File No. CT-98/2 THE COMPETITION TRIBUNAL IN THE MATTER OF THE COMPETITION ACT, R.S.C. 1985, c.C-34, as amended, and the Competition Tribunal Rules SOR/94-290, as amended (the "Rules");
AND IN THE MATTER OF an inquiry pursuant to subsection 0 10( 1) (b) of the Competition Act relating to the proposed acquisition of ICG Propane Inc. by Superior Propane Inc. AND IN THE MATTER OF an application by the Commissioner of Competition for orders pursuant to s. 92 and other provisions of the Competition Act. u It· BETWEEN: 0 THE COMMISSIONER OF COMPETITION Q Applicant 0 -and -D SUPERIOR PROPANE INC. and ICG PROPANE INC. Respondents D 0 REBUTTAL AFFIDAVIT OF RICHARD SCHWINDT, STEVEN GLOBERMAN AND TERRY KEMP TO THE AFFIDAVIT OF COLIN O'LEARY AND ERIC FERGIN 0 0 William J. Miller Jo'Anne Strekaf Steve Robertson u Jennifer Quaid Jamie Bocking Counsel to the Commissioner of Competition c/o Department of Justice 0 Place du Portage, Phase I 50 Victoria Street Hull, Quebec K1A OC9 0 Telephone: (819) 997-3325 Fax: (819) 953-9267 ,~ J
0 1 D File No. CT-98/2 u THE COMPETITION TRIBUNAL u IN THE MATTER OF THE COMPETITION ACT, R.S.C. 1985, c.C-34, as amended, and the Competition Tribunal u Rules SOR/94-290, as amended (the "Rules"); AND IN THE MATTER OF an inquiry pursuant to subsection 0 10(1)(b) of the Competition Act relating to the proposed acquisition of ICG Propane Inc. by Superior Propane Inc. u· AND IN THE MATTER OF an Application by the Commissioner of Competition for an order pursuant to section 92 of the Competition Act
D
BETWEEN: 0 THE COMMISSIONER OF COMPETITION D Applicant ·and· 0 SUPERIOR PROPANE INC. and ICG PROPANE INC. u Respondents D REBUTTAL AFFIDAVIT OF RICHARD SCHWINDT, STEVEN GLOBERMAN AND TERRY KEMP u I, Richard Schwindt, of Mt. Lehman, British Columbia, I, Steven Globerman of a Bellingham, Washington, United States of America; and I, Terry Kemp, of Calgary, Alberta, MAKE OATH AND SAY THAT: ll 1. We were retained by the Commissioner of Competition to provide expert economic evidence in this matter, to review the expert reports submitted by the 0 expert witnesses of the Respondents, and to provide our comments with respect to those reports. u 1]
0 -2-0 2. We have reviewed the expert witness affidavit of Colin O'Leary and Eric Fergin 0 filed herein. Attached hereto and marked as Exhibit "A" is a true copy of our comments with respect to that Affidavit. The contents of Exhibit "A" and the findings and opinions expressed therein are true to the best of our knowledge, 0 information and belief. 0 3. Attached hereto and marked as Exhibit "B" are true copies of our curricula vitae. 0 4. We make this affidavit pursuant to Rule 47(2) of the Competition Tribunal Rules. SWORN BEFORE ME ) 0 AT ) THIS DAY OF September 1999 ) 0 RICHARD SCHWINDT 0 BARRISTER AND SOLICITOR 0 SWORN BEFORE ME ) D AT ) THIS _ DAY OF September 1999 ) D STEVEN GLOBERMAN a BARRISTER AND SOLICITOR 0 a SWORN BEFORE ME AT C~C.),.ltV /~A a THIS ll/ DAY OF September 1999 a A~k BARRISTER AND SOLICITOR ID S:\C3\084\43019\01 \STR\0195-affidavit. u docS:\C3\084\43019\01\STR\O195-affidavit.doc
0 -2-0 u 2. We have reviewed the expert witness affidavit of Colin O'Leary and Eric Fergin filed herein. Attached hereto and marked as Exhibit "A" is a true copy of our u comments with respect to that Affidavit. The contents of Exhibit "A" and the findings and opinions expressed therein are true to the best of our knowledge, a information and belief. 3. Attached hereto and marked as Exhibit "B" are true copies of our curricula vitae. u 4. We make this affidavit pursuant to Rule 47(2} of the Competition Tribunal Rules. D SWORN BEFORE ME ) AT ) f ,. . ·- ~·~ D THIS ...::, DAY OF September 1999 ) RICHARD SCHWINDT 0
ru a SWORN BEFORE ME } AT ) D THIS I_:> DAY OF September 1999 } STEVEN GLOBERMAN D ---> ' GORDON W. RULEY BABR1SIEjfii & bOLIOITOiit u 0 SWORN BEFORE ME } AT } fil THIS _ DAY OF September 1999 } TERRY KEMP D
BARRISTER AND SOLICITOR D 1J
0 D u D 0 D This is Exhibit "An to the Affidavit of 0 Richard Schwindt, Steven Globerman"and Terry Kemp sworn before me at the city of \'.\l:Jb:.,~·~.»tir.t) in the Province 0 of 6. ( . this i 'XJay of September, 1999 0 A Commissioner for Taking Affidavits, etc. D A Notary Public or other such officer entitled to take oathS(SCalfit'Jia s W o n r ~ ic l ' J t J~~ said jurisdiction BARRISTER & a 33066 1st AVENUE. MISSION, B.C. V2V 1G3 826-1266 u D 0 c D D 6 0
0 0 0 0 0 0 g REBUTTAL OF THE REPORT OF COLIN O'LEARY AND ERIC FERGIN u 0 Q Steven Globerman ~ Terry Kemp R. Schwindt D September 1999 D 0 G' D D n 1J
0 .·5 0 I. INTRODUCTION We have been asked by the Commissioner of Competition to review and comment upon the evidence of Colin O'Leary and Eric Fergin. Their opinion is largely based upon a 0 background study prepared by A. T. Kearney (hereafter, Kearney). The Kearney study deals with the identification and quantification of efficiency gains that they claim will be generated by the merger of Superior Propane Inc. (hereafter, SPD and ICG Propane Inc. 0 (hereafter, ICG). 0 Our appraisal proceeds as follows. First, a broad overview of the projected efficiencies is set out. This is followed by a discussion of general issues bearing upon the identification, quantification and credibility of projected, merger-dependent efficiencies. 0 The next section addresses specific estimates (i.e., line items) in the Kearney Report. Our review concludes with a summary evaluation. D II. OVERVIEW OF PROJECTED EFFICIENCIES 0 The Kearney Report breaks the proje~ted efficiency gains into three broad categories: corporate centre, customer support, and field operations. Corporate centre Ill refers primarily to head office functions. Customer support refers mainly to the Ii.I administrative tier that lies between head office and the distribution branches. Field operations involve branch operations including both administrative activities and the actual delivery of propane. Corporate centre savings account for a little more than a third, field operations for nearly half, and customer support for the remaining savings. The savings, D by line item, annually, over ten years, and as a percentage of the ten-year savings are shown in Table 1. 1 D
The organization of Table 1 corresponds to the alphanumeric headings used in the Kearney Report This same scheme is used in our discussion of specific efficiencies later in this report.
0
0 0 Table 1 0 Projected Efficiency Gains 0 A 1 Corporate Leadership A 2 Finance D A 3 Head Office & Administration A 4 Human Resources A 5 Infonnation Systems Operations 0 A 6 Infonnation Technology Capital Costs A 7 Marketing Expenditures 0 A 8 Public Company Costs A 9 Procurement A 10 Supply & Transportation Organisation 0 A 11 Other Corporate Positions Total Corporate B 1 Field Support Administration 0 B 2 Salesforce B 3 Regional Team Support 0 B 4 Regulatory, Safety & Technical Department Total Customer Support a c 1 Field Sites c 2 Branch Managers c 3 Plant & Operations Staff c 4 Delivery Fleet 0 c 5 Delivery Drivers c 6 Service Technicians 0 c 7 Service Fleet c 8 Propane Field Inventory c 9 Customer Equipment D c 10 Supply & Primary Transportation Organisation Total Field Operations a GRAND TOTAL ~ ~ D 2 The ten-year total is in nominal dollars. 0
ro 2 Savings %of ($millions )2 IO yr Annual 10-yr total 1.5 12.7 3.2% 1.4 12.1 3.00/o 1.0 9.7 2.4% 0.6 6.0 1.5% 3.5 29.8 7.4% 1.3 14.8 3.7% 1.4 12.8 3.2% 0.7 6.6 1.6% 2.8 26.l . 6.5% 0.7 6.7 1.7% 0.5 4.2 1.00/o 15.4 141.5 35.3% 3.9 35.7 8.9% 1.9 18.3 4.6% 0.5 3.7 0.9% 0.9 8.0 2.0% 7.2 65.7 16.4% 3.5 40.3 10.1% 2.3 21.3 5.3% 0.6 5.7 1.4% 2.6 33.4 8.3% 3.9 36.3 9.1% 1.8 16.7 4.2% 0.6 8.5 2.1% 0.0 3.1 0.8% 0.0 11.3 2.8% 1.4 17.0 4.2% 16.7 193.6 48.3% 39.3 400.8
1 3 Ill. GENERAL ISSUES a A. Projected Versus Realized Efficiencies O'Leary and Fergin support their conclusions (based upon the Kearny Report), in 0 part, by stating that, + the Superior business model is easily adopted by ICG personnel.. .. 0 • the efficiencies are not on account of a complex or speculative re-engineering of product lines and services, development of new products or services, cross-D selling opportunities or development of a new business mode. 0 + the business lines, products and services of the two businesses are substantially similar in the sense that propane is a commodity type product and other than as a result of service, personnel or price, the two companies' products are D indistinguishable. These statements suggest that the integration of the businesses can be achieved easily, and that the projected gains are not speculative. This is very optimistic in view of the experiences of other enterprises that have engaged in mergers and acquisitions.
There is a large body of research examining the impacts of industrial mergers. Since this research spans a wide time period and different types of mergers, it is difficult to apply the findings directly to the merger in question. Moreover, outcomes of the mergers studied vary in ways that are neither fully identified, nor completely understood.
Notwithstanding these disclaimers, one finding that is robust, and that potentially bears upon the efficiency gains estimates reviewed in this report, is that the anticipated net 0 benefits of mergers frequently exceed the realized net benefits. Indeed, many mergers are ultimately ''unwound," as acquired assets are divested. One fairly common reason for the divergence between expected and actual outcomes of mergers is the difficulty encountered o· by the acquiring firm in integrating the operations and personnel, including management, of the acquired company. Efficient and effective implementation of the acquiring firm's merger strategy is critical to the attainment of anticipated net benefits, and actual implementation is fraught with contingencies and complexities that are often difficult to anticipate.
In light of a strong potential for any given merger to have disappointing ex post results, it is prudent, when forecasting the efficiency gains from a merger, to discount the "base case" scenario to reflect the realistic risks confronting the smooth and successful
0
implementation of the acquiring finn's merger plans. While there are various possible ways to incorporate the risks of unanticipated higher costs, or equivalently lower cost savings, the main point is that a base case merger scenario that does not explicitly 11 incorporate contingencies for unanticipated problems and delays in strategy u implementation should be seen as an upper-limit, and likely excessively optimistic, estimate of the merger gains that will actually be realized.
B. Changing Estimates The difficulty of accurately projecting the costs and benefits of a merger is reflected in the changes in the parties' claimed efficiencies over the past twelve months. Both the sources and the magnitudes of the savings have undergone multiple revisions. Some of these changes are significant. For example, between August 4, 1998 and August 17, 1999 (the Kearney Report), the minimum estimated savings increased by 21%. 3 The 0 earlier estimates were based on the closing of 58 overlapping sites, the later estimates, on 76 closures. On October 16, 1998 the parties provided a more detailed breakdown of the projected savings. Between that date and the later Kearney review, corporate centre 0 savings nearly doubled, and assumed customer loss went from 10 percent ofICG's volume to zero. 4 Different estimates were again provided on November 16, 1998.s LI C. Benchmarks u A useful check of the magnitude of anticipated cost savings, is to compare them with other mergers in the industry. In this regard, SPl's acquisition of Premier Propane in D 1993 could provide valuable insights. u Premier Propane distributed product in British Columbia, Alberta and the Yukon. It operated 24 branches and had 6 agents and one distributor. In 1992 it sold approximately 278 million litres of propane, which was about 9 percent more than SPl's fl volumes in Alberta and B.C. In other words, in tenns of volume within the areas served by Premier, this was a merger of equals. 3 The August 4, 1998 estimate is from Davies Ward & Beck, "Submission to the Director of Investigation and Research Regarding the Proposed Acquisition of ICG Propane Inc. by Superior Propane Inc." The August 17, 1999 estimate is from the Kearney Report. The comparison is between projected savings in present value terms.
4 The October 16, 1998 estimate is from a letter of that date from Milos Barutciski (Davies Ward & D Beck) to Francine Matte (Competition Bureau). We assume that corporate overhead plus IT capital spending are equivalent to Kearney's corporate centre category. s 0 See the letter from Milos Barutciski (Davies Ward & Beck) to James Bocking (Competition Bureau) dated November 16, 1998. 0
4
q
s SPI and Premier had a significant number of overlapping branches. Fifteen of Premier's 24 branches overlapped 13 SPI branches. In terms of volume, about 68 percent of Premier's sales were in overlapping areas. This provided the opportunity for cost savings that in part motivated the acquisition.
Unfortunately, SPI has no record of an ex post evaluation of cost savings attributable to this acquisition. 6 However, there is a record of SPI's ex ante estimates of savings that would be generated by integration of the two organisations. These can be compared with SPI's projected savings from its acquisition ofICG, and the results of the comparison are set out in Table 2.
The second row of Table 2 shows the post-merger SPI volumes in broad trade areas that had been served by the acquired firms (i.e., for Premier, Alberta and B.C., for n ICG, Canada). The third row shows the combined volume in areas with overlapping u branches. The overlap with Premier (73 percent of volume) was less than the overlap with ICG (87 percent of volume). The annual savings are broken out in similar categories: .~ savings from closing the acquired company's head office, savings from rationalization of the administration of field operations and savings attributable to the rationalization of field operations (e.g., drivers and trucks). D To provide a basis for comparison, savings in terms of cents per litre are u calculated. The per litre savings with respect to head office costs are not too different between the two acquisitions. 7 However there is a significant difference in projected savings, on a per litre basis, for both customer support and field operations in overlapping D areas. With respect to customer support, the projected savings from the ICG acquisition are 88 percent more than were anticipated from the Premier purchase, and for field operations, 31 percent more. To our knowledge, there is nothing in the documents that D explains why savings will be so much larger as a result of the proposed acquisition. D a fi 6 D SPI, Answers to Undertakings, #178 (Transcript Undertaking# 158, p. 1002) 7 The per litre saving is calculated by dividing head office savings by the total volume in the broad e trade area. This is admittedly arbitrary. Notionally, in the case of the Premier merger, the ratio suggests that SPI head office expenditures attributable to its sales in Alberta and B.C. were "spread over" an increased volume. In the case of the ICG acquisition, SPI's head office costs are now spread over the merged entity's total volume. 0
0 u 0 Table 2 Comparison of Projected Costs Savings Attributable to SPI's Acquisitions of Premier and ICG fl u Combined volume (million of litres) Combined volume (overlap) (million oflitres) Savings: head office ($000/year) D Savings: customer support ($000/year) Savings: field operations ($000/year) Total Savings ($000/year) D Savings: head office (¢/litre, total volume) Savings: customer support (¢/litre, overlapping volume) D Savings: field operations (¢/litre, overlapping volume) u D. The Identification of Merger-Specific Resource Savings ti The intent of efficiencies analysis in a merger evaluation is to identify and quantify real resource savings that would likely be generated by the merger and could not be obtained in another way. This raises questions as to what constitutes a real resource ll savings, what makes for a merger-specific savings and what point in time should be used for comparisons. ll 1. Real Resource Savings 0 Broadly speaking, a real resource saving exists when an equivalent volume of output can be created with fewer resources (more precisely, when resources expended per u unit of output decline). Resources include land, labour, capital and entrepreneurial effort. Such savings are beneficial to society because the released (saved) resources can be employed to produce other goods and services that will increase total societal welfare. n Real resource savings differ fundamentally from what are called "pecuniary" economies or savings. For example, by means of a merger a firm might be able to extract wage concessions from its employees. In such a case there would be a financial cost
8 The Premier estimates of cost savings were generat~Q. for internal use by SPI management. No attempt was made to purge pecuniary economies or to identify savings that Premier could have achieved on its own. Therefore, these savings undoubtedly exaggerate real resource savings. 0
6 Premier 8 ICG %ICG> Premier 532.1 2,509.0 389.9 2,192.6 2,618 14,150 618 6,570 2,603 19,360 5,839 40,080 0.49 0.56 14% 0.16 0.30 88% 0.67 0.88 31%
0 0 7 u reduction for the merged entity that would show up in its accounts, but there would be no saving of resources. As many workers would be employed as before the merger, only u their wages would be reduced. Since no resources (i.e., workers) would be released to produce additional goods and services, society would not be better off. Obviously the merged entity would benefit, and the workers would suffer from this redistribution. This 0 type of redistribution does not generate efficiency gains and therefore is not counted as a benefit of a merger. 0 It is also possible that a merger will result in increased pecuniary costs as opposed to increased resource costs. For example, as a result of an amalgamation a firm might have to pay higher wages for the same labour. In these circumstances, the increased costs should not be "charged" against the merger because, in fact, no additional resources have been drawn from alternative employment. D The distinction between real and pecuniary savings can be quite subtle as illustrated by a number of the costs and savings identified in the Kearney Report. For D example, in their discussion of savings in customer support, they note that while total employment increases, there are net savings because the eliminated positions had a higher D average salary than the average salary of the added positions. 9 If the new positions require less skill than the old positions, then this would qualify as a real resource saving as the more skilled workers are now released to employ their skills elsewhere. Conversely, if D skill levels are equivalent, this is simply a pecuniary saving since the effective amount of labour will not change. In their analysis of savings within the human resources department, they indicate that SPI will not offer a "wellness program" (an employee D benefit) previously offered by ICG and they count this as a cost saving. In fact, this is largely a pecuniary saving that is no different than a wage concession. 10 D Another area where the distinction between real and pecuniary savings (or costs) is problematic involves the licensing of intellectual property. A case in point is Kearney's a treatment of the avoidance ofICG expenditures on Microsoft software licenses as a saving, and increased licensing costs for SPI as a cost of the merger. In fact, the avoidance (or taking on) of such licensing fees does not result in real resource savings (or 0 costs). As long as the software would have been developed regardless ofICG's or SPI's demand, their use or non-use of the technology does not affect the amount of resources used up in the development of the technology. More simply, Microsoft will develop new D word-processing software regardless ofICG's or SPI's use of that technology. IfICG does not license the technology, resource savings will amount to Microsoft's incremental a costs of supplying ICG with the software (e.g., CD-ROMs) which are trivial. u 9 Kearney Report, p 141. 10 Indeed, it might be that the "wellness" program's benefits (e.g., reduced absenteeism, higher productivity) outweigh its costs, in which case denial of the program might increase SPI's costs. D
D \').; 0 8 u Symmetrically, the resource cost of SPI providing additional employees with access to Microsoft's software is not the licensing fee but the actual incremental resource cost to u Microsoft of producing copies of the technology (also trivial). Of course, the development cost of intellectual property that is specific to either ICG or SPI would qualify as a resource cost (i.e., this technology would not be developed absent ICG's or u SPI's demand). 2. Merger Specific Economies An analysis of efficiency gains focuses on savings that are merger-specific. That is, only efficiencies that could not be achieved absent the merger. In the context of the proposed amalgamation of SPI and ICG this raises two concerns. First, to what extent a could SPI and ICG achieve rationalization on their own? This is particularly relevant to customer support and field operations. Second, to what extent are the projected savings unique to this merger as opposed to an acquisition ofICG by another party? D a. Autonomous rationalization In the period leading up to the announcement of the merger both ICG and SPI D were involved in reorganizations of their respective businesses. This is especially true of ICG in that it was moving towards a more centralized, information technology dependent D operation. SPI was also undergoing change, albeit less dramatic (e.g., they were migrating to new, system-wide software). Under these circumstances it is extremely a difficult to estimate operating efficiencies that each could have achieved as a stand-alone business. It is acknowledged that the Kearney study does attempt to quantify gains that D would have been made by each party absent the merger. Indeed, this complicates some of their analysis to a significant degree. However, it must also be acknowledged that the 0 arresting of IC G's rationalization program in midstream interjects greater uncertainty into the comparison of projected merger related gains with what might have existed had the autonomous rationalizations been allowed to unfold. More simply, this case requires 0 comparison of potential merger related efficiency gains with potential autonomously achieved efficiency gains. This renders the analysis more speculative than it otherwise would be. D b. Efficiencies specific to this merger D Another issue with respect to the identification of merger-specific savings is the uniqueness of the estimated efficiencies to this particular combination as opposed to 0 another. More specifically, can the projected efficiencies be achieved only through the merger of SPI and ICG or could they be obtained through the union ofICG with some other party? 0
0
0 9 0 Arguably, some, perhaps many, of the savings at the customer support and field u operations levels are unique to this merger. SPI and ICG are the only national propane distributors operating in Canada. No other single amalgamation would combine as many field operations as would this one. Of course, ifICG's operations were hived off to 0 multiple, regional propane distributors there would be obtainable savings at the field level. Head office savings are another matter. Were ICG to be acquired by another firm, no matter whether it was involved in propane distribution, there would be head office saVIngs. D 3. Point in time u The estimation of merger related efficiencies in this case is further complicated by the choice of the point in time at which comparisons are made. Conceptually, two comparisons should be made. First, what are the merged entity's resource costs as 0 compared to the sum of the current stand-alone resource costs of the merging entities? Second, as discussed above, what are the savings associated with the merger as compared to savings the merging entities could achieve as stand-alone operations?
The Kearney Report, however, seems to use multiple points in time to make comparisons. In some cases the immediate present is the benchmark. For example, if a currently vacant position is eliminated, severance costs are assumed to be zero. In some cases the more distant future is used. For example, elimination ofICG's future investments in on-truck technology is counted as a saving. In some cases a future, different business form ofICG is assumed. For example, Kearney assumes that ICG would become an IPO and then counts the elimination of public company costs as a D merger related efficiency. Their choices in this regard often serve to maximize projected cost savings. 0 E. Transition and Integration Costs, and Volume Losses a Quite commonly, the horizontal merger of enterprises involves transition and integration costs, and some volume losses. Both of these factors affect the magnitude of anticipated efficiency gains. D 1. Transition and Integration Costs D The projected efficiencies of this transaction are largely driven by the integration of customer support (the second tier of administration) and field operations. These two 0 broad categories of activities account for nearly two-thirds of the estimated cost savings, and both are complex. The proposed integration would involve the merging ofICG's D
0 0 10 0 100,000 customers with SPI's 200,000 customer base, the integration of and rationalization of IC G's 110 distribution sites with SP I's 140 sites, the integration of a u substantial number ofICG's 700 employees into SPI's workforce of 1,300 people, and the integration and rationalization of an extensive delivery fleet. The business involves the distribution of propane, so integration will require the meshing of two complex networks. u Moreover, the two enterprises have adopted fundamentally different operating philosophies. One, ICG, is moving towards a more centralized, information technology dependent model, while the other, SPI, continues to operate a more decentralized system. Given these facts, the integration of these two firms would appear to be a daunting task. However, the Kearney Report identifies very few costs attributable to the actual process of integration.
Of particular concern are the costs of workforce integration. Much of the Kearney Report focuses upon worker redundancy in the merged entity and savings that will be enjoyed as these redundancies are purged. Indeed, a significant proportion of the projected savings is attributable to the 14 percent reduction in employment resulting from the merger. What the Kearney Report largely ignores is the cost of integrating those ICG employees who will be retained in the SPI or.ganisation. 11 D Table 3 shows pre-merger employment in the two companies, as well as the absolute and percentage post-merger increases in SPl's workforce by major activity. While total employment is reduced by 14 percent, SPl's workforce will be increased by a third, and in several activities by substantially more than this.
a a D D 0 11 Kearney allows for a one-time cost ofSI00,000 to recruit 70.5 "new positions" required in the merged organii.ation. 0
0 D 0 Table 3 Additions to the SPI Workforce as a Result of the Proposed Merger 0 Pre-Merger SPI ICG 0 A 1 Corporate Leadership 10 A 2 Finance 21.5 other :financial 3 0 A 4 Human Resources 9 A 5 Information Systems Op. 9 u A 10 Supply & Transportation 10 A 11 Other Corporate Positions 11 B 1 Field Support Admin. 291 135.5 B 2 Salesforce 53 B 3 Regional Team Support 19 B 4 Reg., Safety & Tech. Dept. 11 c 1 Field Sites c 2 Branch Managers 101 c 3 Plant & Operations Staff 129 c 5 Delivery Drivers 428 c 6 Service Technicians 227 GRAND TOTAL 1,332.5 734.0 D Given this dramatic, rapid increase in employment, one would expect significant integration costs. ICG employees would have to be trained in SPI's systems and D procedures. Current SPI employees would have to deal with a new branch configuration, a revised system of customer support, expanded volumes, and, of course, the integration a of the ICG employees. Remarkably, the Kearney Report does not envisage a substantial cost to this process of integration. a Undertaking: To advise whether the response to the efficiencies undertaking assumes a that there would be no additional or incremental training costs or training as a result of adding the ICG employees. If there are any assumed incremental training costs to advise that the amount of these a costs is and how they were estimated. [Transcript U/T 126, p. 812] u D
\~ 11 Post-Merger Total SPI Increase % Iner. 7 17 10 0 00/o 23.5 45 24 2.5 12% 5.5 8.5 3 0 00/o 7 16 12 3 33% 7 16 9 0 00/o 6 16 12 2 200/o 7 18 13 2 18% 426.5 390.5 99.5 34% 43.5 96.5 72 19 36% 7 26 19 0 0% 21 32 20 9 82% 34 135 106 5 5% 18 147 134 5 4% 282 710 630 202 47% 130 357 322 95 42% 2,066.5 1, 776.5 444.0 33%
0
0 12 0 Answer: The response to the efficiencies undertaking assumes that there will not be additional or incremental training costs as a result of adding the ICG 0 employees. D This appears to us to be patently unrealistic. It is simply not credible to assume that the expansion of a workforce by one-third can be accomplished at no, or very little cost. There is clearly an error of omission in this regard. 0 2. Volume Losses D It is quite common for a merged entity to lose volumes in the post-merger period. This can occur for a number of reasons. Dissatisfied SPI customers that had signed up u with ICG might be reluctant to return to SPI. Customers unhappy with the closure of local ICG branches and termination oflocal ICG employees might seek other suppliers. Undoubtedly, during the process of integration errors will be made that will alienate customers. Finally, where alternative distributors exist, they will likely attempt to take advantage of any disruptions of the SPI system. Indeed, SPI personnel acknowledged that there would likely be volume losses due to "integration turmoil," and in its undertakings, ICG noted that it expected a 5-10% loss of customers as a result of the acquisition. 12 However, the Kearney Report has not assumed any volume loss.
There are two ways to view the effects on efficiency gains of likely volume losses. On the one hand, if volumes are lost to other distributors who can reap network and scale D economies from the additional throughput, then SPI's efficiency loss is the other distributor's efficiency gain. On the other hand, if other distributors cannot extract the same amount of savings from the increased volumes, then this difference must be charged 0 against the projected SPI efficiency gains. Since we have seen nothing to indicate that other distributors would not enjoy similar efficiencies, we would accept the implicit assumption that volume losses would not seriously affect the level of obtainable efficiencies.
IV. SPECIFIC EFFICIENCIES The Kearney Report is quite lengthy and complex. In order to make an evaluation of the projected efficiency gains more tractable, the following analysis is organized a according to the format of that report. Each source of cost savings is identified (and is keyed to the tabs in the Kearney Report), and the ten-year sum of savings, and the percent of total savings attributable to the category are set out. The intent is to show the relative D 12 See: Mr. Schweitzer, Competition Hearing, October 21, 1998, p. 22; ICG Response to Undertakings, numbered 243. D
D r1 0 13 a importance of the category before engaging in a more detailed discussion of specific efficiency sources. Several line items are not discussed. This does not imply acceptance 0 of the estimates, but only that no obvious concerns were identified. In several cases we have combined line items when the activities are closely linked. D A. Total Corporate Savings ($141.5 million, 35.3% of total savings) 0 SPI intends to shut down ICG's head office and to perform these duties at its own head office. Broadly speaking, the eradication of duplicate head office activities can 0 generate real resource savings. At issue is the extent of those savings. • A 1 Corporate Leadership($ 12.7 million, 3.2% of savings) D These savings are attributable to the elimination ohop ICG management positions. u Conceptually, this type of savings makes sense in that top management at SPI should be able to accommodate the additional responsibilities due to this horizontal expansion. However, two points are worth making. First, Kearney omits severance costs for a position that is not currently filled (for a one-time avoided cost of$263,000). This is done throughout the report and is a very questionable practice. It is likely that a number of the vacant positions are a result of the uncertainty created by the merger itself. Absent that uncertainty, the positions would be filled and severance would be in order. If they would not have been filled, then the salary D savings could have been achieved without the merger and should be disregarded. D Second, Kearney includes part, but not all of the incentive payments to top management. Ift he total incentives received by ICG management were necessary to keep those employees in ICG's employ, they should be counted as a part of the efficiency gains.
Symmetrically, this raises the question of incentive payments to SPI's management.
SPI has in place a "Management Agreement" that provides incentives to the D manager that are tied to the cashflow distributed to the fund per trust unit outstanding. The formula is set out in Table 4. D
a 0
0 0 14 D Table 4 SPI Management Incentive Scheme 0 Cashflow distributed to Incentive fee entitlement the Fund per trust unit of the Manager less than $1.27 nil First target $1.28-$1.45 15% Second target $1.46-$1.89 25% Third target $1. 90 and greater 0 50% This type of incentive scheme is important to the analysis because if the projected D efficiencies are achieved, a substantial portion of the benefits will become payable to management. That is, the administrative costs of SPI will increase as cost savings increase. Put differently, as one set of costs (e.g., operating costs) declines another 0 (management costs) will increase. At issue then is the magnitude of the increased leadership costs to SPI as a result of the merger. In order to bound our estimates we have imposed several simplifying assumptions. First, we assume the merger has no effect on·revenues (i.e., the merged entity's revenues D are the sum of SPI's and ICG's pre-merger revenues). Second, we assume that ICG's stand-alone cashflows are adequate to finance SPI's purchase oflCG, and that the purchase is bank financed (i.e., the number of unitholders is not increased). 13 Third, we u assume, following the Kearney Report, that the merged entity enjoys cost savings of $40 million per year. Finally, we assume that SPI's pre-merger cashflow equated to exactly a u $1.27 distribution per unitholder (i.e., distributable income had reached the ceiling above which a management incentive fee is payable). 14 Under these assumptions, the projected $40 million/year efficiency gain would generate management incentive fees according to u the schedule set out in Table 5. The average number of unitholders in 1998 was 43,839,000. Therefore, of the $40 0 million addition to cashflow attributable to efficiencies, $7,891,020 would attract a management fee of 15% (i.e., this amount would raise cashflow per unit to $1.45). The next $19,289,160 would attract a fee of25% (this amount would raise cashflow per unit to $1.89), and the remaining addition to cashflow of$12,819,820 would attract a fee of 500/o (this amount would raise cashflow per unit to $2.18). These calculations are
13 ICG's 1998 distributable cashflow was $22.8 million. The purchase price was $175 million. Given current interest rates it does not seem unreasonable to assume that the purchase could be financed from this cashflow.
14 In fact, SPI's distributable cashflow was a little above $1.27/unit in 1998 resulting in a management incentive fee ofS936,000 (Superior Propane Income Fund, "1998 Annual Report," p. 27). 0
D
15 a summarized in Table 5. In effect, the achievement of the $40 million per year in efficiencies would cost SPI an additional $12.4 million per year in management fees. 15 0 Table 5 SPI's Management Incentive Formula
Threshold Addition to Mgt incentive Mgt incentive a $/unit cashflow ($) fee rate fee($) 1.27 0 0 1.28-1.45 7,891,020 15% 1,183,653 1.46-1.89 19,289,160 25% u 4,822,290 >l.89 12,819,820 50% 6,409,910 Total 40,000,000 12,415,853 0 Conceptually, these fees should be treated as a post-merger cost of the merged D entity. Presumably, the projected cost efficiencies would not be achieved without the entrepreneurial efforts of the management team. Assuming the incentives reflect market forces, without them the team would seek employment elsewhere where their skills would 0 be appropriately rewarded. In the result, the fees should be deducted from the projected efficiency gains. D • A2Finance($12.l million,3.0%ofsavings) 0 In this area savings are almost completely related to the elimination of positions. In total, the pre-merger, combined finance workforce is to be reduced by nearly one-half. 0 In several functions this seems optimistic. For example, product accounting and accounts payable staffing varies directly with the number of transactions, the number of vendors and the number of branches and wholesalers. Kearney predicts that the combined staff in this D area will be reduced by one-third. This implies that activity in this area will be reduced by one-third. While there may well be a reduction in the total number of product transactions, it is not obvious why this should decline by such a significant amount. In the 0 case of propane purchases, our understanding is that Petro-Canada arranges this for ICG so we assume that SPI might well engage in more, not fewer transactions for this input. 16 D a IS This calculation is not meant to be precise. It simply attaches an order of magnitude to these increased management costs. 16 The Kearney Report notes at page 120 that Petro-Canada "manages all of the functions surrounding the procurement and primary transportation of propane including ... managing the accounting function for the payment from ICG to the propane suppliers." 0
D 0 16 Kearney also counts the termination of three ICG finance staff dealing with cardlock accounting (for annual savings of $181,860 17 ). No such positions exist in head u office at SPI as this function is carried out at the branch support centre level. However, a review of Kearney's analysis of the customer support functions (which includes the branch support centres) does not explicitly show any increase in staff to accommodate this u increased workload. • A 3 Head Office & Administration ($ 9.7 million, 2.4% of savings) Savings in head office and administration are attributable to occupancy costs and miscellaneous costs (liability insurance, corporate memberships, office supplies and other minor items). Of these, occupancy costs account for about half the savings, and a little more than half of this is rent.
There are serious problems with Kearney's treatment of the rent component of ICG's occupancy costs. First, they inflate the actual rent paid by ICG to Petro-Canada to current "market" values in Calgary. There is .no evidence that the opportunity cost to Petro-Canada of this space is any more than the amount paid by ICG, and therefore the inflation adjustment is not reasonable. Second, and more importantly, Kearney must be consistent in the handling of opportunity costs. At present, ICG pays Petro-Canada u $117,456 in annual rent for its office space. ICG's removal from Petro-Canada premises constitutes a cost saving only if Petro-Canada has an alternate use for this space. While we cannot speak for Petro-Canada, it may be that it has no internal use for the space and is 6 unwilling to sublet to a third party. Kearney goes on to state that SPI has vacant space at its head office that can accommodate the new employees and that this space therefore has no cost. This assumes that SPI has no alternate use for the space and that it is unwilling to D sublet to a third party. lfwe adopt similar assumptions between SPI and Petro-Canada (i.e., no use for the space and an unwillingness to sublet to a third party) then this move to D SPrs premises generates no efficiency gains. All that has occurred is a switching of the burden of excess capacity from SPI to Petro-Canada. As a result, there are no resource savings attributable to rent.
• A 4 Human Resources($ 6.0 million, 1.5% of savings) I This activity involves employee training, management of compensation and benefits, maintenance of payroll records and communication of policy. There are several D areas of concern with respect to cost savings in this activity. First, some savings are pecuniary (e.g., elimination of the "wellness" program, as mentioned earlier, is a pecuniary u 17 Based on data provided in the Kearney Report, p. 67. 0
'-----~. ~ ---- -
D ~l u 17 (l u saving). Second, ICG's entire training budget is eliminated, and four HR positions (out of a total of 16) are eliminated. It is not clear whether HR salaries were included in ICG's a training budget. If they were, elimination of the budget and elimination of the salaries would involve double counting. Third, Kearney does not envisage any increase in SPI's training budget notwithstanding the one-third increase in number of employees and the D addition of 11 branches. The largest component (72 percent) of SPI's training budget relates to "team building training activities at corporate head office and the branch network." Presumably, the increase in branches and employees would require some increase in expenditures. Kearney does allow for a one-time, $100,000 expenditure to recruit for the 70.5 new positions, but claims that integrated ICG employees will be trained through contact with the existing SPI employees on a day-to-day basis. The assumption here is cost-free, on-the-job training. While this type of training may not involve explicit payments it surely will impose implicit, and real, costs (e.g., reduced productivity of those providing the on-the-job training). Finally, severance for currently vacant positions is ignored (for an avoided, one-time cost of $60,000 18 ). As noted earlier, this should be included.
• A 5 Information Systems Operations($ 29.8 million, 7.4% of savings) D • A 6 Information Technology Capital Costs($ 14.8 million, 3.7% of savings) 11 The analysis of cost savings in this dimension is complicated by the fact that SPI is II simultaneously moving to a new system (the J. D. Edwards ERP System) and integrating ICG. It is difficult to evaluate the credibility of this analysis given the very unsettled state D of SPI's information systems operations. In any case, Kearney estimates that the merged entity will continue to employ 139 ofICG's computer-using positions (these positions are not identified in these sections). Presumably these additional computer-using employees D will need to be trained in the J. D. Edwards system. Insofar as they were already trained in ICG's system, there will be a cost to this that should be deducted from the merger savings. 19 D As noted earlier, increased licensing costs for existing software do not constitute a D real resource cost, and avoidance of such licensing fees do not constitute a resource savings. Hence, SPI's increased licencing fees for the addition of J.D. Edwards software users are not costs of the merger. Similarly, avoided payments by ICG for licences to use Iii II Microsoft and PeopleSoft upgrades that would have been developed regardless ofICG's purchases are not resource savings. We do not know whether the $333,000/year charge E 18 Based on three employees at an average severance of $18,880. See Kearney Report. p. 81. 19 E There are several errors in Exhibit 3 of this section. Terminal user costs are $882 times 100 users, which equals $88,200, not $93,492 as reported. This would increase efficiency gains by $5,292. Footnote 4, explaining hardware support, is missing. 0
D 0 18 for PeopleSoft upgrades, counted as savings (at p. 86), or the $2.5 million, one time PeopleSoft upgrade, counted as savings (at p. 98), involved custom work for ICG. If not, they are not resource savings.
• A 7 Marketing Expenditures($ 12.8 million, 3 .2% of savings) Kearney deals with marketing expenses in two sections. Some activities are carried out at head office (considered here) and others are carried out at the branch level. Most of the head office marketing cost savings appear reasonable. Many of the line items do not vary with output (e.g., market research) and therefore would not be expected to increase as a result of the merger. Other, admittedly minor, activities would appear to vary with the size of the operation and probably should be increased. For example, employee awards and educational assistance presumably would increase with the number of employees. Travel expenses would presumably increase with the number of branches.
I As noted earlier, Kearney assumes no loss of volume. Interestingly, and probably very optimistically, this 100 percent retention is to be achieved at a cost of only $157,750. This amount will be spent on two direct mailings (apparently only to ICG customers) and D advertising in local newspapers. 20 • A 8 Public Company Costs ($ 6.6 million, 1.6% of savings) I The inclusion of public company costs is based upon the assumption that absent the merger, ICG would become a public company. However, it is as plausible to assume that absent this merger, ICG would be acquired by another operating company. Given the I uncertainty about ICG's business form if this merger did not take place, these savings should not be included in the calculation. D u • A 9 Procurement($ 26.1 million, 6.5% of savings) This source of savings is attributable to volume or quantity discounts that are to be extracted from trucking, rail and transportation and fleet leasing and purchases. Kearney B states that these discounts are granted because "increased volumes allow the vendor to spread its fixed costs over a greater number of units and reduce its per unit costs E 20 The Kearney Report estimates two customer mailings at a total cost of $100,000. Assuming these letters go only to ICG customers, this implies a per letter cost of only 50¢ per letter. SPI customers, B many of whom will be served by different branches, different branch contact people and different drivers, apparently will not be informed of the changes except insofar as they read local newspaper advertisements. o.
D D 19 u accordingly." 21 However, Kearney provides absolutely no evidence that these projected discounts are in any way attributable to the vendors' lower costs of serving one purchaser instead of two.
Indeed, Kearney indicates that they do not clearly understand the distinction u between real and pecuniary economies in their discussion of truck freight. In this section they claim that no procurement savings "are expected on the ETI shipments since Superior a owns this operation and is already receiving this service at cost. " 22 The implication is that third-party trucking companies are charging something above their costs and this margin can be negotiated away. If there were truly resource savings due to increased volumes,
ETI's costs would be expected to decline. Moreover, if this were the case, Superior, as the owner ofETl, would be able to easily document the resource savings.
With respect to rail freight, Kearney states that the marginal cost of adding a rail car to a train is small, and thus increased volumes will result in lower per unit fixed costs. What they neglect to consider is that the removal of one rail car (i.e., ICG's) will symmetrically increase per unit fixed costs and thus, benefits and costs are equal. 23 Finaliy, it appears that Kearney did not substantiate the magnitude or the basis for these projected discounts with the relevant suppliers. 24 The estimates are based solely upon Kearney's experience in negotiating transportation contracts for other clients.
In short, we believe there may be some very modest transaction costs (e.g., billing, contract monitoring) savings as a result of the merger but that the majority of savings in D this category is pecuniary and should be disregarded. D • A 10 Supply & Transportation Organisation($ 6.7 million, 1.7% of savings) There are two main sources of cost savings in this activity. First, SPI would avoid the charges imposed by Petro-Canada for providing supply and transportation organisation for ICG. Second, SPI would be able to reduce its need to dispose of excess propane D through wholesale markets. 21 Kearney Report, p. 114. 22 Kearney Report, p. 114 23 Actually, no resource savings would be generated by adding one more railcar to a train, regardless of whether another was removed. The railroad company's fixed costs would remain the same, notwithstanding the addition of the marginal car.
24 SPI Undertaking Response #150 (Transcript p. 840), and SPI Undertaking Response #152 (Transcript urr 131, p. 845). D
D D 20 D With respect to the Petro-Canada contract, which requires ICG to pay $924,000 for the services, Kearney notes that about half the payment is for salaries of Petro-Canada employees, and the other half is for infonnation technology licenses and maintenance D costs, and for allocated Petro-Canada overheads including management oversight. 25 These overheads should not be counted as a cost savings because they are likely to persist in the post-merger period. The only change will be that Petro-Canada cannot spread the overheads over ICG functions as well as its own.
The savings associated with cessation of wholesale activities are suspect. It is not clear why SPI is currently in an excess supply position (and thus must wholesale some product) nor why acquisition ofICG's volumes will attenuate the surplus problem. If SPI currently buys more than it needs because of inflexible supply contracts (e.g., a "keep-dry" contract that requires removal of more propane than SPI can use), it is difficult to see how ICG's volumes will lead to a "perfect fit." It seems just as likely that ICG's volumes will reduce surpluses in some geographic areas but create shortages in others, thus requiring SPI to enter into other inflexible supply contracts. Moreover, assuming that SPI can avoid wholesaling excess propane, it would seem that some other party would have to provide the wholesale function. Total production and consumption of propane in Canada would not change so the total amount sold at wholesale into the U.S. would not change, and thus I] I the total amount of wholesaling activity would not change. In the result, there is no resource saving in this regard. n I It is also possible that SPI is in a surplus position in order to avoid any potential shortfall. However, if this is the case, it is unclear why the acquisition oflCG would D reduce the need for this "buffer." Both companies operate largely in the same geographic markets so the acquisition could not be viewed as a diversification. Put simply, the risk of inadequate supply is not reduced by the acquisition ofICG. D B. Total Customer Support ($65.7 million, 16.4% of savings) D This broad category involves the second tier of SPI's administrative structure. D Savings in this area account for approximately 16 percent of total savings. D • BI Field Support Administration($ 35.7 million, 8.9% of savings) E This category of savings accounts for a significant proportion of the total projected efficiency gains. In the SPI organisation there are three levels of administration: head 25 E Avoidance of the infonnation technology license fees may or may not be real resource savings. We do not know whether this technology was developed specifically for ICG or is of a general nature. If it is the latter, no savings are generated as explained earlier. D
D u 21 u office, field and branch. Savings at the customer support level essentially involve the rationalization of this level of administration. u Kearney's analysis of this rationalization is very complicated as it attempts to identify and make comparisons between multiple organizational structures. These include: the current structures ofICG and SPI, the future structures ofICG and SPI had they remained separate organisations (both were reorganizing their customer support systems), and the post-merger structure of SPI. Understandably, this leads to considerable conjecture as to what will exist in the future and what savings are actually attributable to the merger.
About 70 percent of the efficiencies are attributable to salary savings which in tum are due to both the elimination of positions and the reassignment of tasks to lower paid staff.
Staffing for a number of the rationalised functions appears to be directly dependent D upon variables such as the number of customers, number of deliveries, and so forth. Why there would be a disproportionately small increase in staff to handle increased transactions is not made clear. a The Kearney analysis identifies some significant and unexplained differences between the productivity of SPI and ICG staff performing similar functions. Dispatchers D are a case in point. Table 6 (derived from Kearney's Tab B 1, Figures 1 and 4.3) shows the number of bulk deliveries, the number of dispatchers currently employed by SPI and D ICG and the average number of deliveries "handled" per dispatcher. Remarkably, SPI's dispatchers are 3 times more productive than ICG's dispatchers. It appears that Kearney simply divides the number ofICG bulk deliveries by SPI's deliveries per dispatcher to E determine the additional number of dispatchers required. 26 If this is in fact the methodology, we must indicate concern over the substantial productivity differentials between the two companies. This is so for two reasons. First, if the differentials are real, there must be some rationale for why ICG could not increase productivity on its own. Second, if the apparent differentials are explained by a different mix of tasks (i.e., ifthe ICG dispatchers do more), then this should be accounted for in the calculation of savings.
26 Dividing ICG's number of deliveries by SP I's average dispatched productivity (484,230/37 ,260 = 13) yields 13, while Kearney reports that 14 additional dispatchers will be required.
0 22 u Table 6 u Staff Productivity Comparisons: SPI-ICG No. of bulk deliveries No. of dispatchers Deliveries per dispatcher I] u SPI 1,602,167 43 37,260 ICG 484,230 40.5 11,956 D No. of ship-to customers No. of payment Customers per payment processors processor ll SPI 230,653 10 23,065 ICG 144,847 27.5 5,267 No. of ship-to customers No. of credit rep. Customers per credit rep. processors processor SPI 230,653 31 7,440 1.. ICG 144,847 27 5,360 .I Other significant productivity differentials (all in favor of SPI) exist for payment I processors and credit representative processors. With respect to payment processors, the ' higher productivity is due to two factors. SPI assigns most of the processors' functions to the branch, and it outsources most of the accounts receivable payment processing function to a third party. Kearney accounts for the increase in workload at the branch level as these processor positions are eliminated, but does not assume any increase in outsourcing costs. Presumably the cost of outsourcing will increase as the number of accounts increases. Salary savings for credit representative processors are based upon the application of SPI's productivity rate to the number ofICG's ship-to customers. Again, A there is no explanation as to why ICG could not achieve an equivalent level of I productivity on its own. D There are considerable projected savings attributable to the reassignment of tasks to lower paid staff. This raises a number of questions. Is this simply a pecuniary gain? In D other words, if the skills required for the job do not change as tasks are reassigned, and if the workload is similar, then the reassignment results in no real resource savings. Ift he tasks are assigned to a less skilled (and therefore lower salaried) person this raises two E questions. First, why was this not done before if a less skilled person could do the job? Second, if in fact the employee acquires skills to fulfill the job requirements, will the employee remain satisfied with the lower salary? In any case, we suspect that there is some muddling of real and pecuniary economies in savings attributable to reassigilment of tasks.
Finally, here as in other sections of the report, Kearney assumes no training costs. Fourteen dispatchers and sixty branch administrators will be given "on-the-job training" at E no cost to SPI. As noted earlier, this is not reasonable. Even on-the-job training involves 0
0 ~1 23 D costs, although these costs might well be implicit (e.g., loss of productivity of the LI employee/trainer) rather than explicit (e.g., wages paid to a professional trainer). u • B 2 Salesforce ($ 18.3 million, 4.6% of savings) Projected efficiencies in this area are due largely to workforce reductions and, in ~. the overall calculations, are relatively significant. Broadly speaking, a substantial salesforce reduction is to be expected in that Superior, by this proposed merger, is a removing its most significant competitor in most markets and its only competitor in some markets. D A review of Kearney's analysis reveals inconsistencies. For example, their Figure 3.1 indicates a position eliminated at Smithville because "additional salespersons added to D Concord will serve this area." However, Exhibit 2 does not show any increase in salespersons at Concord. D • B 4 Regulatory, Safety & Technical Department ($ 8.0 million, 2.0% of savings) u Savings in this category are all attributable to workforce reductions. Kearney anticipates the elimination of 12 positions. At least one source of savings is quite questionable. Kearney claims significant savings with respect to meter testing. This function involves the testing of bulk delivery truck meters. At present, ICG employs four meter technicians at a total salary cost of$259,052 to perform this function. Kearney claims that these employees can be discharged for a savings of$259,052, and that SPI can outsource this work for the 153 trucks added to its fleet at a cost of $400 per truck, for a total cost of $61,200/year. The net saving in this regard would total nearly $200,000/year. However, given ICG's current fleet of bulk delivery trucks (228 vehicles), the implication is that ICG now spends $1,136 per truck for meter testing while, according to Kearney, SPI has equivalent work done by a third-party for $400/truck. This is difficult to comprehend. In any case, if this is true, ICG could independently discharge these meter technicians and outsource the work for a total cost of $91,200. Thus the efficiency gain would be $30,000, not $200,000 per year.
C. Total Field Operations ($193.6 million, 48.3% of savings) g Projected efficiencies generated at the field operations level are very significant, accounting for nearly half of the anticipated total. These efficiencies are largely attributable to the rationalization of the branch system and the improvement of delivery D logistics. 0
D 24 u u • CI Field Sites($ 40.3 million, 10.1% of savings) a As a result of the merger, SPI intends to shut down a total of 76 field sites. The volumes previously moved through these closed sites will be added to the volumes of the remaining sites. About 84 percent of these savings are attributable to the elimination of D facilities operating costs and the elimination of ongoing site capital expenditures. The remainder is accounted for by the sale of redundant sites (net of decommissioning costs). 27 These are legitimate savings as the sale value presumably reflects the opportunity cost of D the site. Clearly the rationalisation of the site network will generate real resource savings. However, Kearney appears to assume that the remaining sites will be able to carry the s additional volumes without any increase in operating expenditures or ongoing capital expenditures. We do not find this credible. The reason for our incredulity is that many of these costs are related to volumes, staffing levels and numbers of customers. Volumes in D all rationalised trade areas will increase, and, at some, volumes will more than double. Staffing will increase at the branches for a number of reasons. First, the number of customers per branch will increase significantly, and this will increase the number of D administrative staff required to serve these customers. Second, according to the customer service rationalisation plan, many tasks will be reallocated to branch employees (e.g., branch administrators). This will also increase staffing. Third, the increased volumes will require more delivery and service staff (e.g., drivers). Fourth, ICG service technicians who previously worked out of their homes will now work out of the branches.
Equipment located at the branch or operating from the branch will increase. This would include storage tanks, and trucks. This could well require more space, and expanded infrastructure (e.g., paved areas). Storage space for inventories (e.g., parts and customer tanks) would have to be increased.
With respect to the operating costs listed by Kearney, we note that many of these are a function of the volume of business at the site. The following are examples.
+ Utilities: The cost of utilities (e.g., electricity, water, and sewage) would increase with the number of employees and level of activity. • Property Taxes: Taxes would likely increase if expanded operations required physical additions to the site. Increased volumes and staffing could require such additions. • Telephone: This cost would surely increase with the number of customers. Moreover, Kearney earlier noted that branch administrators would assume
27 Kearney notes that at ICG sites that have environmental damage, Petro-Canada will incur the costs for the clean-up (p. 190). Presumably, SPI will be liable for environmental clean-up of D decommissioned SPI sites. Kearney makes no provision for these potential costs. 0
25 many of the responsibilities ofICG's call centre staff. This too would increase telephone requirements at the branch. • Insurance: If this refers to property insurance (as opposed to the liability insurance which is paid at head office), then we would assume that this would increase with any additions to the remaining sites. • Postage: This is dependent upon the number of customers and would increase. • Office equipment rental: This depends upon the number of staff Given the significant downloading of tasks from the customer support level of the organization to the field level, and commensurate branch level staffing, we would expect increases in this regard. • General and ongoing maintenance: These types of general maintenance costs would increase with the activity level at the branch. • Equipment maintenance: Given the increase in equipment at the remaining sites we would expect increases in this category. • Office supplies: These costs would increase with the level of administrative activity.
0 With respect to annual capital expenditures, we assume that these involve maintenance of existing capital. As branch il}.frastructures are expanded to handle increased volumes and staffing, these costs would surely increase. Li Perhaps more importantly, Kearney seems to ignore the capital costs of upgrading site infrastructure to accommodate the increased activity. Presumably, storage tanks D would be moved from some decommissioned sites to the remaining ones. Presumably, office space would have to be increased at some sites to accommodate additional staff. I Presumably, storage buildings would have to be expanded. Presumably, in-yard roads and parking areas would have to be expanded and upgraded to accommodate more traffic. In short, it is difficult to believe that each remaining site has enough idle capacity to absorb a this increase in activity. If they do, Kearney has not shown this. To give a sense of the potential increased activity at a site we have examined changes to operations at the Peterborough branch. Peterborough was chosen because the rationalisation is straightforward. Presently there is one ICG branch and one SPI branch. D SPI intends to close the ICG site and consolidate the operations at the SPI branch location. Table 7 shows the impact of this consolidation. D D u 0
0 30: u 26 Table? 0 Activity Changes at SPI's Peterborough Branch Before After 0 Administrative Staff 2 3 0 Drivers 4 8 Manager 1 1 Dispatchers 0 a 1 Service Technicians 2 2 Plant Operating Staff 1 1 0 Total 10 16 Cylinder Dock yes yes D Cylinder Trucks 1 1 Service/Crane Trucks 4 4 28 D Bulk Delivery Trucks 4 8 Bulk Volume 29 (litre/year ) Storage Tankage 30 (USWG) 60,000 60,000 D Loading 31 (avg. fills/ week) 1.1 3.7 The staffing level will increase by 60 percent. Cylinder operations will be D consolidated at this site which will increase cylinder truck traffic. The bulk delivery fleet will double. The increased fleet will require additional maintenance capacity on the site as D well as general access and parking area. This could require reconfiguration of the site to handle the step change in delivery equipment. Bulk delivery volumes are projected to increase by 220 percent. Such a large increase will mean that both primary deliveries and a bulk truck daily liftings will also increase proportionately. This suggests that the site will have to be reconfigured to handle the significant increase in load factors. u Presently SPI has 60,000 USWG storage capacity at Peterborough. With approximately 1. 1 fills per week the storage capacity is reasonably matched with the SPI a bulk business requirements. It is however, unreasonable to expect that SPI can handle the combined volumes with this storage capacity. What this would mean is that, on average, D We assume trucks can only be eliminated in "whole" units. Kearney Report, p. 277, p. 280. 30 SP! spreadsheet titled "SPI_ICG.XLS." 31 Kearney Report, p. 282, states that storage tanks are filled to the 80% level. A tank would be refilled well before its is empty (they attempt to maintain a minimum fill level of 20%). Thus, effective tank D capacity available when refilling is in the 60% range of its rated capacity. 0 '
0 3{ 27 D 3. 7 loads per week would be to be delivered to the storage tanks. In winter, during peak u demand periods, this would increase to approximately 1 load per day. More realistically, SPI would need to relocate the ICG storage tanks onto the site. This could involve considerable costs, even if it is assumed that space was available (the siting of tanks must u conform to spacing and fire regulation guidelines that could preclude the installation of an additional storage tank). 0 This cursory review of the Peterborough rationalization suggests that there will be significant capital costs involved in moving ICG volumes onto the SPI site. Moreover, 0 both operating and annual capital upgrade costs would surely increase. Generalizing from this case study, we find that Kearney's assumption that the move of these loads to remaining sites will be cost-less (both with respect to capital costs and on-going operating 0 and capital upgrading costs) is unreasonable. D • C 2 Branch Managers($ 21.3 million, 5.3% of savings) D The elimination of branches will obviously allow for the termination of branch managers. Savings will follow from this. W~ note that throughout the sections of the u Kearney Report dealing with customer support rationalization there are numerous examples of tasks that will be reallocated to the branch level. This will place greater administrative burdens upon remaining branch managers. We do not know if Kearney's estimated branch manager salary increases are compatible with these increased D responsibilities. D • C 4 Delivery Fleet($ 33.4 million, 8.3% of savings) • CS Delivery Drivers ($ 36.3 million, 9 .1 % of savings) D Savings in these categories are attributed to the number of bulk delivery vehicles that the combined organisation can eliminate while still meeting the combined annual average delivery.requirements by trade area. The savings are categorised by the operating savings from eliminated vehicles ($15.5 million), the timing difference for vehicle a replacement for the reduced fleet ($14.37 million), and disposal of excess equipment ($3.52 million). The number of delivery drivers is directly related to the number of a delivery trucks. A proper appraisal of Kearney's estimates of truck eliminations was impossible a because the methodology was inadequately explained. Apparently, a national average volume per operating hour (for each truck type) was used to estimate the current capacity 0
0 u 28 D of the combined SPI and ICG fleet within each trade area. Then an undefined "delivery fleet requirements predictive model" was used to determine redundant capacity. 32 D The first step in the Kearney analysis was to estimate extant trading area delivery capacity. This was done by applying a "comparative historical performance" measure (in terms of litres delivered per operating hour) for each vehicle type to the current, combined D fleets in each trading area. We compared these fleet populations with data previously supplied by the parties and found serious inconsistencies. 33 As a result, we are unable to 0 verify the fleet inventories used by Kearney. We then reviewed the "average litres per operating hour" measures that were used 0 to estimate current branch delivery capacity. In order to test the credibility of the application of these averages we compared Kearney's estimate of extant, combined capacity by trade area with their estimate of extant, combined volume by trade area. The 0 results of this exercise are shown in Table 8. a The first column of Table 8 shows the merged entity's trade areas. The second column shows combined (ICG's plus SPI's) sales volumes by trade area as calculated by Keamey. 34 The third column shows annual delivery capacity for each trade area. 35 D Column four is the ratio of capacity to volume sold shown as a percent. The values these ratios (which show the balance between capacity and volume) take are difficult to comprehend. In 21 of the 50 trade areas, capacity exceeds volume (in a number of cases D capacity is more than twice the volume sold), while in 29, capacity is below volume sold (remarkably, in 10 cases, capacity is less than half the volume sold). Capacity as a percent of volume sold ranges from a low of30 percent to a high of273 percent. If these capacity D 32 This model is named in a footnote, but is not described (Kearney, p. 238). u 33 See for example the fleet inventories for Cranbrook/Valemont and Strathroy as listed in the SPI electronic file titled SPI_ICGU.XLS and in the Kearney Report. For Cranbrook/Valemont, the Kearney Report lists 50 percent more single axle bulk truck capacity than the electronic file. For
Strathroy, the Kearney Report lists 25 percent less tandem axle bulk truck capacity and 50 percent D less tractor-trailer capacity. These are very significant differences. 34 u Kearney Report, pp. 278-279, Tab C 8, Exhibit 1. Footnote 3 to this exhibit states that these volumes were identified as "originating at the location in question." We assume these are traditional bulk deliveries and exclude direct deliveries (i.e., deliveries from the point of supply to the end-user). 35 The capacity estimates are based upon data in Exhibit 1 of Tab C 4, pp. 236-237 of the Kearney D Report. At Exhibit 1, Kearney shows the capacity in terms of "litres per operating hour." This figure was derived from their Exhibit 2, Tab C 4 which indicates "litres per operating hour" for single, tandem and tractor propane delivery vehicles based upon an SPI database sample. To D determine a trade area's delivery capacity in terms of "litres per operating hour," Kearney multiples "litres per operating hour" for each vehicle type times the number of this type of vehicle in the trade area. We have converted "litres per operating hour" to an annual figure by multiplying the hourly estimate by 2,080 hours (52 weeks times 40 hours). The June 1999 precursor to the Kearney study D (Exhibit 2 of the Delivery Fleet tab of the Response to the Efficiencies Undertaking) identifies the "annual hours available" as 2,080. 0
D 0 figures, which form the basis for Kearney's estimates of projected fleet reductions, are u accurate, we cannot understand why at present there is so much apparent excess capacity in some trade areas, and such a shortage in others (indeed, we cannot understand how volumes were delivered in 1998). 0 Table 8 Combined Volumes, Combined Delivery Capacity, 0 Capacity as a Percent of Volume and Projected Capacity Reduction (volume and capacity in millions oflitres per year) 0 Trade Area Combined Combined Volwne a Moncton Rimouski Jonquiere Baie-Comeau D St-Romauld Thetford Mines Cap de Madelaine 0 Drummondville Granby Joliette Vimont 0 Gatineau Ottawa Pembroke D Kingston Peterborough Barrie D Concord Strathroy Stratford D Simcoe Walkerton North Bay D Sudbury Porcupine Echo Bay Thunder Bay Kenora Winnipeg/ Brandon Regina/ Yorktonl Swift Current Saskatoon/ Prince Alberti La Ronge Calgary Red Deer/ Rocky Mountain House D 0
29 Cap. as % Kearney% Capacity of Vol. reduction 43.7% 3.72% 20.71% 64.0% 30.2% 18.97% 13.18% 38.2% 222.4% 4.09°/o 9.39°/o 74.7% 72.2% 7.22% 268.9% 1.60% 99.5% 8.38% 1.93% 59.1% 145.5% 3.90% 2.38% 170.5% 6.28% 83.2% 127.2% 16.15% 4.50% 126.3% 114.3% 7.30% 106.9% 3.75% 67.1% 3.67% 3.71% 112.9% 130.7% 4.47% 129.4% 12.87% 140.4% 3.88% 99.6% 5.66% 112.0% 8.66% 54.6% 31.78% 57.3% 29.11% 151.0% 6.95% 104.0% 28.54% 11.03% 78.1% 21.27% 114.2% 87.1% 38.00% 185.3% 9.24% 5.46% 272.8%
0 G Table 8 (continued) 0 Combined Volumes, Combined Delivery Capacity, Capacity as a Percent of Volume and Projected Capacity Reduction 0 (volume and capacity in millions of litres per year) Trade Area Combined Combined Cap. as % Kearney % 0 Volume Edmonton Lloydminster Edson/ Grande Prairie Slave Lake Peace River/ High Level Lethbridge/ Brooks/ Medicine Hat Yellowknife Fort McMurray Castleg ar Cranbrook/ Valemount Kamloop~Invermere Nanaimo/ Port Hardy Coquitlam Prince George Terrace/ Fort Nelson Whitehorse Watson Lake The next step in Kearney's analysis was to apply their "delivery fleet requirements predictive model" to the combined volumes in the trade area. Their description of this model was committed to a footnote and simply says "[t]his model determines the reduction in operating hours required to serve a trade area in the merged organisation." Earlier, they state that the number of delivery trucks required will depend upon:
+ the distance between deliveries (which depends upon the size of the driver's territory and the number of stops the driver need to make during his shift); • the amount of propane delivered (which depends upon customer requirements and how much the vehicle can deliver during a day); • the total number of deliveries, and; 37 • the format in which the propane is delivered (bulk or cylinder). Kearney gives no indication of how their "delivery fleet requirements predictive model" actually models what must be a very complex relationship between these variables. Without disclosure of the model, there is no way to test its credibility.
36 Kearney Report, p. 238. D 37 Kearney Report, p. 231. 0
30 Capacity of Vol. reduction 171.7% 11.53% 70.0% 27.40% 94.0% 23.08% 93.8% 11.54% 117.6% 13.91% 72.8% 13.18% 85.7% 14.81% 63.1% 18.81% 109.0% 13.96% 76.7% 20.18% 72.0% 16.00% 58.6% 9.82% 49.1% 1.82% 92.3% 14.74% 91.1% 76.35% 42.5% 38.40% 74.2% 26.90% 36
0 35' a 31 a Based upon application of their model, Kearney then lists capacity reductions, in terms of fractions of specific vehicles (i.e., single axle, tandem or tractor-trailer), that can be achieved by the merged entity. The calculation of cost savings is based on these a fractional reductions. The choice of what type of vehicle to retire when excess capacity is identified D raises concerns. Generally, the choice is made to retain the more productive tractor-trailer units. Consider the following example. Assume 3, 745 litres per operating hour is deemed to be redundant. Eliminating I tractor-trailer or 3 single axle trucks could reduce this a capacity. On the surface, since drivers are linked to trucks on a one-to-one basis, it is far more cost effective to eliminate the three smaller vehicles, and this is the option generally a chosen by the Kearney methodology. At issue is whether the one tractor-trailer can actually do the work of the three smaller trucks. The report states that "the decision to eliminate single and tandem trucks and not tractor-trailer units is justified given the need D to retain larger fleet delivery vehicles to serve the increased density of the combined trade area." However, this is not realistic because each type of truck is utilised for specific a customer types. Single axle trucks are geared to small residential deliveries while large tractor-trailer units are appropriate for deliveries to industrial/automotive customer sites and customers with larger tankage configura~ions. 38 These customer characteristics will not change with the merger. Moreover, if tractor-trailers could in fact perform the work D done by the smaller units, both companies would be using more of these trucks now. 0 With respect to fleet configuration, it should be acknowledged that as larger volume vehicles are used to service customers previously supplied by smaller trucks, the average efficiency of the larger trucks would fall. Presumably, more calls per daily route u will result (as the larger trucks take on a larger proportion of smaller volume customers), and this will lower "litres per operating hour." D Estimating fractions of vehicle reductions by trade area to determine overall fleet reductions is problematic. Apparently, this is based upon the notion that trucks can be worked across trade area boundaries. This is a credible assumption. However, we believe that shifting trucks between trade areas will result in some reduction of efficiency. For example, there will be costs in terms of non-revenue generating movements between areas, training times involved in training drivers in multiple routes, and so forth. Kearney does not appear to account for these inefficiencies.
Having estimated the extent of fleet reductions, Kearney then estimates capital cost and operating cost savings. Annual operating cost savings associated with the retired vehicles is set at approximately $1.6 million. What the report overlooks is the potential D 38 A residential customer likely would not allow a larger tractor-trailer truck access to a property due to road damage and turning configurations. Also deliveries to small residential tanks is more costly per 0 unit for the large vs. smaller trucks. 0
0 32 0 increase in the operating costs of the remaining fleet. Success of the fleet rationalisation 0 scheme is predicated upon using the remaining fleet more intensively. That implies increased fuel and maintenance costs per vehicle over a year. 39 These increased costs are a ignored in the Kearney Report. Kearney estimates average annual capital expense savings of $1. 4 million over a ten-year period. These savings are attributable to the retired vehicles that will not have to D be replaced in the future. However, the more intensive use of the remaining trucks will probably reduce their economic life. If so, the annual capital costs required to maintain D the fleet increase (i.e., they will wear out faster). Furthermore, an increase in intensity of use will likely increase the frequency of downtime due to scheduled maintenance and a unanticipated repairs. This in tum would require investment in additional "stand-by" capacity. The Kearney report ignores these potential increases in fleet capital costs. a Many of these same concerns identified with respect to bulk trucks also apply to the projected cylinder truck fleet reductions. Kearney's explanation of how they estimated redundant capacity in this fleet is perfunctory. The explanation is committed to a footnote ~ that simply states that "fleet reduction principles developed with regards to bulk vehicles" were applied. 40 The procedure generates a cpnstant 12.95 percent reduction in all trade a areas where both SPI and ICG miiintain cylinder delivery fleets. We find the application of a constant reduction factor over diverse trading areas unrealistic. Finally, fractional reductions are projected and no allowance for increased intensity of use of the remaining fleet is made. · 0 In summary, our review of this section of the Kearney Report raises four broad Il issues. First, we cannot replicate the fleet rationalisation results because the methodology is opaque. Second, we cannot verify the basic fleet inventories upon which the projected gains are based. Third, we cannot verify the credibility of the estimated extant delivery a capacity. Fourth, we believe that increased operating and capital costs of the remaining fleet have been ignored. As a result we cannot accept the estimates of efficiency gains in a this area. Further, because the number of drivers is directly tied to the number of trucks in service, we also cannot accept the estimated reductions in the number of drivers. B 39 Even in the unlikely event that average annual mileage per vehicle post rationalisation did not increase, the average load per kilometre (i.e., the average weight carried by the vehicle) certainly 0 will. The volume of propane will be delivered by a smaller number of vehicles. Furthermore, each vehicle will have to make more "starts and stops." This increases fuel consumption, engine wear, drive-train wear, and brake wear, at a minimum. D 40 Kearney Report, Exhibit 3, p. 240. 0
0
33 D 0 • C 6 Service Technicians($ 16. 7 million, 4.2% of savings) • C 7 Service Fleet($ 8.5 million, 2.1% of savings) u Service technicians are responsible for installing and maintaining customer installations (both tanks and propane appliances) and plant facilities. The number of D service vehicles depends directly on the number of service technicians. Kearney envisages a 10 percent reduction in this workforce. The cost savings are all salary related. We have several questions with respect to these cost savings. First, Exhibit 1 indicates that in the u Northwest Territories ICG employs three technicians while SPI has none. Post-merger, SPI will employ just two technicians in this trade area. It is not clear why ICG could not have performed this rationalization on its own. Second, we have problems understanding the formula applied by Kearney to determine the number of redundant personnel. Apparently they calculate that on average an ICG technician spends 40 percent of his or a her time doing internal work. They then assume that for each site closed, 0.4 of one technician will be redundant. However, an inspection of Exhibit 2 indicates that the a number ofICG technicians per site ranges from 0 to 10. According to their formula, this implies that internal work at ICG sites can range from 0 to 4 full-time technician equivalents. This is not credible. D Kearney ties the reduction of the service fleet to the elimination of service technicians on a one-to-one basis. Presumably the number of service vehicles would be D more directly related to the number of off-site calls than to the number of service people. In turn, the number of calls is presumably a function of the number of accounts, which will not decline in the post-merger period. Without evidence of some sort of efficiencies in D organizing call responses we would not expect such a significant reduction in vehicles. 41 Finally, Kearney notes that the crane truck fleet will be reduced by 11 percent. B They claim that both SPI and ICG maintain excess crane truck capacity and it is this that allows for the reduction (the demand for crane trucks is largely driven by the number of a customers). However, they go on to note that both SPI and ICG lease crane trucks to satisfy short run demand. We wonder why both could not reduce crane truck excess capacity independently by relying more upon leased vehicles. u • C 9 Customer Equipment($ 11.3 million, 2.8% of savings) 0 As a result of the merger, Kearney states that the inventory of customer equipment a (i.e., tanks) held at branches can be reduced. While the number of installations and in-use 41 The Kearney Report does identify "unquantified" efficiency gains attributable to increased customer density (and presumably more efficient routing of service technicians when making service calls). D However, they provide no support for this assertion. 0
34 tanks will not change, the "back-up" inventory will. This makes sense as inventory requirements generally do not increase lock-step with the number of installations. In other words, a merged entity can support the existing population of in-use tanks with a smaller inventory.
We have no dispute with the general concept but we do have some concerns about Kearney's assumptions and calculations. Apparently, Kearney assumes that SPI and ICG each held optimal inventories before the merger, and the merged entity will optimize inventories in the post-merger period. If there were redundant inventories pre-merger, and the merged entity moves to optimal inventories, then the cost savings are exaggerated.
Second, we notice quite different relative inventory levels by size of tank both preand post-merger. Table 9 sets out tank inventory as a percent of tanks in-use by tank size. In both the pre- and post-merger period, these percentages are quite different between tank sizes. Kearney does not explain this. Moreover, the projected rate of inventory reduction differs significantly between tank sizes (see Table 9). For example, the inventory of 2, 000 gallon tanks is reduced by 56 % (and this accounts for 25 percent of the total cost saving), while there is no room for any reduction of the inventory of 1,800 gallon tanks. In fact, reducing the inventories ofj ust two tank sizes (123 gallon and 2,000 gallon) accounts for 59 percent of the cost savings. We find this puzzling, and Kearney provides no explanation for the variations.
Table 9 Customer Tank Inventories
Tank size lnventon:: as % ofln-use Tanks Pre-merger Post-Merger 33 lbs. 13% 9% 100 lbs. 26% 19% 123 gals. 11% 5% 250 gals. 17% 11% 500 gals. 6% 4% 1000 gals. 12% 8% 1750 gals. 24% 20% 1800 gals. 25% 25% 2000 gals. 22% 10% 0 0
% Inventory % of Cost Reduction Savings 28% 3% 26% 2% 55% 34% 33% 1% 39% 17% 34% 14% 14% 4% 0% 0% 56% 25%
0
• C IO Supply & Primary Transportation Organisation($ 17.0 million, 4.2% of savings) Projected savings in this area are· attributable to: an increased utilization of SPI's trucking fleet (ETI), an increased use of the Marysville underground storage facility, and a reduction in the number ofleased railcars.
Kearney posits that SPI will glean savings from an increased use of its trucking operation (ETI). Apparently, ETI has idle capacity in periods of low propane demand. 0 Using ETI to move ICG volumes that were previously transported by third party carriers would produce the anticipated savings. However, this does not generate any real resource a savings. ETI's idle capacity will simply be moved to the third party carriers. Furthermore, in Section A 9, Kearney projected freight cost reductions from third party carriers in return for increased volumes. In this section, volumes are being moved away from third party carriers to ETI. In any case, these are pecuniary, not real resource gains. 0 Efficiencies generated by increased use of the Marysville storage facility are of a D similar, but not identical nature. Kearney states that ICG meets its seasonal demand peaks by drawing down inventories from its suppl~ng refineries' storage facilities or by having the refineries alter their operations (presumably to produce more propane) thereby D impairing the efficiency of the refinery process. Insofar as idle capacity is moved onto the refineries' storage facilities, this is a pecuniary, not a real savings. However, to the extent that refineries do alter their output slates to accommodate ICG's demand, and thereby 0 increase their own costs, this represents a cost savings. However, if this were a serious problem (i.e., a common and costly occurrence) we would presume that the refineries would pass these costs on to ICG and ICG would look for alternatives. Given available underground storage capacity, ICG should be able to obtain this efficiency on its own. Finally, it is our understanding that at Marysville, variable charges are applied on a per unit basis for delivery to and lifting from the storage facilities. Thus the increased use of the Marysville facility is not cost-free.
With respect to the reduction in the "safety stock" ofleased rail cars, the argument appears to be that the "safety stock" is required to accommodate unanticipated demand, n and that by combining volumes, SPI can reduce the stock. This would likely be true if u unanticipated increases in demand affecting ICG's customer base were independent of those affecting SPI's customer base. However, we have no reason to believe that any such independence exists, and therefore cannot accept this rationale for a reduced "safety stock" of railcars. 0
B 0
35
0 36 0 V. CONCLUSIONS 0 Our primary conclusion is that the estimated overall efficiency gains described in the report of Colin O'Leary and Eric Fergin are likely a substantial overstatement of the 0 actual efficiency gains that will be realized from the merger. The overly optimistic estimates ofMssrs. O'Leary and Fergin are due to a number of factors, including: D • the frequent confusion of pecuniary economies for real economies; • the disregard of the potential for cost savings to be realized by ICG in the absence of the ~erger; 0 • unrealistic expectations about labour and capital productivity of existing SPI management, field staff and branch sites; a • ' the unsupported assumptions concerning idle capacity within the extant SPI and ICG organisations; • the disregard of costs that will accompany the more intensive use of SPI's 0 rationalised staff and assets; • the disregard of increased management costs at SPI following the merger; • unrealistic expectations with regard to transition costs; 0 We are confident that the projected cost savings are exaggerated, but it is 0 impossible for us to provide a precise quantification of the overestimate of the merger's efficiency gains. Nonetheless, estimates of the order of magnitude are possible. Of the three organizational areas covered (corporate centre, customer support and field a operations), the most easilr dealt with is corporate centre. If the merger resulted in complete amalgamation of the two enterprises, projected real resource savings (i.e., Kearney's estimates minus pecuniary and speculative savings) would be negated by the D increase in SPI's management costs. The removal of corporate centre savings reduces the projected annual savings by approximately 38%. D Quantification of the over-estimation of resource savings at the customer support a and field operations levels is more difficult. In these areas there are clearly errors of omission. Importantly, the costs of integrating new employees into the SPI organization are largely ignored. The practicability of reassigning tasks from more highly skilled workers to less skilled workers is not tested. The costs ofloading more activities on the remaining branch sites are largely ignored. The methodology for estimating fleet reductions is seriously incomplete and the costs ofloading more intensive activity on the remaining fleet are 'largely ignored. Given these factors we believe that their estimated savings in these areas could be, conservatively, reduced by approximately 20 percent in order to reflect real resource savings. In the result, the removal of the corporate centre savings and a 20 percent reduction of the savings attributed to customer support and field operations would reduce the overall projected savings by approximately $20 million, that is, by one-half.
0
0 37 0 0 Finally, it must be noted that many of the potential efficiencies identified by Mssrs. O'Leary and Fergin are not unique to this merger. Were ICG to be acquired by another operating firm, there would undoubtedly be head office resource savings. Were ICG's 0 branches to be acquired by other incumbent propane distributors, there would undoubtedly be customer support and field operations resource savings. In their report, Mssrs. O'Leary and Fergin do not attempt to either identify or quantify the potential efficiencies that could D be reaped by alternative acquirers. We believe that this is a serious shortcoming in that the potential is significant. 0
0 0 0 0 0 a D a 0 D u n D - 0