1 PUBLIC CT-2010-010 THE COMPETITION TRIBUNAL
IN THE MATTER OF the Competition Act, R.S.C. 1985, c. C-34, as amended; AND IN THE MATTER OF an application by the Commissioner of Competition pursuant to section 76 of the Competition Act;
AND IN THE MATTER OF certain agreements or arrangements implemented or enforced by Visa Canada Corporation and MasterCard International Incorporated.
BETWEEN: THE COMMISSIONER OF COMPETITION
- and-
VISA CANADA CORPORATION and MASTERCARD INTERNATIONAL INCORPORATED
- and-
CANADIAN BANKERS ASSOCIATION and THE TORONTO-DOMINION BANK
Intervenors REPLY EXPERT REPORT OF MIKE MCCORMACK APRIL 23, 2012
Applicant
Respondents
2 PUBLIC I. Introduction 1. I previously filed a report in this matter on March 14, 2011 (the "March 14 Report"). 1 2. I have reviewed the responding expert reports (the "Responding Reports") and witness statements (the "Responding Witness Statements") filed by the Respondents and the Intervenors in this matter (collectively, the "Responding Materials"). As set out below, the Responding Materials do not alter the conclusions set out in my March 14 Report.
3. This report is not an exhaustive reply to every statement or conclusion in the Responding Materials with which I disagree. Below, I address the following claims made in the Responding Materials:
(a) Interchange Fees paid by MasterCard to its Issuers have been less than those Issuers' costs associated with issuing MasterCard credit cards;
(b) current Merchant Service Fees in Canada appear to be generally similar to or below those paid by Canadian merchants in 2003;
( c) merchants benefit from the issuance by MasterCard of premium and super-premium credit cards that carry even higher Merchant Service Fees because these credit cards induce cardholders to spend more at a given merchant and, relatedly, that the difference in the growth of the average transaction values for purchases made with MasterCard's premium credit cards and purchases made with MasterCard's core credit cards is attributable to "the additional products and services provided by the premium card issuing banks";
(d) Canadian Acquirers do not resell Visa or MasterCard Credit Card Network Services to merchants;
(e) merchants will harm consumers and the Respondents if permitted to surcharge; (f) surcharging in Australia has not steered consumers to debit; Defined terms herein have the meaning ascribed to them in my March 14 Report, unless otherwise indicated.
3 PUBLIC -2-
(g) co-branding and strategic alliance agreements reduce the card acceptance costs of participating merchants; and
(h) the cost of credit should be considered when comparing merchants' costs of credit card acceptance to the acceptance costs of cash or debit cards.
4. I also respond to certain comments and assertions regarding my March 14 Report made in the Witness Statement provided by Jeffery van Duynhoven on behalf of The Toronto-Dominion Bank ("TD").
5. The sources I rely upon in this reply report are found in the footnotes below and in Schedule A hereto.
II. Analysis (a) Interchange Fee Revenue is Only One Component of Issuer Revenue 6. In paragraphs 3(a) and 18 of his report submitted on behalf of MasterCard, Peter Dunn claims that "[ o ]ver the past several years, the interchange rates paid to MasterCard issuers in Canada have been less than the issuers' underlying cost of providing the guarantee, prompt funding, and delivery of the transaction". The relevance of this claim is unclear. If it is intended to suggest that Interchange Fees are set at a level below cost, this is incorrect and, in any event, MasterCard and Visa admit that Interchange Fees are not based on costs, but rather are intended to reflect the "value" of credit cards. For example, Michael Bradley (Head of Products for Visa in Canada) testified on discovery in this proceeding that Visa's pricing is based on value not cost:
"Q. . .. And am I right in assuming this from the documents we have just looked at, and I am happy to go to those, if you like, but that Visa's pricing today is not cost based; in other words, the legacy pricing I just took you to was a cost-based pricing model when Visa was a not-for-profit association but, in today's world, you don't sit down and say, 'We are going to price on a cost-plus basis'?
A. I would say cost is always a factor that's associated with how any business prices, including Visa, but no, we don't price per cost now.
PUBLIC 4 - 3 -
Q. You price on a value basis? A. Yes, we do." 2 7. To "illustrate" his claim, Mr. Dunn has produced a one page chart (at Exhibit "B" to his report) which purports to compare MasterCard's average (Canadian) effective interchange rate with the "component costs of [Canadian] Issuers (Losses, Processing & Financial)" in the period between January 1, 2007 and March 31, 2011. Mr. Dunn has not provided the data or any other information used by him to create the chart at Exhibit "B". I am therefore unable to review the underlying computations or data used by Mr. Dunn or to validate the claim his chart is meant to illustrate and support.
8. Leaving this quite serious problem to one side, Mr. Dunn fails to take into account the fact that Interchange Fees are only one source of the revenues earned by Issuers in connection with the issuance of MasterCard credit cards. More particularly, Mr. Dunn ignores the significant net interest income, annual fee income and "other income" (from other fees) earned by Issuers from their MasterCard-branded credit cards. 3 Mr. Dunn ought to have, but has not, quantified total Canadian MasterCard Issuer revenue (including interest, annual fee and other income) and compared the total of those streams of revenue to Issuer "losses and costs". Had he done so, the results would have confirmed (as explained at paragraphs 139 and 140 of my March 14 Report) that issuing general purpose credit cards has been and continues to be a lucrative business for Canadian Issuers.
9. For example, a document produced by TD in this proceeding indicates that in 2010 TD's -~: Visa and MasterCard credit card issuing business realize
2 See, e.g., Examination for Discovery of Michael Bradley on behalf of Visa Canada Corporation ("Bradley Examination"), December 8, 2011, pp. 1070-71, Qs. 3028-29.
Of course, Issuers of Visa credit cards also earn income (and profits) from these additional sources. 4 See
PUBLIC 5 -4-
(b) Canadian Merchant Service Fees Are Not Generally Similar to or Below Those Paid by Canadian Merchants in 2003 5 10. At paragraphs 3(b) and 20 to 25 of his report, Mr. Dunn asserts that average Merchant Service Fee levels in 2011 appear to be generally similar to or below those paid by Canadian merchants in 2003. That assertion is faulty for at least three reasons: (i) the Gadd survey data referred to and relied upon by Mr. Dunn for 2003 Merchant Service Fee levels is ambiguous, incomplete and not representative; (ii) the (alleged) CFIB data referred to and relied upon by Mr. Dunn for 2011 Merchant Service Fee levels is ambiguous, incomplete and not representative; and (iii) Mr. Dunn's assertion is at odds with the Respondents' own data and the experience of Canadian merchants, as reflected in numerous witness statements filed by the Commissioner of Competition (the "Commissioner") in this proceeding.
(i) The Gadd Survey Data is Ambiguous, Incomplete and Not Representative 11. The Gadd survey data was the product of telephone interviews with 642 "small sized [Canadian] merchants(< 50 employees) within certain retail, restaurant & service sectors ... ". 6 In my view, the size Gust 642 merchants) and scope (limited to small sized merchants) of the sample used by Gadd are inadequate; the small size and the narrow scope of the sample cast serious doubt on the reliability of Gadd's data and conclusions.
12. According to The Nilson Report, MasterCard and Visa each had approximately 600,000 merchant outlets in Canada in 2003. 7 Of these 600,000 merchant outlets, Gadd surveyed only a minute fraction. Further, the Gadd survey sample was focused on small merchants (i.e., those with less than 50 employees) within a small number of merchant segments (i.e., retail, restaurant and service sectors), and does not therefore appear to be representative of the actual sizes or
In his report, Mr. Dunn uses the term "Merchant Discount Rates", whereas I use the term "Merchant Service Fees".
6 "Methodology" slide in Gadd International Research Inc., Phase II - Detailed Report (October 2003), Exhibit "C" to Expert Report of Peter T. Dunn ("Dunn Report").
See The Nilson Report Number 811, May 2004. Note: Figures on the numbers of Canadian merchant businesses accepting Visa and/or MasterCard are not available. The industry data usually reports merchant acceptance data by merchant outlet, as opposed to merchant count.
6 PUBLIC - 5
types of Canadian merchants. Gadd also does not appear to have attempted to select a representative distribution of merchants by Canadian Acquirer, or by credit card sales volume.
13. Aside from the problem of unrepresentative sample data, the first of what the Gadd report describes as "Key Findings" is that there was a "lack of information and confusion in the marketplace regarding many aspects of credit card acceptance" (see Figure 1 below). This finding is reflected in merchant responses to questions regarding their Visa and/or MasterCard Merchant Service Fee levels. In particular, of the 642 merchants surveyed, 48% of those accepting only Visa credit cards did not know the Merchant Service Fees they were paying, while 50% of those accepting both Visa and MasterCard credit cards could not identify the Merchant Service Fees they were paying. 8 At most, the Gadd survey reflects the perceived, rather than actual, Merchant Service Fee levels of about one-half of the already limited number of survey respondents.
Figure 1 - "Key Findings" in Gadd Survey, Exhibit "C" to Dunn Report KEY FINDINGS > Phase I findings indicated that there was a lack of information and confusion in the marketplace regarding many aspects of credit card acceptance. > Quantitative findings confirm this - for example, among merchants who accept both MasterCard and VISA:
• Less than 10% aware MSPs set rate • Almost 1/2 unaware that rate is negotiable • 1/ 3 deal with two separate financial institutions instead of one
Gadd polled two sub-groups of merchants, those that accepted Visa only, and those that accepted both Visa and MasterCard credit cards. I could not locate a count of the merchants in each sub-group. See "Anatomy of Acceptance" and "Anatomy of Acceptance Gap" slides in Gadd Survey, Exhibit "C" to the Dunn Report.
7 PUBLIC -6-
(ii) The CFIB Data is Ambiguous, Incomplete and Not Representative 14. For the current 2011 Merchant Service Fee figures that appear in his report, Mr. Dunn relies on "merchant fee examples" (which likely exclude certain Acquirer, Issuer and Network fees) drawn from a preferred rate arrangement between a single Acquirer and a large merchant association. However, these "fee examples" are not accompanied with sufficient detail to determine effective Merchant Service Fee rates under the arrangement in question, let alone for Canadian merchants more generally.
15. More particularly, Mr. Dunn relies on a two-page, undated document (attached as Exhibit "D" to his report) which he says was published by the CFIB. This document provides "merchant fee examples ... based on preferred rates for CFIB members with [Acquirer] Chase ,_ '· Paymentech". 9 The document contains a table which lists and groups certain types of Canadian Issuer Visa and MasterCard credit cards into three Merchant Service Fee categories - "regular", "high/variably high" and "very high" - and provides Merchant Service Fee rate examples ranging from 1.65% (for Visa Classic, Gold and Platinum, which are categorized as having a "regular" Merchant Service Fee rate) to 2.71 % (for MasterCard Premium High Spend, which is categorized as having a "very high" Merchant Service Fee rate).
16. While Mr. Dunn claims that the "CFIB reported merchant rates of 1.68% for 'regular card' transactions and 2.22% for 'premium card' transactions", 10 those rates do not appear in the document attached to Exhibit "D" to Mr. Dunn's report. Nor is it possible to determine from that document the effective Merchant Service Fees paid by merchants under the CFIB-Chase Paymentech arrangement.
17. In any event, it is unclear that the Merchant Service Fee rate examples in Exhibit "D" are applicable or available to other Canadian merchants. Indeed, on its face, the document suggests they may not be. As already noted, Exhibit "D" states that the "merchant fee examples are based on preferred rates for CFIB members" [emphasis added] and, further, that Merchant Service Fee rates payable by merchants outside the CFIB-Chase Paymentech arrangement "may differ based on the card processor or specific agreement".
9 See Exhibit "D" to the Dunn Report. 10 See the Dunn Report, para. 23.
8 PUBLIC
- 7 -
18. It should also be noted that the "merchant fee examples" presented in the table at Exhibit "D" may not be inclusive of all transaction-related fees under the CFIB-Chase Paymentech arrangement. Information on the CFIB website indicates that merchants accepting "non-qualified" cards (e.g., commercial or premium cards) are also required under the arrangement to pay, among other fees, a 0.29% "Non-Qualified Transactions Processing Fee", an "assessment" network fee of 0.06% for Visa credit card transactions and of 0.064% for MasterCard credit card transactions and, on transactions made using a MasterCard credit card issued outside Canada, a cross-border/international service fee of 0.40% plus US$0.125. 11 (iii) The Respondents' Own Data Shows that Average Merchant Service Fee Rates are Significantly Higher Now Than They Were in 2003
19. Apart from the foregoing issues with the comparison made by Mr. Dunn, his opinion that Merchant Service Fees rates have not increased since 2003 is at odds with the Respondents' own data produced in this proceeding. As I discuss in my March 14 Report, the Respondents' documents show that Visa's and MasterCard's respective average Merchant Service Fee rates have increased over the last 15 years, and that they are significantly higher now than they were in 2003 (see Figure 2 below). 12 11 The CFIB document states "Non-Qualified Transactions will be downgraded by the Payment Brand, resulting in higher Interchange fees. For processing such Non-Qualified Transactions, Chase Paymentech Solutions will charge you: (a) the Merchant Discount Rate; (b) an additional fee (the 'Non-Qualifying Amount') equal to the difference between the Interchange fee associated with your Target Qualification Level and the Interchange fee for the Interchange level at which the Non-Qualified Transaction was actually processed; and (c) a Non-Qualified Transaction Processing Fee (which is calculated as a percentage of the amount of the Non-Qualified Transaction). The combined total of the Non-Qualifying Amount and the Non-Qualified Transaction Processing Fee will appear on your statement": see CFIB website, "CFIB, Payment processing tailored to the needs of CFIB members", undated, accessed online at http://www.cfib-fcei.ca/cfib_www_doc/dvaOOOI.pdf.
12 See March 14 Report, paras. 16(n) and 96-103, as well as Figure 12.
9 PUBLIC
- 8 -
Figure 2 - Visa and MasterCard Canadian Interchange & Acquirer Assessment Fees and TD Merchant Services Average Merchant Service Fee 13 20. The Respondents' data (reflected in Figure 2, above) is corroborated by the experience of Canadian merchants, as reflected in witness statements filed by the Commissioner in this proceeding. These witness statements confirm the significant increase in the Respondents' Merchant Service Fees since 2003. For example, in his Witness Statement, Tim Broughton, coowner of C'est What?, Inc. ("C'est What"), states that C'est What's effective cost of credit card
13 Visa, Acquirer Assessment Fee, March 10, 2008 (VISA00072957 at 2957), Payment Volume and Interchange Summary by Country and Cluster VISA00303213 at 3213-3214 Visa Inc. Interchan e Re orting, April 30, 2010 (VISA00312915 at 2916), Visa, Canadian Interchange and Product Plan, 13 July 2010 (VISA00067188 at 7197 to 7201), E-mail message from Richard Morrissey, April 19, 2010, "RE: Canadian Interchange Product Game Plan 041610 vl.docx" (VISA00502845 at 2845), Visa, "Canada Overview", September 30, 2009 (VISA00558283 at 8293), Visa Canada Interchange Summary September 20 I 0 (VISASUPP00006800), Microsoft Excel worksheet (MCW _ CCB _ 00038482).
10 PUBLIC
- 9
acceptance increased from- in 2004 to - the significant and increased card acceptance costs imposed by the Respondents' premium credit cards on his business (with "no additional value"):
"The higher card acceptance costs associated with premium credit cards impose even more significant costs on C'est What, with very little advantage or benefit, if any, to C'est What. The average rate for each premium credit card transaction is (the range is from - to-· Compared to the original - basic rate charged between 2004 and September 2008, this represents an increase of 62%, while yielding no additional value to C'est What. By December 2011, of all transactions (- of dollar volume) at C'est What occurred with a premium credit card." 15 21. Similarly, in her Witness Statement, Marion van Impe of the University of Saskatchewan discusses the significant increase since 2000 in the Merchant Service Fees paid by the University (and in its overall cost of credit card acceptance):
"As noted above, during the period 2000 to 2010, the University accepted both Visa and MasterCard credit cards for tuition payments, including payments made in-person and online. The decision to begin accepting credit cards for tuition payments in 2000 was made at a time when the University was trying to provide more online services. At that time, the University estimated the annual costs of accepting credit cards for tuition payments to be approximately $140,000. This estimated cost was based on a number of assumptions, including an average Merchant Service Fee of- and an acceptance rate of 20% (i.e. 20% of tuition paid by credit card).
However, in the ensuing 10 year period, the average Merchant Service Fee increased by about 20% from to approximately -· At the same time, the percentage of tuition paid by credit card increased to 42% in 2010. As a result of the increasing costs of credit card acceptance and increased use of credit cards, the overall cost to the University resulting from accepting credit cards for tuition payments rose from the original estimate of $140,000 in
14 See Witness Statement of Tim Broughton Statement ("Broughton Statement"), paras. 7 to 14. See also Witness Statement of Craig Daigle, para. 26 ("Daigle Statement").
15 Broughton Statement, para. 15.
in 2011. 14 Mr. Broughton also comments on
11 PUBLIC - 10 -
2000 to $900,000 in the 2009-2010 academic year, an increase of over 600%. This increase made the costs of accepting credit cards for tuition payments unsustainable for the University." 16 [emphasis added]
(c) Merchants Do Not Benefit from the Issuance by MasterCard of Premium and Super Premium Credit Cards that Carry Even Higher Merchant Service Fees
22. At paragraphs 3(c) and 26-30 of his report, Mr. Dunn claims that "one significant benefit merchants receive [from accepting MasterCard premium credit cards] is ... the increase in the average value of a purchase made by a consumer ('average ticket amount') on purchases made on MasterCard high spend and premium high spend cards when compared to the average ticket amount of purchases made on core (non-premium) cards" . 17 Mr. Dunn also asserts that the higher growth (between July-December 2008 and February-July 2011) in average purchase amounts on MasterCard high spend cards (15%), as compared to core cards (2%), can be explained by "the additional products and services provided by the premium card issuing banks", which he describes as "the primary impetus to the increased transaction values." 18 23. I disagree with Mr. Dunn on both points. 24. On the first point, despite the fact that this alleged "significant benefit" to merchants is one of the core justifications offered by MasterCard for the higher Interchange Fees and Merchant Service Fees charged on its premium credit cards, Mr. Dunn cites no evidence to support the claimed causal connection. In my view, there is no such connection.
25. There is no merit to the claim that premium credit cards cause consumers to spend more per transaction than they would if they were using a core credit card. Instead, the differences in average transaction amounts between core and premium credit cards is not attributable to the issuance of premium credit cards with higher Interchange Fees but, rather, to the fact that the premium credit cards are issued to more affluent cardholders who spend more.
16 Witness Statement of Marion van Impe, paras. 15-16. 17 See the Dunn Report, para. 26. 18 Ibid, para. 30.
12 PUBLIC
- 11 -
26. MasterCard premium credit cards are directed by MasterCard and its Issuers at affluent and higher-spending consumers, and MasterCard high spend and premium high spend cardholders are, on average, among the more affluent and higher-spending consumers in Canada. 19 To be issued a high spend card, which attracts higher Interchange Fees, a consumer must have a minimum annual individual income of $60,000 or household income of $100,000, or spend at least $15,000 a year on his card. 20 To be issued a premium high spend card, which attracts even higher Interchange Fees, a consumer must have a minimum annual individual income of $70,000 or household income of $120,000, or spend at least $20,000 a year on her card. 21 In contrast, and by necessary implication, consumers who hold core (non-premium) MasterCard credit cards are relatively less affluent and spend less on credit cards.
27. The claim that premium credit cards cause consumers to spend more per transaction than they would if they were using a core credit card is also undermined by the significant and widespread confusion among Canadian MasterCard cardholders about whether their MasterCard credit cards are, in fact, premium or core credit cards.
28. Similarly, the survey results generated by Benoit Gauthier for this case at the request of MasterCard and Visa emphasize how widespread and significant the confusion among MasterCard cardholders is. According to Mr. Gauthier's report, 51 % of MasterCard credit cardholders who responded to his survey indicated that they hold a premium MasterCard credit card. Mr. Gauthier observes that "[t]he identification of premium MasterCards is not as straightforward [as for Visa credit cards]. MasterCard holders had to identify the issuing institution first (page C-15), then a list of premium card trademark names relevant to the issuing institution were listed to the respondent. Using this method, 51% of MasterCard holders were
19 2 210--22
13 PUBLIC - 12
identified as having a premium card". 23 The conclusion that 51 % of MasterCard cardholders hold premium credit cards casts doubt on the accuracy of Mr. Gauthier's entire survey.
29. The Respondents and their experts have not provided any evidence to suggest that having a premium credit card can induce a MasterCard cardholder to spend more. Rather, the higher average ticket amount for premium card purchases is the product of the fact that the Respondents have issued premium credit cards to more affluent, higher-spending cardholders who spend more on average, per transaction, than less affluent, lower-spending core cardholders. 25 30. On the second point (namely, that rewards and other benefits offered by Issuers are "the primary impetus to the increased transaction values" between 2008 and 2011), I am not aware of any evidence, and (again) none has been cited by Mr. Dunn, that rewards cause consumers to spend more per transaction than they otherwise would in the absence of such rewards or benefits. At best, rewards and benefits may cause affluent, higher-spending consumers to re-allocate the same overall level of spending as amongst different payment methods and, in particular, to consolidate spending on premium credit cards that offer rewards or other benefits. MasterCard
31. The conclusion that there is no causal link between rewards/benefits and increased transaction values is further supported by the fact that at least until August 2010, when the Code
23 See Expert Report of Benoit Gauthier, pp. 12-13. See also p. 4: "In the MasterCard consumer family, several brand names are used for premium cards, depending on the issuer. Therefore, the questionnaire includes eight questions ... to identify respondents with premium cards" [emphasis added].
24
25 See Bradley Examination, December 6, 2011, pp. 373-75, Qs. 1118-25. 26
14 PUBLIC - 13 -
27 28 29 30
15 PUBLIC
- 14 --33. (d) Acquirers Resell Visa and MasterCard Credit Card Network Services to Merchants 34. Several of the Responding Witness Statements contend, contrary to paragraphs 16(g) and 158-163 of my March 14 Report, that Canadian Acquirers do not resell Visa's or MasterCard's Credit Card Network Services to merchants. 32 35. The principal arguments made iri the Responding Witness Statements are as follows: (i) Acquirers do not resell "the same set of services" supplied by Visa and MasterCard because Acquirers also supply additional credit card acceptance-related services (such as the "leasing and selling point-of-sale equipment" and the provision of "ongoing training, service and support of equipment and sales staff' 33 ); and (ii) merchants do not have access to the Visa and MasterCard networks.
36. I disagree with both of these arguments. 37. On the first argument, the ancillary and/or additional services that may also be supplied by Acquirers to merchants who wish to accept Visa and/or MasterCard credit cards cannot obscure the fact that the central and fundamental services supplied by Acquirers to their
31 van Du nhoven ("van Duynhoven Statement"), paras. 38-39, -Witness Statement of Karen Legget, para. 38 and Witness Statement of Brian Weiner, para. 39.
33 van Duynhoven Statement, para. 39.
16 PUBLIC - 15
merchant customers is access to Visa's and/or MasterCard's authorization network and clearing and settlement system in order to allow merchants to accept credit cards as a form of payment.
38. In paragraph 159 of my March 14 Report, I stated that "there would be little, if any, point in supplying (and little, if any demand, for) other ancillary services also typically provided by Acquirers to merchants" in the event that Acquirers ceased supplying Credit Card Network Services. At paragraph 71 of his Statement, Mr. van Duynhoven argues that my statement is incorrect because approximately. (or-of TD's roughly (extrapolating from Mr. van Duynhoven's figures) ~erchant customers 34 accept only Interac debit cards. Mr. van Duynhoven suggests that because those merchants use "other ancillary services" provided by TD this somehow contradicts or diminishes my statement. However, Mr. van Duynhoven's response misses the point (and the statistic provided by him is not inconsistent with, and certainly does not contradict, what I said in my March 14 Report). 35 My point was simply that "ancillary services" supplied by Acquirers (including TD) in connection with the acceptance by merchants of credit cards would be of little or no utility to merchants if those merchants could not actually accept Credit Card Network Services, and therefore not access to the Respondents' respective networks. In other words, if a merchant could not actually accept credit cards (because Acquirers did not supply Credit Card Network Services) there would be little point in that merchant receiving "ancillary services" in connection with the acceptance by that merchant of credit cards. The fact that there is a very small number of TD's merchant customers who do not accept credit cards and who are supplied by TD with "ancillary services" in connection with the acceptance of lnterac debit cards is irrelevant.
39. The second argument in the Responding Witness Statements, that merchants do not connect to the Visa and MasterCard network, is also incorrect. There would be no point in establishing a credit card network if merchants were unable to connect to that network. Access to
34 35 See also paragraph 163 of my March 14 Report.
17 PUBLIC - 16
the Visa or MasterCard networks is required for merchants to be able to receive Credit Card Network Services and to accept Visa or MasterCard credit cards. 36 (e) Merchants Will Not Harm Consumers or the Respondents if Permitted to Surcharge
40. In the Responding Materials, Visa and MasterCard argue in support of the No-Surcharge Rules, claiming that, if permitted to surcharge, merchants will engage in a "bait and switch" or excessive surcharging.
41. For example, in his report, Professor Church states that "[m]erchants, especially merchants that do not expect repeat business, have an incentive to engage in hold up. That is, they have an incentive to free ride on the investments made by the card network and merchants that abide by the Rules (or their equivalent)." 37 Similarly, Professor Elzinga maintains that the , Visa and MasterCard No-Surcharge Rules are necessary to protect the value of the networks' brands because they prevent free riding, baiting and switching, and misleading tactics by merchants "that would penalize cardholders". 38 42. However, experience to date in jurisdictions where surcharging is permitted by Visa and MasterCard contradicts these claims.
43. The validity and seriousness of the claims made by the Visa and MasterCard experts should be assessed in light of the evidence that emerged during the discovery process in this proceeding which establishes that Visa and MasterCard do not systematically track or assemble information about bait and switching, "excessive surcharging" or the impact on their respective networks in jurisdictions where surcharging is permitted.
36 See "Integrated Pricing Policy", February 10, 2010 (VISA00149319 at 9320): "Acquirers enable merchants to interact with the network and allow merchants to accept different types of electronic payments ... The merchant will pay the acquirer a merchant discount rate (l\.1DR) in exchange for the ability to accept payments cards and connect to the various networks." See also "Visa Business Overview", February 2008 (VISA00185983 at 5987): "Fundamentally, we are a network company. In fact, we are the largest payments network company in the world. We connect thousands of financial institutions, millions of merchants, and hundreds of millions of cardholders. These network partners use our products and our processing platform to conduct secure, convenient and profitable electronic payments."
37 See Expert Report of Jeffery Church, paras. 9, 51-53. 38 See Expert Report of Kenneth G. Elzinga ("Elzinga Report"), paras. 227-30. See also Dunn Report, paras. 31-34, and paras. 45-65.
18 PUBLIC - 17 -
44. These claims should also be assessed in light of the fact that Visa and MasterCard permit surcharging in the form of "convenience fees". As discussed at paragraph 170 of my March 14 Report,
there is little or no evidence of "baiting and switching" or of other misleading tactics engaged in by merchants charging convenience fees in Canada or the United States. 41 In the Responding Materials, there is no reference to or evidence of misleading merchant tactics engaged in by Canadian or U.S. merchants that apply convenience fees to MasterCard or Visa credit card transactions.
45. In Australia, the evidence is that the Visa and MasterCard networks have continued to increase their transaction volumes even after the removal of the rule against surcharging by merchants in 2003. Since that date, the number of Visa and MasterCard credit cards in circulation has increased, 42 merchant acceptance of Visa and MasterCard credit cards has increased, 43 transaction volumes and values on the Visa and MasterCard networks have 39 40 See also, e.g., "Re: Surcharging", October 22, 2007 (VISASUPP00006506); "Re: Convenience Fee - Visa", February 5, 2008 (VISA00351283); Proposed Operating Regulations: Convenience Fees in Canada, January 29, 2008 (VISA00056802); Visa Operating Regulations Committee Meeting Minutes, April 21, 2008 (VISASUPP00006525 at 6526); Visa "Convenience Fees - Canada" January 13, 2009, (VISA00101271); and Canada Regional Operating Regulations: Convenience Fees, September 22, 2009 (VISA00468701 ).
41 42 See The Reserve Bank of Australia website, "Payments Data, Schedule Cl - Credit and Charge Card Statistics, Additional Credit Card Statistics", accessed at http://www.rba.gov.au/payments-system/resources/statistics/index.html.
43
19 PUBLIC - 18 -
44 increased and the Respondents' revenues and profits from their credit card operations have increased. 45 46 During this same period, the number of merchants surcharging in Australia has increased. 47 (f) Surcharging in Australia has Steered Consumers to Debit 47. In her Witness Statement, Elizabeth Buse asserts that in Australia "[s]urcharging has not steered customers to debit" and that she has "seen no evidence to indicate that the 2003 reforms led to increased debit use". 48 48. These assertions are contradicted by the statistics maintained by the Reserve Bank of Australia with respect to Australian debit card purchase transaction volumes and values. 49 The RBA statistics indicate that Australian debit card purchase values (AUD$) grew between 2003 and 2011 at an average annual rate of 14.5%, versus Australian credit card purchase values which grew at an annual rate of 7.9% over the same period (see Figure 3). 50 The RBA statistics also indicate that debit card purchase transaction volumes have grown between 2003 and 2011 at
44 See "Visa Volume - Australia and New Zealand" (VISA00579879) and Naffah Deposition, pp. 67-68 F: (MCW _ CCB _ 00166900 at 166918). The number and value of credit card transactions in Australia increased 4.5% and 5.5% respectively on average annually over the 2006/07 to 2010/11 period: "Payments System Board Annual Report" (GSSS5893 _00003027 at 3039).
45 See "Australia Revenues and Ex enses FY2008 to FY2011" (VISASUPP00007583) and 46 47 See RBA Strategic Review of Innovation in the Payments System: The Results of the RBA's 2010 Consumer Payments Use Study (see GSSS5893_00000337 at 0354).
48 See Witness Statement of Elizabeth Buse, para. 32. 49 See Reserve Bank of Australia website "Payments Data, Schedule Cl - Credit and Charge Card Statistics", and "Payments Data, Schedule CS - Debit Card http://www.rba.gov.au/payments-system/resources/statistics/index.html.
50 See ibid and March 14 Report, para. 167.
Statistics", accessed online April 16, 2012 at
20 PUBLIC - 19
an average annual rate of 14.4%, in contrast to the volume of Australian credit card purchase transactions which grew at an average annual rate of 6.1%. Figure 3-Australia Debit Card Purchase Volumes (2003-2011) AUD 160,000,000 AUD 140,000,000 AUD 120,000,000 AUD 100,000,000 AUD 80,000,000 __....-L.--/ AUD 60,000,000 ,,,,,_,-AUD40,000,000 ---AUD 20,000,000 AUDO 2003 2004 2005 2006 (g) Co-Branding and Strategic Alliance Agreements are Not an Effective Way to Reduce Card Acceptance Costs
49. It is suggested in certain of the Responding Witness Statements that co-branding and strategic alliance agreements are an effective way for merchants to reduce their card acceptance costs. 52 .. 50. I disagree for two reasons. 51 See ibid. William Sheedy makes a similar assertion in his Witness Statement: "Use of co-brand card programs by merchants is another effective means of influencing consumer's brand preference and managing merchant payment acceptance costs.": see Witness Statement of William Sheedy, para. 73.
51 ,;' /" ~ .JI' v / 2007 2008 2009 2010 2011
21 PUBLIC - 20 -
51. First, co-branding and/or entering into strategic alliances is a strategy available to a very 52. Second, and more importantly, even for larger merchants, co-branding is not effective in materially decreasing their cost of credit card acceptance.
-·55 53. However, the MBNA co-brand agreement has had little, if any, effect on the Merchant Service Fees paid by Shoppers. As explained by Mr. Daigle, the MBNA Shoppers Optimum
56 - Further, and more importantly: "Shoppers' [Merchant Service] Fees for Visa and MasterCard increased from in 2007 t in 2011, an increase 0£9 -or-' and "Shoppers' average [Merchant Service] Fee increased from -in 2007 to - in 2011 for Visa credit card transactions and from-in 2007 to-in 2011 for MasterCard credit card transactions". 57 f:
(h) It is Inappropriate to Include the Cost of Lending in Credit Card Acceptance Costs 54. In his report, Professor Elzinga asserts that cost comparisons between credit card, debit card, and cash transactions "are difficult to make because they are not apples to apples." 58 53 54 55 56 57 Daigle Statement, para. 26. 58 See Elzinga Report, para. 111.
22 PUBLIC
- 21 -
Professor Elzinga maintains that "in order to make the comparison more like apples to apples, one should subtract lending costs from the cost of credit cards, because this is a feature of credit cards not found with cash or debit. 1159 55. I disagree. 56. This argument ignores the significant interest rates and other fees charged by Issuers to cardholders, in addition to the Interchange Fees paid by merchants to Issuers, to offset the cost of any credit function utilized by cardholders. 60 Indeed, interest income is typically the single largest source of revenues for Issuers. As noted above, a document produced by TD in this proceeding indicates that in 2010 TD's Visa and MasterCard credit issuing business realized
57. Further, Professor Elzinga's argument presumes that cardholders are relying upon the credit function in making a purchase from a merchant. In fact, nearly 70% of Canadians are non-revolvers, or "transactors", that do not carry a balance on their credit cards. 62 (i) Response to Certain Comments and Assertions in the van Duynhoven Statement 58. In his Statement, Mr. van Duynhoven makes several comments and assertions regarding my March 14 Report. Those comments and assertions have two broad themes. First, Mr. van Duynhoven claims that I have misdescribed or misunderstood aspects of the acquiring business or of the Visa Operating Rules. Second, he claims that if the remedy sought by the Commissioner is granted, it will not be effective. For the reasons set out below, I disagree with Mr. van Duynhoven's comments and assertions.
59 See Elzinga Report, para. 112. 61 See ibid. 62 See Canada Region Advisory Board Internal Participant Fact Sheet, undated (MCW_ CCB _ 00033058). See also "Credit Cards: Statistics and Facts - April 2011" (CBA000002 at 0003) which states that "[a] 2010 survey by The Strategic Counsel found that 65% of Canadians pay their balance off in full every month".
23 PUBLIC - 22 -
(i) No Misunderstanding or Misdescription of the Acquiring Business or the Visa Operating Rules
59. At paragraph 38 of his Statement, Mr. van Duynhoven says that I have inaccurately described and/or oversimplified the acquiring business in Canada in paragraphs 46 and 47 and Figures 4 and 5 of my March 14 Report because I did not mention "payment processors". It is unclear if by the term "payment processors" Mr. van Duynhoven is referring to non-bank Acquirers, such as First Data, or to third party processors that provide technology services to Acquirers. In any event, Mr. van Duynhoven fails to offer any basis why the presence of payment processors is material to the conclusions set out in my March 14 Report.
60. Paragraphs 46 and 47 of my March 14 Report were intended to (and do) provide a general description of three and four party payment card systems in Canada. The fact that Canadian Acquirers may have service relationships with third party processors, gateways or POS technology companies does not affect the definitions or illustrations of three and four party payment card systems. In any event, I do reflect several non-financial institution Acquirers in Figure 4 of my March 14 Report, and discuss financial institution sponsorship of non-financial institution Acquirers in paragraphs 70 and 72 of that Report.
61. In paragraph 64(a) of his Statement, Mr. van Duynhoven argues that paragraph 91 and Figure 10 of my March 14 Report are inconsistent with Canadian Acquirers' normal practice of crediting the merchant the full amount of settlement for a payment card transaction, and then charging the applicable Merchant Service Fees at a later stage. Again, it is not clear why the presence of this billing practice is at all relevant to my conclusions. In any event, I am familiar with the Merchant Service Fee billing practice Mr. van Duynhoven describes. Paragraph 91 and Figure 10 of my March 14 Report are intended to illustrate the flow of a payment card transaction, and the net effects of settlement processes between and among the various participants.
62. In paragraph 151 of his Statement, Mr. van Duynhoven claims that paragraph 168 of my March 14 Report mischaracterizes Visa's Operating Rules as permitting Canadian merchants to surcharge through convenience fees. Mr. van Duyhnhoven's claim misstates my evidence. I do not state or suggest in paragraph 168 of my March 14 Report that the Visa Operating Rules now
24 PUBLIC - 23 -
permit, or have ever permitted, surcharging in the form of convenience fees in Canada (although I understand that Visa allows surcharges to be applied to credit card usage in some circumstances in Canada, such as payments for parking tickets by credit card). 63 63. Further, I specifically discuss Visa's convenience fee policy (or lack thereof) in paragraph 170 of my March 14 Report where I state:
"Visa's Operating Rules permit surcharging by some types of U.S. merchants using convenience fees, but Visa has not yet modified its rules to allow Canadian merchants to surcharge.
(ii) If Granted, the Remedy Requested by the Commissioner will be Effective 64. At paragraphs 158 and 159 of his Statement, Mr. van Duynhoven asserts that with "Account Level Processing" ("ALP") of transactions "it is simply not possible for an Acquirer to determine the Default Interchange Rate that applies to a given transaction without having access to the Cardholder's related account information from Visa/MasterCard". Mr. van Duynhoven fails to explain why receiving access to such information would present a problem for Acquirers. Indeed, Visa and MasterCard can provide, and in fact have provided, Acquirers with precisely such access in the United States. As discussed in paragraphs 183 and 184 of my March 14 Report, both MasterCard and Visa have announced and made available to U.S. Acquirers "product inquiry services", which Acquirers and merchants may use to determine a particular card's product type and associated interchange rate at the time of authorization, notwithstanding that Visa's and MasterCard's respective ALP processes have been in use by U.S. Issuers since 2007. 64 63 See, e.g., Bradley Examination, December 8, 2011, pp. 803-805, Q. 2298-99; Proposed Operating Regulations: Convenience Fees in Canada, January 30, 2008 (VISA00020685 at 686); "Re: Acceptance Compliance Program Incident Notification - Newmarket Hydro", June 2, 2010 (VISA00316956).
64 See Digital Transactions News, "As Account-Level Processing Takes Shape, Acquirer Impact Unclear," June 5, 2007, accessed online at http://www.digitaltransactions.net/news/story/1396.
25 PUBLIC - 24 -
65. At paragraph 159 of his Statement, Mr. van Duynhoven asserts that "Canadian Acquirers would not have the technical ability to reliably provide the type of information described in the McCormack report to their Merchant Customers" and that in reaching a contrary conclusion I did not consider that Canadian credit cards are "chip" enabled, with the result that cardholder data is stored in an encrypted form on the chip on each card. Mr. van Duynhoven claims that because ~ ' : . cardholder data is stored in an encrypted form "the encrypted information would need to be sent from the merchant [POS system] to TD's network and back . . . and may cause significant delays".
66. This assertion is incorrect. In either a Visa or MasterCard chip card transaction, following a series of authentication steps, the merchant's POS terminal retrieves and decrypts the cardholder information from the card chip, including the cardholder account number, prior to sending the data to the Acquirer for authorization. 65 Assuming the POS terminal, chip, and cardholder authentication process(e s) succeed, the cardholder account number is accessible to the merchant's POS system in an unencrypted form.
67. At paragraph 161 of his Statement, Mr. van Duynhoven maintains that "if merchants chose to surcharge, they will do so with a blended rate which will not deliver the benefits identified by McCormack in paragraphs 16(v) and 179 of his report." However, Mr. van Duynhoven overlooks the fact that those merchants with combined/blended single MasterCard and Visa Merchant Service Fee rates may request and/or negotiate to have the rates quoted separately by card brand. TD presently provides even small merchants with separate Visa and MasterCard rate quotes. A January 24, 2011 letter from TDMS to the Canadian Chamber of Commerce (the "Chamber") introducing a new "Preferred Rate program" between TDMS and the Chamber quotes a base Visa credit rate of 1.50% and a base MasterCard rate of 1.64% (exclusive of interchange rate and differential charges for premium and commercial cards). 66 65 See Visa website, "Visa Transaction Acceptance Device Guide, version 2.0", March 2011, pages 25, 62-65, accessed online at https://technologypartner.visa.com/Library/Specifications.aspx#42.
66 See letter from TD Merchant Services to Canadian Chamber of Commerce Executives, January 24, 2011, accessed online at http://www.kawarthachamber.ca/files/edocs/TDMS%200ffer%20Letter.pdf.
26 PUBLIC - 25 -
68. At paragraph 169 of his Statement, Mr. van Duynhoven takes issue with paragraph 150 of my March 14 Report insofar as I assert that "[i]f Visa and MasterCard credit card Interchange Fees are reduced, Acquirers will reduce Merchant Service Fees." Mr. van Duynhoven asserts that if a reduction in credit card Interchange Fees occurred, the impact on Merchant Service Fee levels would depend on the terms of a merchant's contract with its Acquirer and (when the r· existing contract expires) the level of competition in the marketplace among Acquirers.
69. In the event of decreases in Canadian credit card Interchange Fee levels, I agree that some merchants would likely need to proactively renegotiate with Acquirers for Merchant
The converse is also true; reductions in Interchange Fees will be passed on by Acquirers in the form of lower Merchant Service Fees.
As such, to remain competitive, Acquirers will be required to pass on decreases in Interchange Fees in the form of lower Card Acceptance Fees for merchants or risk losing merchant customers to rival Acquirers.
70. The evidence from Australia supports the conclusion that if credit card Interchange Fees are reduced in Canada, Acquirers will aggressively compete for merchant Credit Card Network Services business through immediate reductions in Merchant Service Fees. 69 Two months after the RBA-mandated Interchange Fee reductions took effect in Australia in November 2003, average Australian Merchant Service Fees had decreased by approximately 0.26%, from 1.44% to 1.18%. 70 By December 2011, Australian merchants' average Merchant Service Fees for Visa and MasterCard credit cards had decreased by approximately 40.7%, from 1.45% to 0.86% (a
67 68 See also March 14 Report, para. 147: "Canadian Acquirers do compete vigorously based on the small portion of Merchant Service Fees that is retained by Acquirers as service fees (the Acquirers' margin)."
69 See March 14 Report, para. 173. 70 See Reserve Bank of Australia website, "Payments Data, Schedule C3 - Merchant Fees for Credit and Charge Cards", accessed at http://www.rba.gov.au/payments-system/resources/statistics/index.html.
27 PUBLIC
- 26-total of-0.59), which is 0.14% more than the RBA-mandated Visa and MasterCard interchange rate decreases of 0.45%. 71 71. At paragraph 172 of his Statement, Mr. van Duynhoven states that "[i]n the U.S., the largest merchants benefited substantially from the caps on debit card Interchange Fees. In contrast, small-mid size merchants saw very little impact on the fees they were charged by Acquirers." However, the evidence demonstrates otherwise (notwithstanding that Visa and MasterCard debit card Interchange Fee reductions only took effect on October 1, 2011). For example, U.S. Acquirer Heartland Payment Systems ("Heartland"), the 9th largest U.S. Acquirer with 245,420 active merchant outlets, announced on March 6, 2012 that it had delivered " ... More Than [US]$100 Million in Durbin Debit Reform Savings to Merchants" through reductions in merchants' debit card service fees. 72 Heartland is one of the leading providers of Credit Card Network Services in the U.S. to small and medium sized businesses, classifying 171,801 thousand of the merchant locations it provides services to as "active SME bankcard merchants located across the United States" .7 3 April 23, 2012 Mike McCormack
71 See ibid. 72 See Heartland Payment Systems website, Press Releases, "Heartland Payment Systems(R) Delivers More Than $100 Million in Durbin Debit Reform Savings to Merchants", March 6, 2012, accessed online at http://www.snl.com/irweblinkx/file.aspx? llD=40944l7&FID=12855643.
73 See Heartland Payment Systems Form 10-K for Fiscal Year Ended December 31, 2011, page 1, accessed online at . http://www2.snl.com/Cache/13I01307.PDF?D=&O=PDF&IID=4094417&OSID=9&Y=&T=&FlD=13101307, and The Nilson Report Number 990, March 2012.
----------------
28 PUBLIC
Schedule A-Materials Relied Upon CFIB, "Payment processing tailored to the needs of CFIB members", undated Digital Transactions News, "As Account-Level Processing Takes Shape, Acquirer Impact Unclear," June 5, 2007
Expert Report of Mike McCormack dated March 14, 2012 Heartland Payment Systems Press Release, "Heartland Payment Systems(R) Delivers More Than $100 Million in Durbin Debit Reform Savings to Merchants", March 6, 2012
Heartland Payment Systems Form 10-K for Fiscal Year Ended December 31, 2011 Letter from TD Merchant Services to Canadian Chamber of Commerce Executives, January 24, 2011
Reserve Bank of Australia, Payments Data, Schedule Cl - Credit and Charge Card Statistics, Additional Credit Card Statistics
Reserve Bank of Australia, Payments Data, Schedule C3 - Merchant Fees for Credit and Charge Cards
Reserve Bank of Australia, Payments Data, Schedule C5 - Debit Card Statistics The Nilson Report Number 811, May 2004 The Nilson Report Number 990, March 2012
Transcripts of the Examination for Discovery of Michael Bradley, December 5-9, 2011, and answers to undertakings
Visa Transaction Acceptance Device Guide, version 2.0, March 2011
IN THE MATTER OF the Competition Act, R.S.C. 1985, c. C-34, as amended;
AND IN THE MATTER OF an application by the Commissioner of Competition pursuant to section 76 of the Competition Act;
AND IN THE MATTER OF certain agreements or arrangements implemented or enforced by Visa Canada Corporation and MasterCard International Incorporated.
BETWEEN: THE COMMISSIONER OF COMPETITION Applicant -and-
MASTERCARD INTERNATIONAL INCORPORATED
Respondents - and -THE TORONTO-DOMINION BANK THE CANADIAN BANKERS ASSOCIATION
Intervenors
REPLY EXPERT REPORT OF MIKE MCCORMACK APRIL 23, 2012
Davies Ward Phillips & Vineberg LLP Suite 4400, 1 First Canadian Place Toronto, Ontario M5X lBl
Kent E. Thomson (LSUC #24264J) Adam Fanaki (LSUC #38208L) Davit D. Akman (LSUC #44274R) Tel: 416.863.0900 Fax: 416.863.0871
Department of Justice Canada Competition Bureau Legal Services Place du Portage, Phase I 50 Victoria Street, 22nd Floor Gatineau QC KIA OC9
William Miller (LSUC #14443V) Tel: 819.953.3903 Fax: 819.953.9267
Counsel to the Commissioner of Competition
PUBLIC CT-2010-010 THE COMPETITION TRIBUNAL
VISA CANADA CORPORATION and