian economy (Tervita SCC at para 102; Tervita CT at para 262; Superior Propane III at para 196; Superior Propane I at paras 412–413, 429–430).
·
Trade-off analysis or balancing test
[659] Turning to the trade-off analysis, the SCC held that the requirement in section 96 according to which the efficiency gains must be “greater than” and “offset” the anti-competitive effects imports a weighing and balancing of both quantitative and qualitative aspects of the merger. The term “greater than” suggests a numerical comparison of the magnitude of the efficiencies versus the extent of the anti-competitive effects. The term “offset” implies a more subjective analysis related to qualitative considerations — i.e., things that cannot be quantitatively compared because they have no common measure (Tervita SCC at paras 144–145; Superior Propane II at para 95).
[660] In Tervita SCC, the SCC also directed the Tribunal to use an analysis that is as objective as possible. As such, the SCC held that in most cases, the qualitative effects of a merger will be of “lesser importance.” In addition, the SCC stated that “the statutory requirement that efficiencies be greater than and offset the anti-competitive effects would in most cases require a showing that the quantitative efficiencies exceed the quantitative anti-competitive effects as a necessary element of the defence” [emphasis in original] (Tervita SCC at para 146).
[661] Tervita SCC adopted a two-step inquiry for the balancing test in subsection 96(1). First, the quantitative efficiencies of the merger at issue are compared against the quantitative anti-competitive effects to determine whether the quantitative efficiencies are “greater than” the quantitative anti-competitive effects. If the quantitative anti-competitive effects outweigh the quantitative efficiencies, this step will be dispositive in most cases, and the section 96 defence will not apply. Second, the qualitative efficiencies are balanced against the qualitative anti-competitive effects. The Tribunal then makes a final determination as to whether the total efficiencies “offset” the total anti-competitive effects of the merger at issue (Tervita SCC at para 147).
[662] The SCC further held that the Tribunal should consider “all available quantitative and qualitative evidence” (Tervita SCC at para 100, citing Superior Propane I at para 461; Superior Propane III at para 335). For the Tribunal to give weight to qualitative elements in its analysis, they must be supported by the evidence, and the reasoning for the reliance on the qualitative aspects must be clearly articulated (Tervita SCC at para 147).
(ii)
Interpretation of subsection 96(2)
[663]
In this matter, P&H also made specific submissions on the interpretation of subsection 96(2) and relied heavily on this provision in its consideration of the efficiency gains allegedly brought about by the Transaction. This s
ubsection provides in part that in considering whether a merger or proposed merger is likely to bring about gains in efficiency described in subsection 96(1), the Tribunal shall consider whether such gains will result in “(a) a significant increase in the real value of exports.”
[664]
P&H argued that the purpose and effect of the Transaction were to increase export sales by P&H and that increased throughput at export terminals was the “uncontroverted objective” of the Transaction. P&H further submitted that increased throughput at an export facility is a “more efficient utilization of an asset in the purest sense,” and that the Transaction as a whole clearly leads to a significant increase in the real value of exports of grain under subsection 96(2).
[665]
P&H maintained that there were three possible interpretations of subsection 96(2):
· The consideration of efficiencies related to exports is subsumed under the subsection 96(1) analysis, so that all of the tests applied under subsection 96(1) would be applied to exports;
· The consideration of exports under subsection 96(2) is a separate consideration that is not confined to the efficiencies arising in the local or domestic markets. Any efficiencies arising either domestically or in an export market that lead to a significant increase in exports must be separately considered. In this interpretation, the efficiency gains that lead to significant increases in the real value of exports would not be subsumed in the subsection 96(1) analysis including the offset test and the counterfactual but-for test; and
· Subsection 96(2) only deals with efficiency gains that lead to increases in the real value of exports, and exports per se should be considered under the broader analysis under sections 92 and 93 of the Act.
[666] P&H took the position that the second and third interpretations described above are the correct ones to follow.
[667] P&H submitted that the interpretation of subsection 96(2) must accord with Parliamentary intent, as seen in comments made in 1986 by the Parliamentary Secretary to the Minister of Agriculture, the Honourable Pierre Blais, during the House of Commons debates on Bill C-91 (which enacted the Competition Tribunal Act, RSC 1985, c 19 (2nd Supp) (“CTA”) and the Act). P&H argued that mergers should be assessed, in Mr. Blais’ words, “in such a way as to encourage competition between Canadian businesses at home in Canada, without putting them at a disadvantage when carrying out business dealings in international markets.” P&H also referred to the objectives in section 1.1 of the Act, including its references to the expansion of opportunities for Canadian participation in world markets and the promotion of the efficiency of the Canadian economy. In addition, P&H submitted (again citing Mr. Blais) that “when a merger would greatly improve efficiency, thereby increasing exports or substitutions to imports, the Tribunal will have to authorize it.” P&H argued that this reflected a focus on efficiencies that lead to increased exports.
[668] The Commissioner made no written submissions on subsection 96(2). The Commissioner noted in oral argument that P&H made no reference to any reliance on that provision in its Response filed on February 3, 2020. The Commissioner submitted that P&H had acquired the grain volume at the Virden Elevator which LDC would have exported. In other words, these grain volumes were headed to the port at Vancouver for export, regardless of the Transaction. Therefore, P&H had not demonstrated any substantial increase in the real value of exports under paragraph 96(2)(a). The Commissioner also submitted that P&H had not analyzed why keeping the Virden Elevator in its possession (i.e., if the Tribunal makes no order) would lead to a substantial increase in the real value of exports.
[669] This is the first time that the Tribunal has been directly asked to interpret or apply paragraph 96(2)(a) of the Act.
[670] Under modern principles of statutory interpretation, the words of paragraph 96(2)(a) must be read in their entire context and in their grammatical and ordinary sense, harmoniously with the scheme and the object of the Act, and the intention of Parliament (Pioneer Corp v Godfrey, 2019 SCC 42 at para 42; Canada Trustco Mortgage Co v Canada, 2005 SCC 54 (“Canada Trustco”) at para 10; Bell ExpressVu at para 26; Rizzo & Rizzo Shoes Ltd (Re), [1998] 1 SCR 27 at para 21; VAA CT at para 257; The Commissioner of Competition v HarperCollins Publishers LLC and HarperCollins Canada Limited, 2017 Comp Trib 10 at paras 101, 118; Rakuten Kobo Inc v The Commissioner of Competition, 2016 Comp Trib 11 at para 108). In Canada Trustco, the SCC also held that if the words of a provision are precise and unequivocal, their ordinary meaning plays a dominant role in the interpretive process. If the words can support more than one reasonable meaning, the ordinary meaning of the words plays a lesser role. However, in all cases, the Tribunal must seek to read the provisions of the Act as a harmonious whole (Canada Trustco at para 10; 9354-9186 Québec inc v Callidus Capital Corp, 2020 SCC 10 at para 60).
[671] When the legislation that would become the Act was introduced in December 1985, the accompanying guide made the following observations:
The existing merger provision is considered to be unsuitable for the Canadian economy, which is small and open. Canadian firms often have to compete with larger foreign rivals both at home and abroad. In these circumstances, they should not be prevented from obtaining economies of scale which improve their competitive position. An effective merger law for Canada must weight [sic] the advantages of economic efficiency against the disadvantages of a lessening of competition.
[…]
The new merger law will also provide a defence in situations where the gains in efficiency that would result from the merger would more than offset the costs due to the lessening of competition. It is important for the performance of the economy that significant cost savings brought about by mergers, for example, through scale economies or other efficiencies, be allowed. Moreover, the Tribunal will be invited to consider whether the gains in efficiency resulted in increased exports or increased import substitution.
Minister of Consumer and Corporate Affairs, Competition Law Amendments: A Guide, (1985) at pp 16–17.
[672] In Tervita SCC, Justice Rothstein, speaking for a majority of the court, observed the following at paras 87 and 167:
[87] A stand-alone statutory efficiencies defence was considered “particularly appropriate for Canada because a small domestic market often precludes more than a few firms from operating at efficient levels of production and because Canadian firms need to be able to exploit scale economies to remain competitive internationally” (Campbell, [citation below] at p. 152; see also House of Commons Debates, vol. VIII, 1st Sess., 33rd Parl., April 7, 1986, at p. 11962; Minister of Consumer and Corporate Affairs, Competition Law Amendments: A Guide (1985), at p. 4). In the context of the relatively small Canadian economy, to which international trade is important, the efficiencies defence is Parliamentary recognition that, in some cases, consolidation is more beneficial than competition (ibid., at pp. 15-17).
[…]
[167] While the efficiencies defence applies in this case under the terms of s. 96 as written, this case does not appear to me to reflect the policy considerations that Parliament likely had in mind in creating an exception to the general ban on anti-competitive mergers. As discussed above at para. 84 [sic: 87] in the historical examination of s. 96, the evidence suggests that the efficiencies defence was created in recognition of the size of Canada’s domestic market and with an eye toward supporting operation at efficient levels of production and the realization of economies of scale, particularly with reference to international competition. […]
[Emphasis added.]
[673] In section 1.1 of the Act, Parliament set out the overall purpose of the statute: “to maintain and encourage competition in Canada.” Section 1.1 contains four objectives or benefits of maintaining and encouraging competition in Canada (Superior Propane I at paras 407–408, 410), which are listed in the provision:
| Purpose of Act
|
|
| 1.1 The purpose of this Act is to maintain and encourage competition in Canada in order to promote the efficiency and adaptability of the Canadian economy, in order to expand opportunities for Canadian participation in world markets while at the same time recognizing the role of foreign competition in Canada, in order to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy and in order to provide consumers with competitive prices and product choices. |
1.1 La présente loi a pour objet de préserver et de favoriser la concurrence au Canada dans le but de stimuler l’adaptabilité et l’efficience de l’économie canadienne, d’améliorer les chances de participation canadienne aux marchés mondiaux tout en tenant simultanément compte du rôle de la concurrence étrangère au Canada, d’assurer à la petite et à la moyenne entreprise une chance honnête de participer à l’économie canadienne, de même que dans le but d’assurer aux consommateurs des prix compétitifs et un choix dans les produits. |
[674] The FCA has held that section 96 as a whole gives primacy to the statutory objective of economic efficiency, because it provides that if efficiency gains exceed and offset the effects of an anti-competitive merger, the merger must be permitted to proceed even though it would otherwise be prohibited by section 92 (Superior Propane II at para 90).
[675] The commentary on the proper interpretation of the language in subsection 96(2) has not been extensive, despite its presence in the Act since 1986.
[676] The Competition Bureau’s 1991 MEGs stated that the words “described in subsection (1)”:
make it clear that section 96(2) does not operate to expand the class of efficiency gains that may be considered in the trade-off analysis. Accordingly, this provision is simply considered to draw attention to the fact that, in calculating the merged entity’s total output for the purpose of arriving at the sum of unit and other savings brought about by the merger, the output that will likely displace imports, and any increased output that is sold abroad, must be taken into account.
[677] The principal drafter of the 1991 MEGs, now a judicial member of the Tribunal, set out the same point in a 1991 article, adding that, “[a]pparently, there was never any intention on the part of the drafters of subsection 96(2) that this section has any significant role in the trade-off assessment” (Crampton 1991 at pp 967–968).
[678] A book by the same author further explained:
If an increase in the real value of exports […] cannot be shown to lead to one or more of the types of “gains in efficiency” [in subsection (1)], there is nothing that can be found in subsection 96(2) that will change this fact. In short, this provision simply requires the Tribunal to assess whether any significant increase in the real value of exports […] that is attributable to the attainment of efficiencies, will give rise to further efficiency gains.
Crampton 1990 at p 548 [Emphasis in original.]
[679] On this view, the increased exports are a source of efficiencies that must be taken into account under subsection 96(1) and Parliament enacted subsection 96(2) to fulfil a “guidance function” (Crampton 1990 at p 549).
[680] While this approach would not give efficiencies that meet the requirements of subsection 96(2) an independent role, the Crampton 1990 text offered additional analysis. It noted that there is a reasonable argument that Parliament intended subsection 96(2) to provide more than a guidance function, and that export-related efficiencies merit, and were intended to receive, additional qualitative weighting, that should be decisive only if the trade-off analysis under subsection 96(1) “yields an inclusive result.” The text listed four possible reasons. The significant increase in the real value of exports may: 1) give rise to other important but unquantifiable benefits to the economy; 2) promote the adaptability of the Canadian economy; 3) expand the opportunities for Canadian participation in world markets; and 4) have been considered to be of such particular importance as to merit specific mention in the legislation. In a footnote, the author referred to remarks by the then-Director of Investigation and Research (”Director”), who noted in a speech that information provided by merging parties about efficiency gains that allow the firm to significantly increase the real value of exports would be given “additional weight” in favour of the merging parties under subsection 96(1) (Crampton 1990 at pp 549–550).
[681] At least one commentary has taken a more robust view of subsection 96(2). While recognizing the positions taken in the 1991 MEGs and by the Director, Dr. Neil Campbell remarked that in enacting the provision, Parliament appeared to be creating an “independent category of efficiencies” (Campbell at pp 155–156, citing a discussion of “national champions” in A.N. Campbell and J.W. Rowley, Industrial Policy, Efficiencies and the Public Interest – the Prospects for International Merger Rules, (Center for Trade Policy and Law, 8th Annual Conference on Trade, Investment and Competition, Ottawa, May 1993)).
[682] The Competition Bureau’s 2004 MEGs and 2011 MEGs both advised merging parties to provide the Bureau with information that “establishes that the merger will lead them to increase output owing to greater exports,” noting in a footnote that “increased output in this context is generally only possible with an associated decrease in price” (see 2011 MEGs at para 12.12; 2004 MEGs at para 8.10).
[683] The text of subsection 96(2) is situated within section 96. The chapeau language of subsection 96(2) provides that in considering whether a merger or proposed merger is likely to bring about gains in efficiency described in subsection (1), the Tribunal “shall” consider whether “such” gains “will result in” a significant increase in the real value of exports and a significant substitution of domestic products for imported products
.
[684] With respect to the language in subsections 96(1) and (2), the Tribunal notes that:
· In Tervita SCC, Justice Rothstein observed that subsection 96(2) is “logically subservient to” subsection 96(1) (Tervita SCC at para 152);
· The analysis of any claimed efficiencies that meet the requirements of paragraphs 96(2)(a) and (b) occurs during the Tribunal’s consideration of whether the merger is likely to bring about the gains in efficiency described in subsection 96(1);
· The word “shall” is mandatory or directory, not permissive. The Tribunal is required to consider subsection 96(2) if the evidence supports its relevance in a proceeding;
· The words “such gains” [emphasis added] in subsection 96(2) refer to the gains in efficiency in subsection (1); and
· The words “will result in” suggest that the Tribunal consider the causal relationship between the gains in efficiency in subsection (1) and the existence of the two factors in paragraphs 96(2)(a) and (b). Parliament also chose to use “will result in” [emphasis added], which implies proof with greater certainty than, for example, “may result in.”
[685] Paragraph 96(2)(a) contemplates that the Tribunal consider the gains in efficiency in subsection (1) that will result in a significant increase in the “real value” of exports, which suggests that evidence must be adduced as to the quantum, in dollars, of the increase. To engage the provision, the increase must be shown to be “significant,” not small, trivial or inconsequential.
[686] In light of the foregoing, the Tribunal considers that the proper interpretation of subsection 96(2) should:
Adhere to the language of subsection 96(2), read in tandem with subsection 96(1);
Seek to implement the overall purpose and the objectives of the Act, as set out in section 1.1;
Be compatible with the approach to the trade-off analysis established in Tervita SCC, including the two-step quantitative and qualitative assessments described by Justice Rothstein at paragraph 147; and
Be compatible with the five screens established by the Tribunal in Tervita CT to implement the trade-off analysis in subsections 96(1) and (3).
[687] The Tribunal further observes that, by creating the efficiency exception enabling parties to avoid the application of section 92 when the elements of section 96 are satisfied, Parliament has in fact “withdrawn efficiency consideration, or at the very least, consideration relating to efficiency enhancement, from eligibility in the determination of whether competition has been prevented or substantially lessened” [emphasis in original] (Crampton 1990 at p 257). Gains in efficiency are thus excluded from the assessment of anti-competitive effects under section 92.
[688] In view of the language in subsection 96(2) and the discussion in the various legal sources above, the Tribunal adopts the following framework for the interpretation and application of subsection 96(2):
The Tribunal recognizes that a merger may make a firm more competitive in international markets through proven efficiencies realized in Canada (i.e., efficiencies that accrue to the Canadian economy).
Such proven efficiencies may be claimed and included in the Tribunal’s trade-off analysis, as established in Tervita SCC and Tervita CT.
Efficiencies that are claimed to result in a significant increase in the real value of exports should be analyzed within the established trade-off analysis under subsection 96(1).
The language of subsection 96(2) does not alter or expand the types of gains in efficiency under subsection 96(1) that may be considered in the trade-off analysis.
If a respondent raises a claim that efficiencies will result in a significant increase in the real value of exports, that respondent has the same burden to prove such efficiencies as with any other claimed efficiencies.
Specifically, if such claimed efficiencies are quantifiable, they must be quantified or, at a minimum, estimated, as contemplated by Tervita SCC.
If proven and quantified, such efficiencies will then be assessed along with other proven quantitative efficiencies in the first step of the trade-off analysis.
If such efficiencies are proven but cannot be quantified or estimated, they may be considered and given weight by the Tribunal at step two of the trade-off analysis, as qualitative efficiencies.
[689] Not resolved in this framework are the questions of whether and how the Tribunal should provide any additional recognition to efficiencies that result in a significant increase in the real value of exports under paragraph 96(2)(a).
[690] The Tribunal does not have the benefit of recent legal, economic, or business commentary on the role that subsection 96(2) should play in the efficiencies trade-off analysis established by Tervita SCC. The Tribunal also believes that, since the Act came into force in June 1986, there have been material developments affecting the Canadian industrial policy and international trade policies, as well as significant changes in some sectors of the Canadian economy, that could affect the proper approach to paragraphs 96(2)(a) and (b).
[691] Given the Tribunal’s conclusions elsewhere in these Reasons under section 92 and subsection 96(1), it is not necessary to resolve all questions related to paragraph 96(2)(a) in this proceeding. The Tribunal nonetheless considers that, without limiting other possible options, the following approach to paragraphs 96(2)(a) and (b) could be contemplated.
[692] As noted above, the Tribunal is of the view that the language of subsection 96(2) does not alter or expand the types of gains in efficiency under subsection 96(1) that may be considered in the trade-off analysis. However, in order to implement the objectives set out in section 1.1 of the Act and other indications of Parliamentary intent and objectives, the Tribunal could provide additional qualitative recognition to some claimed efficiencies in the trade-off analysis if a firm is able to demonstrate, with clear and cogent evidence, that the specified efficiencies achieved in Canada also meet the requirements of paragraphs 96(2)(a) or (b). Such proven efficiencies could achieve one or more of the objectives listed in section 1.1, in addition to the promotion of the efficiency of the Canadian economy inherent in any proven efficiencies under subsection 96(1).
[693] For example, a merger could be proven to result in efficiencies that accrue to the Canadian economy which will also result in a significant increase in the real value of exports and expand the opportunities for Canadian participation in world markets. A merger might, for example, enable a firm to enter a new overseas market due to the lower costs of production in Canada that result from the merger. In that type of scenario, the Tribunal could give the proven efficiencies that achieve the requirements of paragraph 96(2)(a) some additional, qualitative weight at stage two of the trade-off analysis under subsection 96(1). Specifically, at stage two, the Tribunal could give some qualitative weight to proven and quantified efficiencies that will result in a significant increase in the real value of exports; or the Tribunal could give additional qualitative weight at stage two to proven qualitative efficiencies. The weight could vary, for example, with the strength of the efficiencies’ proven ability to achieve at least one of the objectives of the Act as set out in section 1.1 (other than promoting the efficiency of the Canadian economy). Regardless of how the Tribunal decides to proceed in a future case, the Tribunal will determine how such efficiencies may be weighed at stage two based on the evidence and the applicable law, including Tervita SCC and Tervita CT.
(b)
Evidentiary burden on efficiencies
[694] As discussed above, the Commissioner has an initial burden under section 96 to prove the anti-competitive effects, including both quantitative and qualitative effects (Tervita SCC at para 122, citing Superior Propane I at paras 399, 403, Superior Propane II at para 154 and Superior Propane IV at para 64; Tervita CT at para 232).
[695] The SCC in Tervita SCC instructed that, to discharge his burden, the Commissioner must quantify any quantifiable anti-competitive effects, at least by estimate (Tervita SCC at paras 125, 126, 137, 165; Tervita FCA at para 127; Tervita CT at para 243). To meet that burden, the Commissioner must ground the calculations or estimates in evidence that can be challenged and weighed (Tervita SCC at para 125). If effects are realistically measurable, failure to at least estimate the quantification of those effects will not result in the effects being assessed on a qualitative basis (Tervita SCC at para 100, citing Superior Propane III at para 233 and Superior Propane IV at para 35). Instead, quantifiable effects that are not quantified are considered to be equal to zero and have no weight (Tervita SCC at paras 128–129, 137, 140, 142, 151, 157, 159, 165).
[696] While the Commissioner has the burden to prove the anti-competitive effects, the merging parties bear the onus of proving the remaining elements of the defence under section 96 (Tervita SCC at paras 136, 165; Superior Propane II at paras 154, 157; Tervita CT at para 233). The merging parties’ onus is to prove the extent of the efficiency gains and that those efficiency gains from the merger will be greater than and will offset the effects of any prevention or lessening of competition resulting from the merger (Tervita SCC at paras 89, 122; Superior Propane I at para 403).
[697] In the Tribunal’s view, the same requirements imposed on the Commissioner for proof of anti-competitive effects under section 96 should also be imposed on the merging parties (in this case, P&H) to discharge their onus to prove the remaining elements under section 96. Thus, if a claimed efficiency is quantifiable, it must be quantified or at least estimated. That quantification or estimate must be grounded in evidence that can be challenged and weighed. If the claimed efficiency is quantifiable but is not quantified or estimated, then it will be treated as a zero and given no weight. An unquantified claimed efficiency that could have been quantified, but was not, will not be considered as a qualitative efficiency (Tervita SCC at para 124; Superior Propane IV at para 35). Claimed qualitative efficiencies, if any, must also be supported by evidence that can be challenged and weighed.
(2)
Tribunal’s assessment
[698] In the Tribunal’s view, and for the following reasons, P&H has not met its burden of demonstrating, on a balance of probabilities, that the Virden Acquisition is likely to bring about cognizable gains in efficiency. As a result, such gains would not be greater than, and would not offset, the anti-competitive effects of any lessening of competition resulting from the Virden Acquisition.
(a)
The anti-competitive effects of any lessening of competition
[699] In light of the Tribunal’s conclusions on efficiencies, there is no need to deal extensively with the anti-competitive effects of any lessening of competition resulting from the Virden Acquisition.
[700] Suffice it to say that there is no fundamental dispute between Dr. Miller and Ms. Sanderson on the magnitude of the consumer surplus loss, suggested to be CAD $540,000 per crop year in wheat and less in canola according to Dr. Miller’s model. This estimate is conditional on accepting the diversion ratios and mark-up used by Dr. Miller.
[701] Ms. Sanderson only takes issue with Dr. Miller’s producer surplus calculations. She claims that it should include profits captured by Elevators and Crushers falling outside the relevant geographic markets. Even accepting Ms. Sanderson’s critique, when profits associated with all transfers to rival Elevators and Crushers are included, there is still some DWL in wheat, but not in canola. Taking into account the producer surplus transferred to rival Elevators and Crushers which do not belong to Dr. Miller’s geographic market, Dr. Miller’s DWL for wheat decreases to some CAD $30,000, while the DWL in respect of canola is eliminated.
[702] In sum, no matter how it is defined or calculated, the DWL is certainly greater than zero for wheat.
(b)
The gains in efficiencies
(i)
P&H’s claimed efficiencies
[703] At the outset of this proceeding, P&H’s Response claimed the following efficiencies from the Acquisition: improved FGT scale efficiencies and cost savings; elimination of the margin that LDC formerly paid to use the Vancouver export terminal owned by Kinder Morgan; output expansion and improved scale economies at the former LDC Elevators; and administrative efficiencies.
[704] P&H did not elaborate on its claimed efficiencies during the discovery process.
[705] Mr. Heimbecker’s initial witness statement identified four areas of claimed efficiencies that became the focus of the parties’ submissions at the hearing. They were as follows: 1) increased throughput at the Virden Elevator; 2) network logistics efficiencies arising from optimizing the shipment of grain to “freight-logical” terminals; 3) efficiencies at the Vancouver terminals, particularly the FGT; and 4) efficiencies related to the conversion of local Elevators to include the retail sale of crop inputs.
[706] At the hearing, the Commissioner did not cross-examine Mr. Heimbecker on his evidence about these proposed efficiencies. Mr. Heimbecker was not qualified to provide expert evidence on efficiencies and acknowledged that he is not such an expert. He provided P&H’s factual foundation for its efficiency claims. As discussed above, on a pre-hearing motion, some of his evidence was struck out (Parrish & Heimbecker at paras 72–73). Ultimately, P&H did not file an expert report to support or quantify its position on efficiencies, either initially or in reply to Mr. Harington’s expert report. P&H solely relied on the initial and reply witness statements of Mr. Heimbecker.
·
Increased throughput at the Virden Elevator
[707] Regarding the increased throughput at the Virden Elevator, P&H offered Mr. Heimbecker’s evidence comparing the throughput at the Virden Elevator in 2019 when it was operated by LDC with Virden’s 2020 throughput when operated by P&H. Mr. Heimbecker also explained that P&H’s Elevators had a higher “turn rate” than LDC’s former Elevators. An Elevator’s “turn rate” is calculated as the purchases of grain by an Elevator in a given period divided by the storage capacity. According to Mr. Heimbecker, P&H also forecast a higher turn rate for each of the former LDC Elevators during 2020. He attributed the higher turn rate to four factors: P&H’s superior port access (particularly its access to the Superior terminal in Thunder Bay, Ontario, which LDC did not use); port storage and ship loading speed in Vancouver; its larger network of Elevators from which it could source grain; and the fact that P&H purchased a larger variety of grain than LDC did.
[708] In his initial witness statement, to which Mr. Harington responded, Mr. Heimbecker compared the Virden Elevator’s 2019 results with its actual results from January to July 2020 and P&H’s forecast for the balance of 2020. He then applied P&H’s grain margins for wheat and canola to the increased volumes to find an annual “efficiency,” in dollars, for each of wheat and canola.
[709] In his reply witness statement, Mr. Heimbecker reiterated his evidence about P&H’s Elevator turn rates, its better port access at both Vancouver and Thunder Bay terminals, and its larger network of Elevators, all compared with LDC. He also updated his initial Virden throughput evidence with actual 2020 results up to October 31, 2020.
[710] Mr. Heimbecker also addressed the overall increase in grain production from 2019 to 2020 raised by Mr. Harington in his expert report. He concluded that, even adjusting for the increased production, the Virden Elevator showed increased throughput for wheat and canola when comparing the period from January 1 to October 31, 2019 with the period from January 1 to October 31, 2020. Using the same wheat and canola margins from his initial evidence, Mr. Heimbecker then re-determined the “additional value” created, in dollars, from the increased Virden throughput through to October 31, 2020, adjusting LDC’s 2019 numbers for grain industry developments.
·
Network logistics efficiencies
[711] Turning to network logistics efficiencies, Mr. Heimbecker explained in his initial witness statement that Elevators have a natural, “freight logical” terminal. Elevators operated by P&H fall into catchment areas depending on their physical location. Those catchment areas are for terminals in either Vancouver or Thunder Bay. The catchment area determines rail freight costs from an Elevator to a terminal. The Virden Elevator is in the Thunder Bay catchment area. As a result, the wheat and canola purchased from farms and delivered to the Virden Elevator will generally be shipped to P&H’s Superior terminal in Thunder Bay for sale to its eastern grain customers.
[712] Mr. Heimbecker testified that P&H decided that four of the Elevators it purchased in the Transaction could shift their grain from delivery to West Coast terminals in favour of delivery to the Superior terminal in the east. At the same time, some grain movement from Elevators already owned by P&H could be switched to delivery westward to Vancouver at P&H’s two most efficient terminals, namely, AGT and FGT. According to Mr. Heimbecker, the additional eastbound throughput from the addition of the four new Elevators “will now allow P&H to increase the efficiency of its network.” In particular, Mr. Heimbecker stated that P&H acquired the Virden and Rathwell Elevators in the Transaction and would shift their grain eastward, enabling additional tonnage to be shipped westward to Vancouver from other Elevators in the P&H network.
[713] Mr. Heimbecker advised in his reply witness statement that the addition of the Virden Elevator and its additional volume was a “necessary addition to P&H’s elevator network to achieve these logistics benefits” (Exhibits CA-R-121 and P-A-122, Reply Witness Statement of Mr. John Heimbecker (“Heimbecker Reply Statement”), at para 43).
·
Efficiencies at West Coast terminals
[714] Mr. Heimbecker also identified efficiencies at the West Coast terminals. He testified that the amount of grain that would be processed through P&H’s terminals in Vancouver would increase as a result of the Transaction. In his reply witness statement, Mr. Heimbecker further noted that the FGT’s location on Vancouver’s South Shore bypasses the congested rail corridors to the West Coast, which move grain from the Prairies to the West Coast terminals. As a result of its location, FGT would provide significant rail cycle time of 50% which many other grain purchasers would not enjoy. Mr. Heimbecker did not provide any evidence supporting his claim in improved rail car cycle time, nor did he provide any explanation or operating plans from the railways that attest to those savings.
·
Crop inputs efficiencies
[715] With respect to crop inputs efficiencies, Mr. Heimbecker testified that the Transaction allows P&H to compete more effectively with rival grain companies by converting the Elevators acquired in the Transaction into dual-purpose facilities. Previously, they were purely grain facilities, whereas P&H would now create a “one-stop shop” location that would include both the delivery of grain for sale and also enable farmers to purchase crop inputs. The sales of crop inputs at the newly acquired Elevators would provide P&H with increased margins. Mr. Heimbecker provided evidence of the cost per Elevator to convert a location into a combined grain/crop inputs facility and, based on his past business experience, he estimated the increased margins at XXX XXXXXXXXXXXXXXX. Mr. Heimbecker acknowledged that some of the crop inputs sales to be made at the Virden Elevator would come from sales made by other grain companies and crop input retailers. Mr. Heimbecker also advised that the expansion of crop inputs would not only benefit P&H through increased sales and margin but would also increase overall grain production.
(ii)
The Commissioner’s response
[716] The Commissioner’s Reply filed on February 17, 2020 stated that the Transaction would not generate cognizable gains in efficiencies to the extent claimed by P&H. The Commissioner reminded that his Application to the Tribunal sought the divestiture of just one Elevator, leaving nine others acquired by P&H as part of the Transaction. The contemplated orders requested would therefore not impact P&H’s ability to achieve the alleged efficiencies. The Commissioner further denied that any cognizable efficiencies would be greater than or offset the anti-competitive effects of the Virden Acquisition.
[717] As already noted, the Commissioner filed Mr. Harington’s expert report, which addressed each of the efficiencies claimed in Mr. Heimbecker’s initial witness statement. Mr. Harington’s report provided his analysis of how the five screens identified in Tervita CT applied to each of P&H’s claimed efficiencies.
·
Increased throughput at the Virden Elevator
[718] The Commissioner submitted that this claimed efficiency was caught by four of the five Tervita CT screens, namely, screens 1, 2, 3, and 5, summarized above in these Reasons. On Tervita CT screens 1 and 2, the Commissioner submitted that the increased volumes at the Virden Elevator would have occurred irrespective of the Acquisition, given the overall upward trend in grain production (measured by delivery volumes to Elevators) from 2019 to 2020. The Commissioner also noted that at the end of 2019, inventory levels at the Virden Elevator were low and had to be replenished by P&H after it acquired that Elevator. According to the Commissioner, those additional purchase volumes could partially explain the increase in delivered volume to Virden.
[719] The Commissioner further submitted that P&H had failed to demonstrate that the increased volumes at the Virden Elevator were not a redistribution of income between two persons under subsection 96(3) (Tervita CT screen 3). He argued that any increases in volume not attributable to the overall industry increase in grain production were merely a wealth transfer from other Elevators and did not represent a cognizable efficiency under section 96. According to the Commissioner, P&H also had not shown that its per unit variable operating costs of the Virden Elevator for the 10 months of 2020 were lower than LDC’s operating costs at this Elevator the previous year.
[720] The Commissioner further submitted that P&H had not shown that any increase in volume at the Virden Elevator could not have been achieved by an alternative purchaser of that Elevator (Tervita CT screen 5).
[721] Mr. Harington’s expert report concluded that increased throughput at the Virden Elevator was not a cognizable efficiency. In his view, any increase in throughput in P&H’s Virden Elevator and terminal network would have to arise from increased Canadian grain production and not be a pecuniary redistribution of throughput between P&H’s facilities and other facilities. To be a cognizable efficiency, any increased throughput would have to result from an increase in grain production brought about by the Virden Acquisition that was not “cannibalized” volume from other entities, and would not likely occur in the event of an order under section 92. Mr. Harington concluded that the claimed efficiencies based on an increase in throughput at the Virden Elevator did not qualify as a cognizable efficiency under these criteria.
·
Network logistics efficiencies
[722] Regarding the claimed logistics efficiencies, Mr. Harington noted in his report that P&H did not quantify the cost savings arising from shipping volumes to a more “freight logical” terminal and did not demonstrate that any of the savings would be lost in the event of a remedial order by the Tribunal. In addition, P&H did not show what proportion of any cost savings was attributable to the purchase of the Virden Elevator as opposed to the Rathwell facility.
·
Efficiencies at West Coast terminals
[723] Turning to the claimed efficiencies at the West Coast terminals, Mr. Harington’s expert report found that increased volumes at more efficient Vancouver area terminals were not a cognizable efficiency that would be lost in the event of an order of the Tribunal, because the same volume was still going through the less efficient Superior terminal in Thunder Bay; that is, the “only difference is the Elevators from which these volumes to Thunder Bay are coming.” Mr. Harington also noted that P&H did not offer any proof of increased volume of grain production due to the Acquisition and did not provide a comparison of operating costs at the Vancouver terminals with similar costs at another terminal through which the grain previously travelled.
·
Crop inputs efficiencies
[724] On crop input efficiencies, Mr. Harington’s expert report opined that the benefit of increased crop inputs sales would be a redistribution of income rather than a real resource saving. Any margin on crop inputs sales earned by P&H by the conversion of facilities was, in his view, a “pecuniary redistribution of income between P&H and farmers,” even if some portion of those sales were new sales rather than sales that would otherwise be made by rival retail suppliers.
[725] Mr. Harington noted the absence of any evidence (apart from Mr. Heimbecker’s opinion) that an additional crop inputs retail location would increase crop inputs sales in the area of the Virden Elevator rather than redistribute sales within the area, or would lead to more use of crop inputs by farmers and increase grain production as a result.
(iii)
The gains in efficiency under subsection 96(1)
[726] For the following reasons, the Tribunal concludes that P&H has not proven cognizable efficiencies under section 96.
·
Increased throughput at the Virden Elevator
[727] P&H advanced the position that the Acquisition had caused an increase in throughput at the Virden Elevator, by an increase in volume of grain delivered to and processed by that Elevator between 2019 and 2020.
[728] Mr. Heimbecker’s evidence about Elevators’ turn rates was not contradicted, nor did the Commissioner cross-examine him on it. Mr. Heimbecker testified that P&H had access to port terminals both in the Vancouver area and in Thunder Bay, compared with LDC which had no access to Thunder Bay and did not export from the Superior terminal. He also testified that the Virden Elevator is in P&H’s catchment area for Thunder Bay and that grain from Virden will generally be shipped east by P&H to its Superior terminal for sale to its eastern grain export customers. He testified that P&H’s most efficient port terminals are the AGT and the FGT.
[729] The Tribunal accepts that there are likely some time and transportation (i.e., rail) cost savings to move the Virden Elevator grain to the Superior terminal in Thunder Bay rather than to a Vancouver port terminal, which would presumably contribute to a higher turn rate at Virden for P&H than what existed when LDC sent its grain to Vancouver. However, P&H did not adduce evidence of any transportation cost savings. Having said that, Mr. Heimbecker’s evidence was also that P&H’s Superior terminal is not as efficient as the AGT or FGT in Vancouver. In light of that evidence, the Tribunal is unable to measure any benefits to the Canadian economy that would allegedly result from the change of the Virden Elevator grain being shipped west to the Vancouver port by LDC before the Transaction, to it being shipped east to the Superior terminal in Thunder Bay by P&H afterwards.
[730] Mr. Heimbecker provided a forecast that P&H would increase the Virden Elevator’s turn rate from historical rates of XX in 2017 and XX in 2018 while LDC operated Virden, to a turn rate of XX in P&H’s fiscal year 2021. Mr. Heimbecker did not present turn rates (actual or forecast) for the years 2019 or 2020, as the Virden Elevator was emptied by LDC following the agreement on the Transaction, and P&H needed to replenish it. The Virden Elevator’s forecast increase of Xx from 2018 to 2021 ranks just Xx of the 10 former LDC Elevators’ forecast increases, and is relatively small compared with some increases at other Elevators (which went up by as much as Xx). The Tribunal further observes that the Virden Elevator’s turn rate of XX in fiscal year 2021 will be the lowest of any of the former LDC Elevators. The Tribunal does not find this evidence about increased turn rates at the Virden Elevator to be particularly compelling or persuasive.
[731] The evidence about the relative inventory numbers between 2019 and 2020 must be assessed carefully. Mr. Harington testified that the abnormally lower inventory due to lower purchases by LDC in 2019, as well as the abnormally higher purchases by P&H in 2020 to make up the needed volume for inventory, would essentially have a “two-times” effect on the Tribunal’s ability to compare the two years. To compare them, each year would have to be adjusted (2019 upward and 2020 downward).
[732] In the Tribunal’s view, there are several additional uncertainties and anomalies associated with P&H’s evidence that undermine the reliability of any comparison of the change in throughput over these two years. These concerns prevent a sufficiently accurate quantification of any change in throughput following the Transaction and raise considerable doubt about the existence of any increase in throughput brought about by the Virden Acquisition.
[733] First, there is uncertainty about the baseline volumes in 2019. Mr. Heimbecker testified that LDC diminished the inventory levels at its Elevators, including at the Virden Elevator, prior to the closing of the Transaction on December 10, 2019. P&H did not provide the Tribunal with the Virden Elevator’s inventory level information as of that date. There was no evidence from LDC as to when it began to lessen its grain purchases (or if it simply stopped sometime before the closing), or about how much less grain it acquired at the Virden Elevator or decided not to store. The evidence also does not contain inventory levels or grain delivery volumes to the Virden Elevator in any prior years (e.g., 2018, 2017, or 2016) for historical comparison to 2019.
[734] Second, there is uncertainty about which product’s inventory decreased in 2019. In his initial witness statement, Mr. Heimbecker testified that there was little or no canola inventory at most of the LDC locations, including at the Virden Elevator. He testified that, by summer of 2020, P&H had “significantly increased canola purchases at Virden over the levels purchased there previously by LDC.” In his reply witness statement, Mr. Heimbecker explained that one reason why P&H missed its three-month throughput forecast for May to July 2020 was “low level of inventory left by LDC. LDC did not have grain in the pipeline when P&H purchased the assets which prevented P&H from selling grain in the beginning of the year” [emphases added] (Heimbecker Statement at para 51; Heimbecker Reply Statement at para 24).
[735] At the hearing, Ms. Sanderson testified, having looked at the Virden Elevator’s purchase data, that she did not see any sort of change on the canola side in December 2019 or January 2020 compared to the year prior. She attributed the running down of inventories just to wheat.
[736] Third, there were anomalous monthly deliveries of both canola and wheat that make 2019 and 2020 harder to compare. Mr. Heimbecker provided monthly MT deliveries of canola to the Virden Elevator from January 1 to October 31, 2020. Of the approximately XXXXXXXXXXXX in those ten months, XXXXXXXXXXXX was delivered in September 2020. That month saw deliveries of XXXXXX in 2019, implying an increase of over XXXx comparing September 2019 with September 2020.
[737] There was also an anomalous month in the data for wheat. Ms. Sanderson testified that November 2019 wheat purchases by LDC were more than twice larger than October 2019 and were “by far and away the largest” quantity purchased in any month from January 2019 to July 2020 (the period of available actual data when she prepared her report) (Consolidated Transcript, Confidential A, at p 1895). Ms. Sanderson’s review of volumes from September to December 2019 allowed her to conclude that LDC increased purchases in November and may have run down inventories in December. As noted, the Transaction closed on December 10, 2019, and the parties did not provide the Tribunal with inventory level information at the Virden Elevator as of that date.
[738] Fourth, in order to compare the 2019 and 2020 deliveries to the Virden Elevator, both parties agreed that the 2020 deliveries had to be adjusted for an overall industry increase in grain production over 2019. Mr. Heimbecker used the overall grain production increase in Canada of 13.1% for the crop years ending July 31, 2019 compared with July 31, 2020 to make adjustments. However, Mr. Harington’s presentation of the industry evidence on increased deliveries indicated that the increases in both all-wheat production and all-grain production in Saskatchewan and Manitoba were each different from the 13.1% used in the calculations: higher for Saskatchewan (+15.8% and +17.9%, respectively) and noticeably lower for Manitoba (+6.8% and +7.0%, respectively). The Virden Elevator’s purchases come from those two provinces.
[739] Mr. Heimbecker’s revised calculations in his reply witness statement continued to use the 13.1% Canada all-grain increase, referring to the Manitoba increases in all-wheat production and all-grain production and tendering his calculations as a “conservative” approach. However, Mr. Heimbecker acknowledged in cross-examination that overall canola deliveries in Manitoba increased by 18%, comparing January to October 2019 with January to October 2020. At the same time, Mr. Heimbecker was also careful to distinguish between grain or canola production and delivery.
[740] The Tribunal finds that an adjustment of 13.1% to account for increased grain production in Canada does not account accurately for grain delivery increases to the Virden Elevator from farms in Saskatchewan and Manitoba from 2019 to 2020, and negatively affects the Tribunal’s ability to quantify any increase in throughput at the Virden Elevator over the period.
[741] In light of all this evidence, the Tribunal finds that there is considerable uncertainty about the existence and quantum of any actual increase in throughput at the Virden Elevator following the Acquisition. The Tribunal cannot conclude that any specific amount of increased throughput volume of wheat, or canola, or grain, has been demonstrated or satisfactorily quantified for 2020 after the Transaction.
[742] The Tribunal is also unable to conclude that any increase or any quantifiable increase in throughput was brought about by the Virden Acquisition for the purposes of Tervita CT screen 2. P&H offered no expert assistance to assist the Tribunal in identifying and quantifying throughput increases brought about by the Virden Acquisition (if any) and to distinguish it from higher throughput caused by rising overall grain deliveries to replenish inventory (the quantum of which was itself debated, as discussed above).
[743] In addition, while the monthly trend comparing 2019 to 2020 for canola was generally rising from March to September 2020, the spike in canola deliveries in September 2020 (comprising nearly half of the delivered volume for the calendar year) is inconsistent with P&H’s position that there are sustainable, ongoing efficiencies in throughput at the Virden Elevator. As Mr. Harington testified, one month does not make a trend.
[744] The Tribunal concludes that P&H has not demonstrated, with clear and convincing evidence, any section 96 efficiencies with respect to increased throughput at the Virden Elevator.
·
Network logistics efficiencies
[745] With respect to network logistic efficiencies, the Tribunal agrees with the Commissioner that P&H has not quantified its alleged efficiencies arising from the optimization of shipments to “freight logical” terminals in Vancouver. The evidence does not establish how much additional or incremental grain has been or will be diverted to the grain terminals in Vancouver (from other western terminals or from Thunder Bay), nor any of the cost savings that resulted from processing grain at the Vancouver terminals compared with the Superior terminal, nor any cost savings associated with processing and transporting grain a shorter distance from an Elevator to a terminal. While the possibility of a synergy or an efficiency arising from better network logistics has some intuitive attraction, the evidence does not support the existence or any quantification of such an efficiency for the purposes of section 96.
[746] Applying the principles articulated in Tervita SCC, this proposed efficiency must be considered a zero.
·
Efficiencies at West Coast terminals
[747] The Tribunal also concludes that the evidence does not support any proven efficiencies arising from any additional grain flowing through the terminals in Vancouver used by P&H. Again, P&H did not offer any quantification of the incremental volumes of grain from the Virden Acquisition, or any cost savings as a result of processing such grain, in the Vancouver terminals or by a shift in volumes from other grain terminals to its terminals. Moreover, Mr. Heimbecker did not provide any evidence supporting his claim of improved rail car cycle time nor did he provide any explanation or operating plans from the railways that attest to those savings. Applying the principles set out in Tervita SCC, this proposed efficiency must also be considered a zero.
·
Crop inputs efficiencies
[748] Turning finally to the claimed crop inputs efficiencies, the Tribunal concludes that P&H has not proven any section 96 efficiencies related to crop inputs. An increase in margins accruing to P&H is not a benefit to the Canadian economy for the purposes of an efficiency under section 96. In addition, P&H has not measured, quantified, or even estimated any increase in output as alleged (i.e., any increase in the sale of crop inputs in the area around the Virden Elevator, nor any increase in local grain production). Needless to say, such claims in efficiencies were, by their nature, clearly quantifiable.
[749] The Tribunal therefore concludes that P&H has not demonstrated any cognizable efficiencies, whether quantitative or qualitative, under subsection 96(1) of the Act. Since P&H failed to meet its burden, the efficiencies are assigned a weight of zero.
(iv)
The impact of subsection 96(2)
[750] P&H submitted that the Tribunal should also apply subsection 96(2), since its objective in entering the Transaction was to increase its competitiveness in domestic and international export markets and to maximize the profitability of its export business. P&H emphasized that the Transaction served to enhance the return on its investment in the FGT through the expansion of its network, by acquiring access to increased throughput of grain sourced from the acquired LDC Elevators. P&H emphasized that it had no anti-competitive intention. Its intention was not to exercise monopsony power in local input markets, but rather to increase output in export markets.
[751] P&H pointed to its focus — and the focus of the industry as a whole — on exporting canola and wheat. It noted that its pricing mechanism (i.e., the Workback Algorithm) tied local prices in the input market to prices in the export market. The acquisition of the grain volumes represented by the LDC Elevators, including the Virden Elevator, enabled large increases in the volume processed by the FGT without displacing any planned volumes from pre-Transaction P&H Elevators. Those increased throughput volumes would be achieved quickly, rather than waiting for XX years as P&H originally anticipated and planned for the FGT.
[752] P&H referred to Mr. Brooks’ evidence and to a report from the Saskatchewan Wheat Development Commissioner (found in Exhibits CA-R-240 and P-R-241, Compilation of Miscelleaneous Documents) to support a submission that grain producers’ (i.e., farmers’) interests in increased production were aligned with grain companies’ interests in increasing export capacity.
[753] P&H submitted that the Elevators acquired in the Transaction:
[…] will be capable at full utilization to fill the XXXXXx metric tonnes of capacity at the FGT. This capacity utilization represents a 9% increase in export sales capacity at the West Coast and approximately 6.5% of total Canadian export capacity. The contribution of the LDC Elevators to export sales pre-acquisition was XXXXXX MT. At full capacity, within the P&H network, these elevators will add XXXXXX MT of export throughput to FGT for a net increase of XX. Hence the transaction as a whole clearly leads to a significant increase in the real [value] exports of grain within the meaning of section 96(2).
[Emphasis added.]
[754] The Commissioner responded that there was no evidence to indicate that the Transaction had actually led to a real increase in the value of exports under subsection 96(2). The Commissioner contended that P&H had acquired the grain volume that LDC was exporting through Vancouver terminals, so that P&H could export the same grain volume through its terminals in Vancouver. The grain was all headed for export, regardless of the Transaction. In addition, no analysis had been done to demonstrate that allowing P&H to keep one Elevator (i.e., the Virden Elevator) would lead to a significant increase in the real value of exports.
[755] The Tribunal disagrees with P&H’s submission that the increase in the real value of exports from the Transaction “as a whole” may be considered under subsection 96(2). That provision refers to the gains in efficiency mentioned in subsection 96(1), which in turn refers to the efficiencies brought about by the merger or proposed merger “in respect of which the application is made.” Consideration of a substantial increase in the real value of exports under paragraph 96(2)(a) must therefore focus on the specific merger being challenged by the Commissioner in the proceeding. In this case, the Commissioner’s Application is solely made in respect of P&H’s Acquisition of the Virden Elevator, not in respect of the Transaction as a whole.
[756] During oral argument, the Tribunal asked whether there was anything in the evidence that parsed out the impact of the Virden Elevator on exports. P&H’s answer was no. It advised that it would need access to third party information and data about grain volumes from other grain companies at the port and their efficiencies there, as well as those companies’ variable operating costs to reach the port facilities, in order to measure such impact.
[757] During argument, the Tribunal requested that P&H specifically refer to the evidence on the additional (incremental) volume attributable to the Virden Elevator that would be exported through the FGT. P&H did not provide a satisfactory answer in substance, pointing to its evidence of increased throughput at the Virden Elevator. However, as noted above, this evidence does not support the quantification of any volume of increased throughput at the Virden Elevator. The Tribunal also notes that volumes of grain from the Virden Elevator were delivered to Thunder Bay (not Vancouver) and that the Superior terminal in Thunder Bay closed during the winter. Because P&H redirected the Virden Elevator grain eastward to Thunder Bay, any increased volume to the FGT that could (in theory) be attributable to the Virden Elevator would have to be sourced from another Elevator(s). P&H did not provide an analysis to support such an attribution.
[758] Reduced to its essence, P&H’s submission was that more of the FGT’s capacity would be used, and sooner, as a result of the Transaction as a whole. The evidence of how much additional capacity would be used (or the value of that grain in dollars), and when, is insufficient to show a significant increase in the real value of exports resulting from the Virden Acquisition. The Tribunal agrees with the Commissioner that it is likely that most of the additional volumes anticipated to be processed at the FGT post-Transaction are volumes that would otherwise have been processed at other terminals but will now be diverted to the FGT because P&H now owns all of the source Elevators. Further, P&H confirmed during argument that the additional volumes are for all types of grains, not just wheat and canola. Lastly, the Tribunal appreciates the logic of Mr. Heimbecker’s evidence concerning congested rail lines. Even accounting for that evidence and the general claim that the FGT would be a more efficient terminal (it was scheduled to become fully operational shortly after the hearing), the Tribunal is not satisfied that the evidence demonstrates a significant increase in the real value of exports for the purposes of paragraph 96(2)(a).
[759] The Tribunal pauses to again note that P&H’s intention or objectives in entering the Transaction are not relevant, or material, to the Tribunal’s analysis.
[760] Given the analysis above with respect to subsection 96(1), the Tribunal finds that P&H has not shown a causal connection between any proven efficiencies under subsection 96(1) and an increase in the real value of any exports. The Tribunal therefore concludes that the requirements of subsection 96(2) have not been met.
(c)
The trade-off analysis
[761] In light of the Tribunal’ conclusions on efficiencies, there is no need to deal with the trade-off analysis in this case.
(3)
Conclusion on efficiencies and section 96
[762] For the reasons detailed above, the Tribunal concludes that P&H has not demonstrated, with clear and convincing evidence, its claimed efficiencies and that it would not have met its burden of demonstrating, on a balance of probabilities, that its claimed gains in efficiency would be greater than, and would offset, the anti-competitive effects of any lessening of competition resulting from the Acquisition.
VIII.
Conclusion
[763] For the above detailed reasons, the Commissioner’s Application is dismissed. In light of this conclusion, no remedial action will be ordered.
IX.
COSTS
[764] The parties were unable to come to an agreement as to costs.
[765] The Commissioner submits that he should be awarded a lump sum amount of CAD $2 million inclusive of counsel fees and disbursements if he is successful. If the Application is dismissed, then the Commissioner argues that P&H should be awarded CAD $2 million inclusive of counsel fees and disbursements. However, if the Application is dismissed and the Tribunal finds that P&H’s section 96 efficiencies defence would not have been successful, then the Tribunal should, in the Commissioner’s view, deduct CAD $500,000 from the lump sum cost award, in recognition of the costs the Commissioner incurred in order to respond to that defence. While the Commissioner recognizes that P&H was entitled to rely on the efficiencies defence, he argues that if a respondent pleads the defence but does not adduce sufficient evidence to make it out, then the Tribunal should use a costs award to recognize the significant costs incurred by the Commissioner (and ultimately, Canadian taxpayers) to respond. If there is no financial deterrent associated with an unsuccessful efficiencies defence, the Commissioner submits that in the future, respondents will claim efficiencies as a matter of course, causing significant financial burden on the Commissioner regardless of whether raising the efficiencies defence was justified.
[766] P&H, in turn, seeks costs payable as a lump sum in the amount of CAD $2,206,958.18, inclusive of fees, disbursements and taxes, if the Commissioner’s Application is dismissed. This sum represents approximately CAD $209,000 for legal fees and approximately CAD $1,998,000 for disbursements (both inclusive of taxes). Should the Application be allowed, P&H indicates that it takes no position with respect to the Bill of Costs submitted by the Commissioner, save for one item — i.e., the preparation and filing of the Commissioner’s motion materials dealing with confidentiality designations — which P&H maintains is an ineligible cost in view of the Tribunal’s order dismissing the Commissioner’s confidentiality motion without costs. As for the matter of costs relating to the efficiencies defence, P&H submits that the merits of the efficiencies defence becomes moot if a substantial lessening of competition is not found under section 92, and that the result of the case must drive costs, not obiter dicta. According to P&H, this is not a case of divided success (citing Allergan Inc v Sandoz Canada Inc, 2021 FC 186 at paras 31, 43).
[767] Both parties submitted bills of costs and affidavits in support.
A.
Legal principles applicable to costs
[768] In VAA CT, the Tribunal noted that section 8.1 of the CTA grants jurisdiction to the Tribunal to award costs of proceedings before it in accordance with the provisions governing costs in the FC Rules (VAA CT at para 817). Under subsection 400(1) of the FC Rules, the Tribunal has “full discretionary power over the amount and allocation of costs and the determination of by whom they are to be paid.” A non-exhaustive list of factors that the Tribunal may consider when exercising its discretion is set out in subsection 400(3).
[769] Costs ordinarily follow the outcome of the proceeding, in that the successful party is usually awarded costs (see, for example, Heiltsuk Horizon Maritime Services Ltd v Atlantic Towing Limited, 2021 FCA 26 at para 182; MacFarlane v Day & Ross Inc, 2014 FCA 199 at para 6; VAA CT at para 816; FC Rule 400(3)(a)).
[770] The costs regime does not indemnify the successful party for all of its legal fees and disbursements, absent very unusual circumstances. Costs are only partial compensation for the actual costs incurred in litigation. As noted in VAA CT, an award of costs represents a compromise between compensating a successful party and not unduly burdening an unsuccessful party (VAA CT at para 817, citing Apotex Inc v Wellcome Foundation Ltd (1998), 159 FTR 233 (FCTD), 84 CPR (3d) 303, aff’d (2001), 199 FTR 320 (FCA)).
[771] The objectives of a costs award include having the unsuccessful party make a “reasonable contribution” to the successful party’s costs of litigation, having regard to the Tariff in the FC Rules (NOVA Chemicals Corporation v Dow Chemical Company, 2017 FCA 25 (“NOVA Chemicals”) at paras 13, 21). Although the Tariff amounts may be inadequate in complex litigation, nevertheless, an increased costs award cannot be justified solely on the basis that a successful party’s actual fees are significantly higher than the Tariff amounts. The burden is on the party seeking increased costs to demonstrate why their particular circumstances warrant an increased award (NOVA Chemicals at para 13, citing Wihksne v Canada (Attorney General), 2002 FCA 356 at para 11).
[772] The approximation of lump sum costs is a matter of judgment rather than an accounting exercise (Consorzio del Prosciutto di Parma v Maple Leaf Meats Inc, 2002 FCA 417 at para 8). While the costs awarded should have a fair relationship to the actual costs of litigation, the question for the Tribunal is what, in the circumstances, are necessary and reasonable legal costs and disbursements (Nadeau Ferme Avicole Ltée v Groupe Westco Inc, 2010 Comp Trib 1 at para 49).
[773] Disbursements must be reasonable, necessary, and justified (NOVA Chemicals at para 26; VAA CT at para 821). Expert-related costs are not automatically recovered in their entirety and can be adjusted by the Tribunal if they do not appear reasonable (VAA CT at para 822).
[774] In VAA CT, the Tribunal took into account that success on the issues in dispute, particularly the legal issues in dispute, was divided; although the respondent was successful overall, the Commissioner prevailed on certain issues (VAA CT at paras 819, 827). The Tribunal reduced the costs award to the respondent to reflect the time spent on issues on which the Commissioner prevailed but the respondent persisted in spending time, and based on the reasonableness and necessity of the disbursements.
[775] As paragraph 400(3)(a) of the FC Rules contemplates, success overall in the proceeding remains a principal factor, subject to additional considerations relevant to the circumstances and claims made. The Tribunal may consider the factors in paragraphs 400(3)(b) to (n.1) of the FC Rules and any other matter it considers relevant under subsection 400(3)(o).
[776] The Tribunal favours lump sum costs awards (VAA CT at para 825).
B.
Tribunal’s assessment
[777] As the successful party, P&H is entitled to an award of costs. The Tribunal will fix the costs payable by the Commissioner in this proceeding.
(1)
Legal fees
[778] With respect to legal fees, P&H claimed approximately CAD $209,000 in respect of legal fees, calculated based on Tariff B, Column IV. Its claims included time spent for preparation of pleadings, affidavits of documents, preparing for and attending oral examinations for discovery, preparing for and attending case management conferences, preparation of witness statements, and the attendance of two counsel at the hearing. The Commissioner submitted that, if P&H is successful, its quantum of costs should be determined on the usual basis of the middle of column III, and should also be less than the amount fixed by the Tribunal in VAA CT (which was CAD $70,000) because the matter occupied fewer hearing days.
[779] As noted, both parties took positions to increase or decrease an award of costs by large amounts beyond the Tariff. P&H increased its requested award based on its actual legal fees and the complexity of this proceeding. P&H advised that it believed that an award for legal fees in the amount of CAD $900,000 would be appropriate (its actual legal fees were approximately CAD $3.6 million), but only claimed CAD $209,000 under Tariff B, Column IV. For his part, the Commissioner sought to decrease an award to P&H if it did not succeed on section 96 issues. The Commissioner submitted that P&H’s costs award should be reduced by CAD $500,000, which the Commissioner argued represented the amount he paid to respond to P&H’s position under section 96 by having to quantify the anti-competitive effects (through Dr. Miller) and file Mr. Harington’s expert report on efficiencies under section 96.
[780] The parties also argued several novel points — i.e., issues that had not been argued to the Tribunal in previous litigated proceedings. The Commissioner’s position on product market raised new issues for the Tribunal’s consideration, while P&H raised issues under section 96 that had not been considered previously.
[781] The most important overall factor in arriving at a costs award is which party succeeded. Here, the Tribunal dismissed the Commissioner’s Application. In addition to the overall result, the Tribunal recognizes that this proceeding involves a public official with a statutory mandate to administer and enforce the Act; both parties are highly sophisticated with very experienced counsel; and the legislative setting contemplates significant pre-litigation disclosure through the merger review process and pre-hearing disclosure, as well as well known elements and burdens of proof under sections 92 and 96 of the Act. The Tribunal also finds that proceedings under section 92 involve complex legal and factual matters that support higher costs awards under the Tariff B, Column IV in the FC Rules (as claimed by P&H).
[782] Although the Commissioner succeeded on several preliminary issues, the Tribunal does not find that those arguments diminish P&H’s entitlement to an award of costs in this case.
[783] The Tribunal does not agree with the Commissioner’s position that the costs award to P&H should be reduced by an overall lump sum amount of CAD $500,000 because P&H would not have succeeded on its section 96 defence. Although the cost of Mr. Harington’s services are known (i.e., CAD $259,000), the balance to arrive at the claimed amount of CAD $500,000 is merely an assumption or guess without a sufficient evidentiary basis.
[784] That said, however, the Tribunal finds it appropriate under FC Rule 400(3)(o) (and by analogy to other paragraphs in FC Rule 400(3)) to take into account the specific circumstances of this proceeding related to the section 96 evidence and arguments in its overall assessment of legal fees, as follows:
· The overall burden of proof under section 96 was on P&H;
· P&H raised efficiencies in its pleading. The Tribunal notes that P&H did not provide details of its position on efficiencies at examinations for discovery and did not file an expert report for the hearing — even though it advised it would do so during the discovery process. It only filed Mr. Heimbecker’s fact evidence (which included some efficiencies arguments that P&H did not initially plead);
· The Commissioner did not waste time on section 96 issues during the fact portion of the hearing; he did not cross-examine Mr. Heimbecker on his evidence related to alleged efficiencies;
· The Commissioner prevailed on section 96 issues. Even if the Tribunal’s conclusions on section 96 were, strictly speaking, unnecessary for the Tribunal to decide given the outcome of its analysis under section 92, the Commissioner had no practical alternative but to respond to the section 96 efficiencies defence raised by P&H and to do so with an expert report;
· The Commissioner had to prepare for and conduct discovery on section 96 issues, quantify the anti-competitive effects in accordance with the principles established in Tervita SCC, file an expert report, address section 96 issues at the hearing, and respond to issues related to the proper interpretation of subsection 96(2), all of which affected the time spent by legal counsel;
· The Tribunal considers that P&H’s approach to the section 96 issues in this proceeding tended to unnecessarily increase the Commissioner’s costs and increase the time spent on the proceeding. A considerable part of the Commissioner’s legal costs in relation to section 96 and its disbursement for Mr. Harington’s report could have been avoided.
[785] Exercising its discretion, the Tribunal concludes that the appropriate costs award to P&H for legal fees in this matter should be fixed at CAD $157,000, which represents approximately 75% of P&H’s legal fees as claimed under Tariff B, Column IV.
(2)
Disbursements
[786] The Tribunal has considered the positions of both parties with respect to each of the claims made by P&H for the disbursements it incurred in this litigation.
[787] P&H claims expert fees in the amount of approximately CAD $1.61 million. Having regard to the Tribunal’s positive treatment of Ms. Sanderson’s evidence, but also to the overall reasonableness of the quantum claimed by P&H to be reimbursed by the Commissioner, the Tribunal finds CAD $1.2 million to be a reasonable sum in respect of expert fees.
[788] The Tribunal recognizes that this hearing was conducted not only electronically (as is standard at the Tribunal) but entirely virtually, and in very unusual circumstances owing to the COVID-19 pandemic. P&H decided that its counsel would travel to Winnipeg, at or close to its witnesses (particularly Mr. Heimbecker), and presumably close to P&H’s offices. For the hearing, P&H set up an operations centre at a hotel, necessitating the rental of a large room (to maintain physical distancing and set up the appropriate equipment for a virtual hearing with all the required computers and technical support).
[789] P&H claims a disbursement of approximately CAD $126,000 for hotel conference rooms and audio visual display equipment used during examinations for discovery and, later, during the facts portion of the hearing. The Commissioner submitted that these amounts were excessive, noting his own claim for just CAD $2,200. The Tribunal does not accept the Commissioner’s comparison of P&H’s costs for four witnesses to testify in Manitoba, versus the 10 witnesses called by the Commissioner. Having decided that counsel would travel to Winnipeg (which the Tribunal does not find appropriate to question in this case), the Tribunal finds that the rental of space and equipment was reasonable given the COVID-19 restrictions in Manitoba at the time. Although some charges on the invoices related to food and package deliveries and the amounts charged for space and equipment appear quite high on a daily basis, the Tribunal finds it appropriate that the Commissioner make a reasonable contribution to this expense in the amount of CAD $50,000.
[790] P&H claimed payments made for case law searches in third party legal databases in the amount of approximately CAD $32,000. The evidence reveals more than 300 searches, done mostly in the two months leading up to the hearing. It is unclear whether those searches related to two motions argued and decided during the same period, or to the legal issues arising in the hearing itself. (On one motion, costs were awarded in the cause whereas no costs were awarded in the other.) The Tribunal recognizes that most legal research is done online and at the time, the law firm personnel were likely working from home and without a law library. It is also clear that some of the legal issues raised by the Commissioner’s position and by P&H (for example, preliminary issues and the interpretation of subsection 92(2)) required legal research. Although the Commissioner did not object in his submissions, the Tribunal finds that the high number of searches and absence of details as to what the searches concerned (motions as opposed to hearing; issues raised by each party) support a reasonable claim for CAD $8,000.
[791] P&H sought reimbursement for air travel to and from Winnipeg in the amount of CAD $31,500. Its claim was based on 50% of the actual cost of a private jet for two counsel to travel to Winnipeg and back for discovery, and later for the fact portion of the hearing. Reviewing the invoices, it appears that at both the discovery and hearing stages, the aircraft flew from Winnipeg to Toronto (without passengers on board other than flight crew) to pick up P&H’s counsel and returned the same day. The aircraft made a trip to return counsel to Toronto upon completion of both the discoveries or portion of the hearing, and then flew back to Winnipeg without passengers on board other than flight crew. The invoices reflect charges for round-trips even though counsel were on board one way only. The Tribunal will allow a claim for CAD $4,500, which (on the evidence) approximates a full fare economy air ticket for two counsel to fly between Toronto and Winnipeg for discovery and for the fact portion of the hearing.
[792] P&H claims approximately CAD $31,600 for transcripts of the examinations for discovery and the hearing and approximately CAD $10,600 for data hosting, which was necessary for the virtual hearing at the Tribunal. The Tribunal allows these claims in their entirety.
[793] The Tribunal allows claims for photocopies and printing in the amount of CAD $800 and for hotels and meals during examinations for discovery and at the hearing in the aggregate of the amount of CAD $8,000 (based on a contribution to the cost of hotel rooms for two counsel and a reasonable per diem for meals).
[794] P&H claimed approximately CAD $6,000 in courier costs attributable, for example, to counsel working from home during the pandemic and materials sent by counsel to the panel members and the Tribunal Registry during the hearing. The Tribunal notes that most (approximately CAD $4,200) of P&H’s claim in that regard relates to a single package sent from Winnipeg to Toronto to return materials and equipment after the fact portion of the hearing ended. The Tribunal considers CAD $2,000 as a reasonable contribution towards courier costs.
[795] P&H claimed meals in the amount of approximately CAD $1,350 in addition to those claimed by hearing counsel, which the Tribunal notes was not an appropriate claim for costs purposes.
[796] A claim for conference calls in the amount of CAD $127 is de minimis in this context. The Tribunal notes that certain calls occurred before the litigation began.
C.
Conclusion on costs
[797] In light of the foregoing, the Tribunal awards costs for legal fees in the lump sum amount of CAD $157,000, inclusive of applicable taxes. The total of all disbursements allowed is CAD $1,315,500, inclusive of applicable taxes.
[798] The Tribunal therefore concludes that the Commissioner shall pay an all-inclusive aggregate lump sum amount of CAD $1,472,500 to P&H in respect of costs of this proceeding.
X.
ORDER
[799] The Application brought by the Commissioner is dismissed.
[800] Within 30 days from the date of this Order, the Commissioner shall pay to P&H an amount of CAD $1,472,500.
[801] These Reasons are confidential. In order to enable the Tribunal to issue a public version of the Reasons, the Tribunal directs the parties to attempt to reach an agreement regarding the redactions to be made to these Reasons in order to protect confidential evidence and information. The parties are to jointly correspond with the Tribunal Registry by no later than the close of business on November 14, 2022, setting out their agreement and any areas of disagreement concerning the redaction of the confidential version of the Reasons. If there is any disagreement, the parties shall separately correspond with the Tribunal setting out their respective submissions with respect to any proposed, but contested, redactions from these confidential Reasons. Such submissions are to be served and filed with the Tribunal Registry by the close of business on November 14, 2022.
| DATED at Ottawa, this 31st day of October, 2022
|
| SIGNED on behalf of the Tribunal by the Panel Members.
|
| (s) Denis Gascon J. (Presiding Member)
(s) Andrew D. Little J.
(s) Ramaz Samrout
|
COUNSEL OF RECORD:
For the applicant:
| Commissioner of Competition
|
| Jonathan Hood
|
| Ellé Nekiar
|
For the respondent:
| Parrish & Heimbecker, Limited
|
| Robert S. Russell Joshua Abaki
Carolyn Wong
|
Schedule “A” – Relevant provisions of the Act
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Schedule “B” – List of Exhibits
| Exhibit No.
|
Description
|
| P-A-001
|
Witness Statement of Mr. Alistair Pethick
|
| CA-R-002
|
Purchase Receipts from Pethick Farms (Confidential – Level A)
|
| P-R-003
|
Purchase Receipts from Pethick Farms
|
| CA-R-004
|
2018 Purchase Receipts from Pethick Farms (Confidential – Level A)
|
| P-R-005
|
2018 Purchase Receipts from Pethick Farms
|
| CA-R-006
|
Excel Table Summarizing Purchases of Grain by Pethick Farms (Confidential – Level A)
|
| P-R-007
|
Excel Table Summarizing Purchases of Grain by Pethick Farms
|
| CA-R-008
|
2019 Purchase Receipts from Pethick Farms (Confidential – Level A)
|
| P-R-009
|
2019 Purchase Receipts from Pethick Farms
|
| CA-R-010
|
Purchase Contracts from Pethick Farms (Confidential – Level A)
|
| P-R-011
|
Purchase Contracts from Pethick Farms
|
| CA-R-012
|
Grain Purchase Order Agreement of Mr. Pethick (Confidential – Level A)
|
| P-R-013
|
Grain Purchase Order Agreement of Mr. Pethick
|
| P-A-014
|
Witness Statement of Mr. Harvey Brooks
|
| P-R-015
|
PDQ Update Notice from Alberta Wheat Commission
|
| P-R-016
|
“Where’s My Region?” Map from PDQ
|
| P-R-017
|
Rosetown “Where’s My Region?” Map from PDQ
|
| P-R-018
|
Virden “Where’s My Region?” Map from PDQ
|
| P-R-019
|
Moosomin “Where’s My Region?” Map from PDQ
|
| P-R-020
|
Fairlight “Where’s My Region?” Map from PDQ
|
| P-R-021
|
Wheat Market Outlook and Price Webpages from Sask Wheat
|
| P-R-022
|
FOB Wheat Prices and Export Basis Prices Calculation pdf from Sask Wheat
|
| P-R-023
|
Sask Wheat Strategic Plan 2018-2020
|
| P-R-024
|
Sask Wheat Port Capacity Article
|
| CB-A-025
|
Witness Statement of Mr. Chris Lincoln (Confidential – Level B)
|
| P-A-026
|
Witness Statement of Mr. Chris Lincoln
|
| CA-R-027
|
Chris Lincoln Deliveries 2016-2019 (Confidential – Level A)
|
| P-R-028
|
Chris Lincoln Deliveries 2016-2019
|
| CA-R-029
|
Chris Lincoln Grain Purchase Agreements (Confidential – Level A)
|
| P-R-030
|
Chris Lincoln Grain Purchase Agreements
|
| CA-R-031
|
Balance Sheet for CKB Lincoln Farms for Wawota and Maryfield (Confidential – Level A)
|
| P-R-032
|
Balance Sheet for CKB Lincoln Farms for Wawota and Maryfield
|
| P-A-033
|
Witness Statement of Mr. Ian Wagstaff
|
| CA-R-034
|
Ian Wagstaff Deliveries 2016-2019 (Confidential – Level A)
|
| P-R-035
|
Ian Wagstaff Deliveries 2016-2019
|
| P-A-036
|
Witness Statement of Mr. Dean McQueen
|
| CA-A-037
|
Witness Statement of Mr. Dean McQueen (Confidential – Level A)
|
| CB-A-038
|
Witness Statement of Mr. Dean McQueen (Confidential – Level B)
|
| CB-R-039
|
Competition Bureau RFIs (Confidential – Level B)
|
| P-R-040
|
Competition Bureau RFIs
|
| CA-R-041
|
Viterra’s Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-042
|
Viterra’s Response to the Competition Bureau RFI
|
| P-R-043
|
Canadian Grain Commission Statistics Webpage
|
| P-R-044
|
Canadian Grain Commission Grain Deliveries at Prairie Points Webpage
|
| P-A-045
|
Witness Statement of Mr. Ray Elliott
|
| P-A-046
|
Bunge Limited General Terms and Conditions
|
| P-A-047
|
Witness Statement of Mr. Brett Malkoske
|
| CB-A-048
|
Witness Statement of Mr. Brett Malkoske (Confidential – Level B)
|
| P-A-049
|
Reply Witness Statement of Mr. Brett Malkoske
|
| CB-A-050
|
Reply Witness Statement of Mr. Brett Malkoske (Confidential – Level B)
|
| CA-A-051
|
G3 Purchase Contract Terms and Conditions (Confidential – Level A)
|
| P-A-052
|
G3 Purchase Contract Terms and Conditions
|
| CA-R-053
|
G3’s Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-054
|
G3’s Response to the Competition Bureau RFI
|
| P-A-055
|
Witness Statement of Ms. Darcy Jordan
|
| CA-R-056
|
Cargill’s Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-057
|
Cargill’s Response to the Competition Bureau RFI
|
| P-A-058
|
Witness Statement of Ms. Kara Hawryluk
|
| CA-R-059
|
LDC’s Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-060
|
LDC’s Response to the Competition Bureau RFI
|
| P-A-061
|
Witness Statement of Mr. Jeff Wildeman
|
| P-A-062
|
Ceres Standard Terms and Conditions
|
| CA-R-063
|
Ceres’ Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-064
|
Ceres’ Response to the Competition Bureau RFI
|
| P-A-065
|
Witness Statement of Mr. Michael Irons
|
| CA-R-066
|
ADM’s Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-067
|
ADM’s Response to the Competition Bureau RFI
|
| P-A-068
|
Witness Statement of Mr. Bryce Geddes
|
| P-A-069
|
Richardson Terms and Conditions for 2015-2016 Purchase Contracts
|
| P-A-070
|
Richardson Terms and Conditions for 2016-2017 Purchase Contracts
|
| P-A-071
|
Richardson Terms and Conditions for 2017-2018 Purchase Contracts
|
| P-A-072
|
Richardson Terms and Conditions for 2018-2019 Purchase Contracts
|
| CA-R-073
|
Richardson Yorkton’s Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-074
|
Richardson Yorkton’s Response to the Competition Bureau RFI
|
| CA-R-075
|
Richardson Grain Elevators’ Response to the Competition Bureau RFI (Confidential – Level A)
|
| P-R-076
|
Richardson Grain Elevators’ Response to the Competition Bureau RFI
|
| P-R-077
|
Witness Statement of Mr. Kristjan Hebert
|
| P-A-078
|
Print off from the HGV Website
|
| P-A-079
|
Article “Stop Leaving Money in the Field. Learn the 5% Rule”
|
| P-A-080
|
Extracts from Mr. Hebert 5% Rule Presentation
|
| P-A-081
|
Transcripts from “Lessons Learned in Marketing” - Phoenix Group Podcast, dated January 15, 2018
|
| CA-A-082
|
Email from Mr. Hebert, dated March 21, 2018 (Confidential – Level A)
|
| CB-A-083
|
Email from Jeremy Krainik, dated October 24, 2018 (Confidential – Level B)
|
| CA-A-084
|
P&H Contracts and GPOs with HGV (Confidential – Level A)
|
| CA-A-085
|
HGV Transactions from the 2016 to 2021 Crop Years (Confidential – Level A)
|
| CA-A-086
|
Settlement Receipts of HGV with Other Grain Elevators (Confidential – Level A)
|
| CA-A-087
|
Email from Jeremy Krainik, dated November 29, 2019 (Confidential – Level A)
|
| CA-A-088
|
Spreadsheet Containing HGV Transactions Collected from Third Party Data (Confidential – Level A)
|
| P-A-089
|
Email from Jeremy Krainik, dated October 24, 2018
|
| P-A-090
|
P&H Contracts and GPOs with HGV
|
| P-A-091
|
HGV Transactions from the 2016 to 2021 Crop Years
|
| P-A-092
|
Settlement Receipts of HGV with other Grain Elevators
|
| P-A-093
|
Email from Jeremy Krainik, dated November 29, 2019
|
| P-A-094
|
Spreadsheet Containing HGV Transactions Collected from Third Party Data
|
| P-R-095
|
Witness Statement of Mr. Timothy Duncan
|
| CB-A-096
|
Tim Duncan GPO, dated January 6, 2020 (Confidential – Level B)
|
| P-A-097
|
Tim Duncan GPO, dated January 6, 2020
|
| CB-A-098
|
Tim Duncan Fixed Price Contract, dated January 9, 2020 (Confidential – Level B)
|
| P-A-099
|
Tim Duncan Fixed Price Contract, dated January 9, 2020
|
| CA-A-100
|
Spreadsheet Containing Tim Duncan Transactions Collected from Third Party Data (Confidential – Level A)
|
| P-A-101
|
Spreadsheet Containing Tim Duncan Transactions Collected from Third Party Data
|
| CA-A-102
|
Tim Duncan Settlement Receipts for 2016 to 2020 (Confidential – Level A)
|
| P-A-103
|
Tim Duncan Settlement Receipts for 2016 to 2020
|
| P-R-104
|
Witness Statement of Mr. Edward Paull
|
| CB-A-105
|
Ed Paull’s P&H GPOs and Contracts (Confidential – Level B)
|
| P-A-106
|
Ed Paull’s P&H GPOs and Contracts
|
| CA-A-107
|
Ed Paull’s Spreadsheet Containing Deliveries from 2016-2019 (Confidential – Level A)
|
| P-A-108
|
Ed Paull’s Spreadsheet Containing Deliveries from 2016-2019
|
| CA-A-109
|
Ed Paull’s Receipts and Tickets with Third-Party Grain Companies (Confidential – Level A)
|
| P-A-110
|
Ed Paull’s Receipts and Tickets with Third-Party Grain Companies
|
| CB-A-111
|
Text Exchange between Mr. Paull and Mr. Klippenstein, dated April 2, 2018 (Confidential – Level B)
|
| P-A-112
|
Text Exchange between Mr. Paull and Mr. Klippenstein, dated April 2, 2018
|
| CA-A-113
|
Transaction Level Data for Mr. Paull Collected from Third-Party Companies (Confidential – Level A)
|
| P-A-114
|
Transaction Level Data for Mr. Paull Collected from Third-Party Companies
|
| CA-R-115
|
Witness Statement of Mr. John Heimbecker (Confidential – Level A)
|
| P-R-116
|
Witness Statement of Mr. John Heimbecker
|
| CA-R-117
|
P&H Protein Spreads up to October 6, 2020 (Confidential – Level A)
|
| P-R-118
|
P&H Protein Spreads up to October 6, 2020
|
| CA-R-119
|
Updated Table of Sales by Commissioner’s Farmer Witnesses to Virden and Moosomin Elevators to December 16, 2020 (Confidential – Level A)
|
| P-R-120
|
Updated Table of Sales by Commissioner’s Farmer Witnesses to Virden and Moosomin Elevators to December 16, 2020
|
| CA-R-121
|
Reply Witness Statement of Mr. John Heimbecker (Confidential – Level A)
|
| P-R-122
|
Reply Witness Statement of Mr. John Heimbecker
|
| P-A-123
|
Screenshot from the P&H Direct App Taken October 22, 2020
|
| CB-A-124
|
P&H PDQ Spot Pricing Information for Virden, January 2020 (Confidential – level B)
|
| P-A-125
|
P&H PDQ Spot Pricing Information for Virden, January 2020
|
| CB-A-126
|
P&H PDQ Deferred Pricing Information for Moosomin, April 2020 (Confidential – Level B)
|
| P-A-127
|
P&H PDQ Deferred Pricing Information for Moosomin, April 2020
|
| CB-A-128
|
P&H Contract with Tim Duncan (Confidential – Level B)
|
| P-A-129
|
P&H Contract with Tim Duncan
|
| CB-A-130
|
P&H Contract with Arrowdale Farms, dated April 18, 2018 (Confidential – Level B)
|
| P-A-131
|
P&H Contract with Arrowdale Farms, dated April 18, 2018
|
| CB-A-132
|
P&H PDQ Deferred Pricing Information (Confidential – Level B)
|
| P-A-133
|
P&H PDQ Deferred Pricing Information
|
| CB-A-134
|
Read-in Brief of the Commissioner of Competition (Confidential – Level B)
|
| P-A-135
|
Read-in Brief of the Commissioner of Competition
|
| CB-A-136
|
P&H Email Subject: Canola Basis, dated May 5, 2017 (Confidential – Level B)
|
| P-A-137
|
P&H Email Subject: Canola Basis, dated May 5, 2017
|
| CB-A-138
|
P&H Email with Alberta Wheat and Barley Subject: Bid Price, dated October 17, 2019 (Confidential – Level B)
|
| P-A-139
|
P&H Email with Alberta Wheat and Barley Subject: Bid Price, dated October 17, 2019
|
| P-A-140
|
Email from Mr. Hebert, dated March 21, 2018
|
| P-A-141
|
Canola Data from Exhibit 1 to Mr. Heimbecker’s Reply Witness Statement
|
| CB-A-142
|
P&H Email Subject: Canola Price Comp, dated January 5, 2018 at 12:10 pm (Confidential – Level B)
|
| P-A-143
|
P&H Email Subject: Canola Price Comp, dated January 5, 2018 at 12:10 pm
|
| CB-A-144
|
P&H Email Subject: Canola Price Comp, dated January 5, 2018 at 12:19 pm (Confidential – Level B)
|
| P-A-145
|
P&H Email Subject: Canola Price Comp, dated January 5, 2018 at 12:19 pm
|
| CB-A-146
|
Moosomin Purchases for January 5, 2018 (Confidential – Level B)
|
| P-A-147
|
Moosomin Purchases for January 5, 2018
|
| P-A-148
|
P&H Email Subject: Gain From Your Grain, dated December 13, 2018
|
| CB-A-149
|
P&H Email Subject: Gain From Your Grain, dated February 16, 2017 (Confidential – Level B)
|
| P-A-150
|
P&H Email Subject: Gain From Your Grain, dated February 16, 2017
|
| CB-A-151
|
P&H Email Subject: CWRS Protein Spreads, dated August 29, 2017 (Confidential – Level B)
|
| P-A-152
|
P&H Email Subject: CWRS Protein Spreads, dated August 29, 2017
|
| CB-A-153
|
P&H Email Subject: Competitors Protein Spreads, dated January 5, 2018 (Confidential – Level B)
|
| P-A-154
|
P&H Email Subject: Competitors Protein Spreads, dated January 5, 2018
|
| CB-A-155
|
Competitors Protein Spreads (Confidential – Level B)
|
| P-A-156
|
Competitors Protein Spreads
|
| CB-A-157
|
Notifiable Transaction Filing of P&H (Confidential – Level B)
|
| P-A-158
|
Notifiable Transaction Filing of P&H
|
| CB-A-159
|
P&H Basis Contract no. 157710 dated January 5, 2018 (Confidential – Level B)
|
| P-A-160
|
P&H Basis Contract no. 157710 dated January 5, 2018
|
| CB-A-161
|
P&H Fixed Price Contract no.157710-1 (Confidential – Level B)
|
| P-A-162
|
P&H Fixed Price Contract no.157710-1
|
| CB-A-163
|
P&H Basis Contract no. 157688 dated January 5, 2018 (Confidential – Level B)
|
| P-A-164
|
P&H Basis Contract no. 157688 dated January 5, 2018
|
| CB-A-165
|
P&H Fixed Price Contract no.157688-1 (Confidential – Level B)
|
| P-A-166
|
P&H Fixed Price Contract no.157688-1
|
| CB-R-167
|
P&H Shipped Tonnage Budget Fiscal 2021 (Confidential – Level B)
|
| P-R-168
|
P&H Shipped Tonnage Budget Fiscal 2021
|
| P-A-169
|
Expert report of Dr. Nathan Miller
|
| CA-A-170
|
Expert report of Dr. Nathan Miller (Confidential – Level A)
|
| CB-A-171
|
Expert report of Dr. Nathan Miller (Confidential – Level B)
|
| P-A-172
|
Reply expert report of Dr. Nathan Miller
|
| CA-A-173
|
Reply expert report of Dr. Nathan Miller (Confidential – Level A)
|
| CB-A-174
|
Reply expert report of Dr. Nathan Miller (Confidential – Level B)
|
| P-A-175
|
Qualifications of Dr. Nathan Miller
|
| P-A-176
|
Demonstrative of Dr. Nathan Miller
|
| CA-A-177
|
Demonstrative of Dr. Nathan Miller (Confidential – Level A)
|
| P-R-178
|
Joint List of Issues for the Concurrent Expert Session
|
| CA-R-179
|
Joint List of Issues for the Concurrent Expert Session (Confidential – Level A)
|
| P-R-180
|
Expert report of Ms. Margaret Sanderson
|
| CA-R-181
|
Expert report of Ms. Margaret Sanderson (Confidential – Level A)
|
| P-R-182
|
Slides of Ms. Margaret Sanderson
|
| CA-R-183
|
Slides of Ms. Margaret Sanderson (Confidential – Level A)
|
| CB-R-184
|
Slides of Ms. Margaret Sanderson (Confidential – Level B)
|
| P-R-185
|
Qualifications of Ms. Margaret Sanderson
|
| CA-A-186
|
USB Key Containing the Back-up Documents to the Expert Reports of Dr. Nathan Miller (Confidential – Level A)
|
| CA-R-187
|
USB Key Containing the Back-up Documents to the Expert Report of Ms. Margaret Sanderson (Confidential – Level A)
|
| CA-A-188
|
Moosomin Business Plan for Fiscal Year 2020 (Confidential – Level A)
|
| P-A-189
|
Moosomin Business Plan for Fiscal Year 2020
|
| CA-A-190
|
Moosomin 2017 Business Plan (Confidential – Level A)
|
| P-A-191
|
Moosomin 2017 Business Plan
|
| CA-A-192
|
Relevant Results from Ms. Sanderson’s HMT Calculations (Confidential – Level A)
|
| P-A-193
|
Relevant Results from Ms. Sanderson’s HMT Calculations
|
| P-R-194
|
By-Laws and Rules of the Minneapolis Grain Exchange
|
| P-A-195
|
Expert report of Mr. Andrew Harington
|
| CA-A-196
|
Expert report of Mr. Andrew Harington (Confidential – Level A)
|
| CB-A-197
|
Expert report of Mr. Andrew Harington (Confidential – Level B)
|
| CA-A-198
|
USB Key Containing the Back-up Documents to the Expert Report of Mr. Andrew Harington (Confidential – Level A)
|
| CA-A-199
|
Alistair Pethick Compendium (Confidential – Level A)
|
| P-A-200
|
Alistair Pethick Compendium
|
| CA-A-201
|
Ian Wagstaff Compendium (Confidential – Level A)
|
| P-A-202
|
Ian Wagstaff Compendium
|
| CA-A-203
|
Chris Lincoln Compendium (Confidential – Level A)
|
| P-A-204
|
Chris Lincoln Compendium
|
| CA-A-205
|
Ceres Compendium (Confidential – Level A)
|
| P-A-206
|
Ceres Compendium
|
| CA-A-207
|
Bunge Compendium (Confidential – Level A)
|
| P-A-208
|
Bunge Compendium
|
| CA-A-209
|
G3 Compendium (Confidential – Level A)
|
| P-A-210
|
G3 Compendium
|
| CA-A-211
|
Cargill Compendium (Confidential – Level A)
|
| P-A-212
|
Cargill Compendium
|
| CA-A-213
|
LDC Compendium (Confidential – Level A)
|
| P-A-214
|
LDC Compendium
|
| CA-A-215
|
Richardson Compendium (Confidential – Level A)
|
| P-A-216
|
Richardson Compendium
|
| CA-A-217
|
Viterra Compendium (Confidential – Level A)
|
| P-A-218
|
Viterra Compendium
|
| CA-A-219
|
Dr. Miller Compendium (Confidential – Level A)
|
| P-A-220
|
Dr. Miller Compendium
|
| P-A-221
|
Dr. Miller Public Documents Compendium
|
| CA-A-222
|
Additional Documents to Be Relied upon Compendium (Confidential – Level A)
|
| P-A-223
|
Additional Documents to Be Relied upon Compendium
|
| CA-A-224
|
ADM Purchase Contract (Confidential – Level A)
|
| P-A-225
|
ADM Purchase Contract
|
| CA-R-226
|
Compilation of Documents re Alistair Pethick (Confidential – Level A)
|
| P-R-227
|
Compilation of Documents re Alistair Pethick
|
| CA-R-228
|
Compilation of Documents re XXXXXXXXx (Confidential – Level A)
|
| P-R-229
|
Compilation of Documents re XXXXXXXXx
|
| CA-R-230
|
Compilation of Documents re Chris Lincoln (Confidential – Level A)
|
| P-R-231
|
Compilation of Documents re Chris Lincoln
|
| CA-R-232
|
Compilation of Documents re XXXXXXX (Confidential – Level A)
|
| P-R-233
|
Compilation of Documents re XXXXXXX
|
| CA-R-234
|
Compilation of Documents re Tim Duncan (Confidential – Level A)
|
| P-R-235
|
Compilation of Documents re Tim Duncan
|
| CA-R-236
|
Compilation of Documents re Ian Wagstaff (Confidential – Level A)
|
| P-R-237
|
Compilation of Documents re Ian Wagstaff
|
| CA-R-238
|
Compilation of Documents Listed in Exhibit B to Ms. Sanderson's Expert Report (Confidential – Level A)
|
| P-R-239
|
Compilation of Documents Listed in Exhibit B to Ms. Sanderson's Expert Report
|
| CA-R-240
|
Compilation of Miscellaneous Documents (Confidential – Level A)
|
| P-R-241
|
Compilation of Miscellaneous Documents
|
| CA-R-242
|
Compilation of Additional Documents Added to Agreed Book (Confidential – Level A)
|
| P-R-243
|
Compilation of Additional Documents Added to Agreed Book
|
| CA-A-244
|
Agreed Statement of Facts (Confidential – Level A)
|
| P-A-245
|
Agreed Statement of Facts
|
| CB-A-246
|
Agreed Statement of Facts Related to Two Farmers (Confidential – Level B)
|
| P-A-247
|
Agreed Statement of Facts Related to Two Farmers
|
| P-R-248
|
Map with All Farm Locations
|
| P-R-249
|
Map with Farmer Witnesses, Elevators and Crushers and Canadian Rail Lines
|
| P-R-250
|
Map with Farm Locations in Dr. Miller’s Geography
|
| P-R-251
|
Map with Farmer Witnesses and Elevators and Crushers
|
| P-R-252
|
Interactive Elevators and Crushers Map
|
| P-A-253
|
List of Elevators
|
Schedule “C” – Map representing farm locations as well as Elevators and Crushers in Dr. Miller’s geography
[1]
Where the words “Tribunal” or “panel” are used and the decision relates to a matter of law alone, that decision has been made solely by the judicial members of the Tribunal.
[2]
The Tribunal pauses to note that the Commissioner is not claiming that the Acquisition substantially prevents competition. Hence, in these Reasons, the Tribunal’s analysis will be limited to the Commissioner’s alleged substantial lessening of competition.
[3]
In his oral and written submissions, the Commissioner often refers to the Basis as the “basis price.”
[4]
In these Reasons, the word “MEGs” will also be used to refer to the Competition Bureau’s merger guidelines more generally.
[5]
Throughout these Reasons, all references to a SSNIP are meant to include a SSNDP when the context requires it.
[6]
To assist the reader of these Reasons, the Tribunal reproduces at Schedule “C” a map representing the location of farms and the location and identity of the Elevators and Crushers operating in Dr. Miller’s geography (Exhibit P-R-250, Map with Farm Locations in Dr. Miller’s Geography).
[7]
In the corridor of concern, a total of 80 farms sold wheat, 77 farms sold canola, and 82 farms sold either wheat or canola.
[8]
In a situation involving the purchase of a product and potential monopsony power, the determination to be made is whether prices are or likely would be materially lower than in the absence of the merger. In this discussion on the analytical framework, all references to “price increases” or “material price increases” are meant to relate to mergers involving the sale of a product and potential monopoly power. For mergers involving the purchase of a product and potential monopsony power, all references would be to “price decreases” or “material price decreases.”