Modelling the ease of entry in merger analysis: can financial analysis move the ball?

Malcolm B Coate; Arthur J Del Buono

doi:10.1093/jaenfo/jnx013

Coate, Malcolm B;Del Buono, Arthur J

2018-04-01 00:00:00

Abstract In competition policy, ease of entry, when present, generally trumps competitive concerns and allows market behaviour, such as a merger, to proceed unchallenged. Thus, the entry issue plays a role in every antitrust study. That said, it is surprising that entry analysis is inconsistently defined, both in the US courts and at the Agencies. Research suggests that the key problem revolves around the analysis of the likelihood of entry. This article addresses the likelihood problem head-on, suggesting that this question be addressed with a discounted cash flow analysis to assess the profitability of a hypothetical entry able to deter or defeat the potential anticompetitive effects caused by the merger. We present a comprehensive discounted cash flow model and broaden the analysis to evaluate the major risks of the project. We also discuss how to evaluate the results of the model and discuss various objections to and limitations of the modelling technique. Our methodology provides practitioners with an innovative approach to evaluate the profitability and thus the likelihood of entry in an antitrust analysis. I. INTRODUCTION In the US, as in other jurisdictions, ease of entry effectively trumps a wide range of antitrust concerns. Horizontal mergers that lead to duopoly or even monopoly market structures are not anticompetitive if entry will deter or defeat the adverse market impact of the transaction.1 Below cost pricing on the part of a monopolist is not predatory if profitable entry precludes the monopolist from recouping its foregone profits.2 Vertical restraints are not anticompetitive when entry or the threat of entry blocks the firms in question from raising price or reducing quality-adjusted output. Only the perse standard used for horizontal price fixing agreements enables an antitrust challenge to market conduct in the absence of evidence on entry.3 Traditional antitrust analysis focused on the lack of “barriers to entry” as a proxy for ease of entry. In theory, if barriers to entry did not exist, then new entry into the relevant market must be considered easy and thus, anticompetitive effects are extremely unlikely, if not impossible. This approach proved very difficult to implement, because economists could not agree on a definition for barriers.4 In the early 1990s, it became clear that the concept of a barrier must be defined in light of the economic model of competition applied in a particular analysis.5 Within this understanding, differences in definition collapse to differences in modelling structure. Of course, this does not solve the problem, as different economists propose different modelling structures and thus adopt different implicit barrier definitions. The key take-away is that case-specific analysis must drive the entry study.6 Ease of entry, therefore, must be an empirical question. Practitioners have long sidestepped the methodological problems by implementing a case-by-case entry analysis as part of the merger review algorithm.7 Starting in 1982, the US Merger Guidelines advanced a hypothetical entry test focused on the likely magnitude of the entry that would occur within two years in response to a small, but significant and non-transitory price increase (usually 5 per cent).8 In 1992, this approach evolved into the three prong (timeliness, likelihood, and sufficiency) standard that addressed the ability of entry to deter or defeat an anticompetitive effect.9 Although the timeliness and sufficiency concepts proved relatively manageable in the case-specific analyses, likelihood analysis often remained difficult to fully characterize.10 In some situations, Stiglerian barriers such as government fiat proved entry would not occur (hence entry was not likely), while in other cases, prohibitively large economies of scale or scope implied entry was not feasible. However, in numerous other situations, Federal Trade Commission (FTC) staff analysis simply mentioned scale economies existed and sometimes noted the market was declining. Usually, no further analysis was undertaken to support the inference on the lack of likelihood. In a few investigations, the staff reported the opinions of likely entrants, but tended to neglect to explain the reason for the opinion. In a few other cases, the staff discussed innovative quantitative analyses designed to study the actual profitability of entry.11 This article will follow up on the innovation of quantitative profitability analysis discussed in that entry study and provide a detailed overview of how to construct a financial model of entry. The 2010 revision of the Merger Guidelines is compatible with direct modelling of entry as it moves away from generic benchmark analyses (ie the two-year standard for entry or the computation of minimum viable scale) and highlights the importance of actual evidence on the entry decision.12 The history of actual entry in response to non-transitory price increases is considered particularly relevant. Examples of historical entry would define a clear road map for a financial analysis. Likewise, when only a few companies are considered to have the potential to enter the market, their specific financial situations control the review.13 Here, the entry model would be based on the business opportunities available to these firms. By focusing the study of entry on the output expansion needed to ‘deter or counteract any anticompetitive effect of concern’ in a timely manner, the entry model would yield insights into the three key questions in the 1992 Merger Guidelines. This article employs a common financial analysis technique, Net Present Value (NPV), which is used in the normal course of business to evaluate the profitability of a long-lived investment. NPV is defined as the difference between the positive and negative cash flows over the lifetime of an investment, discounted to the present using an appropriate rate.14 If entry, sufficient to maintain the competitiveness of the market, proves to have a significantly positive NPV, then it is reasonable to infer that entry is likely in response to less than competitive behaviour on the part of the merger partners. The NPV approach is general enough to allow customization to the key facts relevant to the antitrust case. As a bottom line, a financial model would offer insight into the ease of entry and thus aid the evaluation of the merger’s likely effect.15 Section II starts with an overview of entry analysis and provides perspective linked to both the case law and the regulatory regime. The basic modelling analysis is introduced in Section III, with the NPV model presented in detail. Section IV offers some real world complications, drawing examples from both theoretical and application concerns. Section V concludes. II.ENTRY ANALYSIS IN PRACTICE Structuralism in US merger policy reached its zenith in 1966, with the Vons Grocery decision. Justice Stewart dissented, with the infamous comment ‘The sole consistency that I can find is that in litigation under Section 7, the Government always wins.’16 Earlier in the opinion, the Justice highlighted the numerous successful entrants as well as absence of any barriers to entry, but these facts had no impact on the majority opinion.17 The merger was blocked; ease of entry was seen as irrelevant. Structuralist policy was shattered by the General Dynamics decision in which the Supreme Court reversed course and mandated a forward-looking review of the competitive process to control merger review.18 Market structure would continue to play a role in the analysis, as would numerous other factors. By the early 1980’s, ease of entry had become a key consideration in the merger review process. Entry analysis developed along a two-track path, one in the courts and the other at the Agencies. Inter-relationships exist between the two tracks, with the Agencies trying to guide the courts, and the case law controlling the Agencies’ law enforcement activity. However, the two lines of analysis remained remarkably different, with the courts retaining an interest in barriers to entry and the Agencies highlighting the issues of timeliness and sufficiency. Neither institution does a good job addressing the likelihood of entry. Entry analysis in the courts Ease of entry was formally established as a key consideration to rebut the structural presumption in Waste Management.19 Here, the court found that easy entry into the commercial trash collection business negated the structural concern associated with a merger that created a leading firm with a share of almost 50 per cent. Entry also played an important role in a number of other cases in the 1980s including Calmar, Promodes, and Country Lake Foods.20 In Calmar, the court focused on the threat of potential entry, while in Promodes, the combination of actual entry and lack of barriers precluded the alleged anticompetitive effect. Country Lake Foods introduced the idea of the large buyer as a potential entrant (or sponsoring entry). Without barriers to entry to block backward integration of customers into the market, prices could not be elevated. These cases all suggest that entry analysis should be customized to the facts in the particular case. In 1990, two important decisions, Syufy and Baker Hughes, set the legal standards for future entry analysis. Both cases involved mergers in highly concentrated markets and both mergers were found not to violate the law primarily because of entry considerations. In Syufy, the court rejected a DOJ challenge of a consummated merger to near monopoly due to ease of entry, finding ‘low entry barriers or other evidence of a defendant’s inability to control prices or exclude competitors’.21 A few months later, another appellate court addressed the Department of Justice‘s ‘quick and effective’ characterization of the entry standard in Baker-Hughes.22 The court rejected the DOJ’s approach as ‘unduly onerous’, noting it would shift the overall burden of proof to the defendant, mandate an untenable level of proof, and ignore numerous other sources of evidence.23 Relevant entry considerations include evidence of actual entry, facts suggesting barriers do not exist, and information on the impact of potential entry.24 In both cases, the court ruled that ease of entry trumps competitive concerns. Entry analysis at the agencies Insights from the merger guidelines The 1982 Guidelines recognized that entry would significantly reduce the probability of a merger challenge, if easy entry could prevent existing competitors from raising price for a significant period of time. To evaluate entry, the DOJ posited a hypothetical test linked to the amount of entry likely to occur within two years of an anticompetitive price increase.25 The Guidelines recognized sunk costs, stagnant markets, scale, and scarce resources as relevant concerns, although no clear algorithm was given for the likelihood analysis.26 The 1992 and 2010 revisions of the Merger Guidelines built on the 1982 foundation to provide the analyst with the three-part test of timeliness, likelihood, and sufficiency through which to evaluate the entry question.27 If entry is timely, likely, and sufficient, then anticompetitive pricing is considered unlikely to occur and if it is tried, entry will quickly compete the price down to the competitive level.28 The 1992 Guidelines detailed the three core lines of analysis. The first study focuses on the timeliness of entry, with information gathered to trace the path of entry from the initial planning stage to the point where the entry has a significant impact on competition. A two-year standard for the analysis is proposed, although longer times would be allowed in special case circumstances. The second study evaluates the likelihood of entry. Likelihood is considered synonymous with the long run profitability of the entry at competitive prices and thus the analysis concentrates on whether the entrant could cover its costs (including a return on the assets that must be invested in the market to compete). If the market conditions were such that an entrant could not expect to operate at the required scale, entry would not be profitable and thus not likely. The third test, sufficiency, ensures that timely, likely entry is sufficient to offset the competitive effect of concern. This analysis would rarely be necessary in a homogenous goods market, as timely, likely entry could be duplicated until it suffices to offset the competitive effect of concern.29 However, if the product is differentiated, the competitive effect of concern may be localized in ‘product space’. In these cases, the entrant(s) must introduce products in sufficient quantity to deter or defeat the price increase in the segment of the market affected by the merger.30 Thus, the logical approach to entry analysis is to first define a type (or types) of entry (eg upscale brand or full line) that would be sufficient to deter or defeat the specific anticompetitive effect of concern and then evaluate the timeliness and likelihood of that (those) type(s) of entry.31 This would represent a change in the order of the review, as the staff tends to sequentially proceed through the timeliness, likelihood, and sufficiency analysis for each type of entry considered potentially relevant to the merger in question.32 Insights from the staff entry analyses As noted in the introduction to this article, the Entry Study’s analysis of FTC merger investigations identified fact-based evidence for timeliness and sufficiency findings, but limited empirical support for the likelihood analyses.33 In some situations, readily identifiable barriers preclude entry into the market and thus a review of those factors (eg government regulation) served to show entry was not likely. When no obvious entry barrier existed, staff moved on to the qualitative discussion of entry conditions. These analyses, focused on viable scale, were often not supported with sufficient factual detail to clearly show entry was not likely. The research sample combined the 138 markets studied in the Entry Study, with a follow-on review of 42 more recent market analyses associated with mergers from 2006–2012. This created a database of 180 market studies undertaken in the 1993–2012 period. Summary results for both the Coate Entry Study and the 2006–2012 update are presented in Table 1. Concerns with the likelihood of entry were raised in 118 of the 180 matters, with 40 offering some type of clear barrier to entry analysis to justify the conclusion that entry was not likely to occur. These barrier findings include 21 matters in which some form of government regulation made entry unlikely, another 10 in which some type of exogenous (eg patent) or endogenous (eg strategic entry deterrence) barrier precluded entry and 11 in which switching costs facing customers served to deter entry.34 This leaves 78 markets in which the likelihood of entry was evaluated with other qualitative, but market-specific facts, often involving discussion of scale and its effect on the prospective profitability of entry. Table 1. Results of the FTC timeliness of entry study Coate (2008) Update Total 1993–2005 2006–12 1993–2012 Basic analysis Mergers reviewed 138 42 180 Likelihood issues identified 93 25 118 Actual barriers to entry 33 7 40 Questionable entry impediments 60 18 78 Innovative entry analyses Identifiable actual entrant 18 15 33 Financial analysis of entry 7 1 8 Impact on likelihood of entry Questionable entry impediments 60 18 78 Identifiable actual entrant 5 9 14 Financial analysis of entry 3 1* 4* Unsubstantiated claims 52 9 61 Coate (2008) Update Total 1993–2005 2006–12 1993–2012 Basic analysis Mergers reviewed 138 42 180 Likelihood issues identified 93 25 118 Actual barriers to entry 33 7 40 Questionable entry impediments 60 18 78 Innovative entry analyses Identifiable actual entrant 18 15 33 Financial analysis of entry 7 1 8 Impact on likelihood of entry Questionable entry impediments 60 18 78 Identifiable actual entrant 5 9 14 Financial analysis of entry 3 1* 4* Unsubstantiated claims 52 9 61 * One claim overlaps with identifiable entrants and thus is only counted once in the grand total. Two techniques for structured likelihood analyses were observed in the FTC documents. Firstly, the staff tried to identify and then interview the most likely entrants. If these firms could be identified and interviewed, the staff would have valuable information with which to understand the likelihood of entry issue. In the best-case situation, the specific entrants would disclose the reasons for their expected course of conduct and those reasons would provide a justification for a likelihood finding. Specific entrants were interviewed in 33 matters, of which entry was found to be expected in 12 and unexpected in 21. Secondly, the staff modelled the profitability of entry to determine if entry should be considered likely.35 Here, the basic idea involves showing entry into the market covers all relevant costs and therefore is economically profitable at competitive prices (although short run deviations from the competitive price may be used to trigger the entry). If the model shows entry is profitable, then it is reasonable to conclude that such entry is likely. Unprofitable entry is obviously unlikely. Some of the models were quite complex, focusing on the financial returns to entry, while others were simple, tending to build on specific information collected in the investigation. Financial modelling was observed in eight markets (five matters in which entry seemed likely and three in which it was not). Table 1 also shows that of the remaining 78 matters addressed in which entry was asserted to be unlikely without evidence on entry impediments to prove the claim, interview analysis was useful in 13 and modelling analysis was relevant in another three studies and both analyses were used in one. This left a rather large sample of 61 claims of likelihood impediments in which the obstacles were not clearly justified.36 As Table 1 indicates, the 2006–12 update showed the FTC had improved the likelihood analysis with the rate of unsubstantiated likelihood findings declining from 87 percent (52/60) in 1993–2005 to 50 percent (9/18) in the 2006–12 update.37 In light of the focus in the 2010 Merger Guidelines on actual entry decisions, it is possible that this result represents a conscious effort on the part of the FTC staff to collect evidence on perceived intentions of market participants from roughly 2006 onwards.38 Overall, the review of the recent merger investigations shows that the FTC’s likelihood analysis is becoming more fact-specific, but also identifies room for further improvement. By modelling the profitability of entry over time when no clear evidence of strong barriers to entry exists, a merger analyst can increase the evidence available to the decision-maker. Case law versus policy, the potential for convergence The courts and the regulators have applied different techniques to evaluate the likelihood of entry. For the courts, reviews have focused on evidence of actual entry (and if the information had been available, likely entrants), along with a theoretical analysis of barriers to entry. In contrast, regulators made limited use of evidence on actual entry39 and found strong evidence on barriers to entry in only 40 of the 180 markets surveyed. Alleged impediments to likely entry were identified in another 78 studies. The staff supplemented these findings with information from likely entrants, along with studies of entry profitability. What is innovative here, are the studies of entry profitability. By showing entry is profitable, the regulators and, if they adopted this approach, the courts, would focus the analysis more on facts and less on theoretical impediments to entry. This would represent an important improvement for both institutions, because the theory of entry is more complicated than the courts appreciate and, as seen in the review of the FTC activity, difficult to implement. Timeliness and sufficiency of entry findings should usually be integrated into a profitability model. The Guidelines no longer impose a hard and fast rule for the timing of entry. Rapid entry is now valued for its ability to quickly return the market to the competitive level. Antitrust authorities prefers rapid to slow entry, but the costs of rapid entry may preclude that option. If delays in entry need to be built into the financial model, the analysis would recognize that incumbents could maintain premium prices for an extended period of time. To the extent that any long run merger-specific efficiency benefits do not overcome these short-term pricing effects, the regulator could move against the merger based on the combination of short run injury and the lack of offsetting efficiencies. Sufficiency would also be addressed within the entry analysis, because the type of entry modelled would need to be sufficient to deter or defeat a price increase.40 Entry not directed at the competitive effect of concern would be downplayed by the merger analysis. Overall, when the profitability model is complete, incorporating the timeliness and sufficiency analyses, the analyst could make a reasonable prediction for the likelihood of entry. Kirkwood and Zerbe introduced a profitability analysis to determine if entry is likely.41 They assume that the analyst can determine the sales levels, the future growth or decline of the overall market, the firm’s cost structure, the incumbent’s reactions to the entry, and level of support offered by the customers. In their model, the entrant initially discounts price below the market level (itself elevated due to the anticompetitive effect of the merger) for the first two years and then obtains the pre-merger market price.42 If entry is profitable and the market price returns to the competitive level, the defendant meets its burden. 43 To prevail, the plaintiff would be expected to demonstrate that the entry would not be sufficient to restore competition. Clearly innovative, their model can be further developed to present a business analysis that would justify entry into the market. Such an analysis would not require omniscience about market conditions or the responses of market participants to various actions by other participants. The analysis would instead require reasonably reliable and complete information available to the firms in the market to build a model that the managers themselves could use to assess the attractiveness of entry. Existing but confidential case-specific FTC work, some of which is included in the Coate Entry Study discussed above, employs NPV analyses of entry (along with other less sophisticated approaches), on occasion modifying work found in internal documents of the parties. Confidentiality issues preclude discussing or generalizing these models. However, the basic ideas are relevant. Modelling actual investments better illustrates the risks of entry and enables the evaluation the value of the entrant as a going concern. Likewise, an annual focus on cash flow allows the model to track the cash needs of the investment and to estimate simple statistics such as the payback period. In Section III, we present a generic pro-forma study of hypothetical entry designed to address numerous concerns, while remaining manageable given the time and resources available in a merger review. III.MODELLING ENTRY IN MERGER ANALYSIS Theoretical considerations An entry model is a discounted cash flow analysis, wherein a project is evaluated to determine if it is likely to generate a positive NPV for the firm. In NPV analysis, a project’s incremental cash revenues and costs are projected into the future and their period differences, or net operating cash flows, are discounted back to the present using an interest rate representing the opportunity cost of capital. Incremental investment outlays, also discounted to the present as needed, are subtracted from the present value of the net operating cash flows to determine the net present value of the project. Positive NPV projects are ‘profitable’ in that they generate discounted returns in excess of the relevant cost of capital. Entry into a new product or service is highly unlikely to be undertaken unless the project generates a positive NPV for the firm. At the same time, just because an investment has a positive NPV does not mean management will undertake it. Responsible managers rank and prioritize potential discretionary investments, undertaking those projects having the highest NPV’s that fit the firm’s strategy, have acceptable uncertainties, and can be executed with available resources. Regarding resources, in business, all investments are subject to budget constraints, with legally required investments and those needed to keep the firm running usually taking priority over even the most value-enhancing discretionary projects. Still, clearly positive NPV investments meet the key initial screens in determining whether management will green-light them. This strongly supports the contention that, they are ‘likely’ to be undertaken from an antitrust perspective. This contention can be enhanced by feedback obtained from interviews with potential entrants, especially if interviewees reveal that entry has been or is being considered, details on relevant considerations compatible with the assumptions in the model, and what threshold NPV must exceed for entry to become a likely project. The NPV model requires information on the prospective business’ expected cash flows profiled over time and the appropriate discount rate for the entry project. Cash flows can be derived by estimating the elements of the new business’ operating statement and values for required investments in fixed assets, working capital, research and development and marketing necessary to start and scale the operation. These estimates can be based on a forecast of what is most likely to occur or built up as expected values from a series of predictions for potential scenarios whose various elements are adjusted by each scenario’s relevant probabilities. Best-practices require that the model embody a solid understanding of the project being analysed including the underlying strategy, consequent operations and their scalability, required investments, tax issues, and key operating and risk levers. The model should also include sensitivity analyses on the key parameters. The goal of the modelling effort is to provide a reasonable and complete (ie realistic) financial representation of entry within the constraints imposed by the available information. In business, a cross-functional team, including representatives from such areas as operations, production, purchasing, marketing, sales, and finance, would be put together to help prepare the entry analysis. For a pro-forma antitrust analysis of entry, a merger analyst would assemble relevant information from a diverse set of sources to build a reasonable and complete model. To that end, the analyst will have to review documents and data supplied by the parties and consult cognizant employees of the parties to fill in the blanks and validate the model.44 These employees may prove invaluable in helping the analyst identify reasonable estimates and assumptions to simplify the model and enable its timely completion. Information is particularly useful if one of the parties had actually undertaken an entry project in recent years. The analyst should also consult with management of the parties for insight on evaluating the results of the model. Contact with cognizant employees and responsible management will help keep the model from becoming a mere copy of existing business data presented in a new format and help prevent a mechanical evaluation of results. Throughout the process, care must be taken to avoid introducing bias into the results; in effect, the analyst must make a credibility finding on each assumption. Whenever feasible, the information should be checked with interview results from representatives of competitors and suppliers in the marketplace. In some situations, these competitors and suppliers may also supply information to build the model and thus, reduce the potential for bias. Modelling is an iterative process of gathering data, building the model, and validating its assumptions, structure, and results with cognizant and responsible personnel. Modelling begins by defining the details of the product(s) or service(s) required to deter or defeat the competitive effects of concern from the merger, along with the supporting operations of the new business. Based on the information collected, the analyst will estimate the output level associated with achievable and efficient scale, the time needed to reach that scale, the business’ operating structure, and the investments required. The analyst will project the direct and indirect costs of the product or service and estimate the costs of supporting activities. Then the analyst will forecast how these costs scale up with volume. The result defines the operating costs section of the entry model, including estimates for such inputs as labour; supplies and materials; various production and distribution overheads; and selling, general, and administrative services. To the extent that this analysis uncovers new questions not addressed in the initial research, the analyst will have to return to the various sources of information for more details. The cost analysis will help define the required investments in physical assets. Assumptions about growth through a start-up phase will help determine such intangible investments as pre-launch marketing campaigns.45 Other working capital investments can be modelled as a percentage of revenues. Revenues will be calculated using estimates of price and volume as the primary variables.46 Economists posit two theoretical approaches that can guide the analysis; the choice is determined by the fact situation associated with the merger at issue. First, in the differentiated goods model, the entrant sets price, and demand conditions (including the pricing decisions of the incumbents) determine the volume sold into the marketplace. In some cases, the pre-merger price of one of the merging firms would be a reasonable starting point, with some customers assumed to switch to the entrant, due to the expected higher post-merger prices charged by the incumbents. Alternatively, a price slightly below the post-merger level could be considered to reflect the short run opportunities to profit by undercutting the post-merger price increase. Over time, the price would be expected to evolve to the competitive level. The entrant is able to build sales by continuing to capture accounts from incumbents, earning a significant share of the new customers to the market, and growing its business with its initial customers. Sales gains can be tempered if incumbents are able to react to entry and expand output.47 However, if the anticompetitive price increase has alienated the merged firm’s customers, recovery of lost sales is likely to be difficult. Unless good evidence is available on the responses of the merged firms, it may be best to consider this factor in the robustness analysis. In the second, simpler, economic model, the product sold in the market is considered homogeneous, and the entrant chooses an output level. Price is then determined by the overall market demand conditions in combination with the output choices of the incumbents. In an anticompetitive merger, incumbents are expected to restrict output and thus create an obvious potential for the entry of new capacity. Short run market growth, the existence of inefficient capacity, and alienated customers searching for new suppliers could all create the potential for the entrant to obtain sales over time. The market price would need to be predicted from an understanding of demand conditions and expectations on the sales of the incumbents. Prices slightly below the post-merger level could be considered for early periods, with the level predicted to fall to the competitive level as the entrant’s sales grow. Alternatively, it could be reasonable to set price close to the pre-merger level, by assuming an immediate return to competitive conditions if the entry is large or incumbents are expected to quickly respond. Aggressive incumbent responses seem best considered in the robustness analysis. The long run volume available to any entrant depends on its share of the market and the overall demand for the product.48 Growth markets will obviously be more attractive for entry (all else equal), although static markets may represent profitable targets when the entrant can easily capture sales from incumbents. The model should project a time-period long enough to establish the potential for a profitable steady-state business. The financial results from this steady-state equilibrium can be used to forecast the continuing value of the project. The resulting model would represent the expected case, one based on the most reasonable estimates for the various inputs. Once the model is completed, the analyst can determine the NPV of the entry project given the appropriate discount rate. The rate for the project is the expected rate of return offered by comparable investment opportunities.49 Other important financial metrics can also be determined; including how long until the initial investments are recovered, whether the bulk of the return is in the discreetly forecasted period or in the continuing value of the project, and the internal rate of return (IRR).50 At this point, the analyst can begin to drill down into the major uncertainties inherent in business analyses, focusing on the valuation effects of key parameters, especially price and volume. In theory, the entrant sets either price or quantity. In practice, competitive realities will determine both the price and volume realized by the entrant. Thus, these two factors are difficult to accurately predict and largely beyond managerial control, and both are significant sources of uncertainty.51 They are also key drivers of the entrant’s profitability and of keen antitrust interest. Microsoft Excel’s Data Table tool allows the analyst to vary the two key input variables and calculate NPV at various combinations of the two inputs. By studying the change in profits associated with variations in price and volume, it is possible to obtain insight on the robustness of the model and address any concerns associated with market uncertainty.52 The price inputs to the Data Table should address a range of possibilities with particular focus on prices significantly lower than the price in the expected case to identify where NPV turns negative. It is also useful to consider some prices above the expected one to understand the upside of the project. For prices well below the expected price, the model gauges the risk to the project’s NPV of a major deterioration in demand conditions. Prices above the expected one highlight the potential value of improved market conditions to increase overall demand. Price ranges may also reflect uncertainty associated with the incumbents’ responses to entry in a homogeneous market. High prices could imply that the incumbents will choose to significantly restrict output, while low prices allow for the situation in which incumbents hold or even expand output in response to entry. For differentiated products, the price variation could also be affected by unexpected changes in the strategies implemented by the entrant’s new rivals, as well as customer acceptance of the entrant’s product. The volume inputs should also cover a significant range above and below the quantity achieved after the ramp-up period modelled in the expected analysis. The highest quantity considered in the Table should be the maximum realistic output achievable for the new product or service given the cost structure and investments built into the model. For example, stronger-than-expected demand conditions in the market imply a volume well above expectations. The lowest volume could also be linked to a much lower level of industry growth than expected, but in this case, the low growth depresses the output achievable at any price level. Incumbent responses may also justify the study of variation in volume for differentiated goods. A high volume could imply that incumbents do not adjust their prices in response to entry and the new firm can build a larger than expected market position. Low volume suggests the opposite; rivals lower prices in the face of entry and the new firm captures fewer sales than expected. Regardless of the cause of the variation, the Data Table tool will calculate NPVs for the various combinations of prices and quantities input by the analyst. The output in the table will reveal the lower bounds for price and volume at which the entry project is still NPV positive. More complicated simulations are also possible, although these tend to require multiple spreadsheets to evaluate. For example, both the initial pricing scenario and volume ramp-up can be customized to potential market conditions. Possibly, the entrant will have to sell at a significant discount to the market level to build share or maybe the entrant can match the higher price charged by the incumbents in the first period. Alternatively, various ramp-up levels for volume can be considered, with the firm building its market position slowly or quickly. Another analysis may assume that the firm benefits from economies associated with ‘learning-by-doing’. Here it is the cost structure that declines as the firm obtains experience in the marketplace. The NPV for these and many other scenarios can be easily simulated with minor adjustments to the entry model. In each simulation, the Data Table tool can illustrate the price and quantity effects on NPV, allowing the analyst to see the robustness of the entry prediction under each scenario. An example of entry analysis To operationalize the general principles outlined above, we present a detailed hypothetical example of an entry model. Assume a merger analyst interviewed knowledgeable personnel of the merging parties, reviewed various documents, validated his assumptions and inputs with responsible managers, and established the following model inputs for entry into the market for Product A: General Assumptions All cash flows occur at the end of their periods. Scale that can be realistically achieved in the market: 100,000 units per year. Timing of achieving that scale: 3 years with linear growth (33,000 in first year, 67,000 in second, 100,000 in third), followed by 2 per cent annual volume growth in the years after reaching 100,000 units. The appropriate real discount rate for this kind of project is determined to be 10 per cent. The discrete forecast period is five years, with a continuing value calculated for succeeding years assuming the 2 per cent yearly growth rate continues. Operating Assumptions All costs and revenues (and related cash flows) are incremental and stated in real dollar terms. The entry price is estimated to be $75.00 per unit, which compares with current market price of roughly $76.00. Unit variable cost is $12. First-year staffing costs, including benefits, of $200,000, growing linearly to $600,000 per year at 100,000 units. Other manufacturing overhead of $1,000,000 per year. Annual marketing and selling expenses of $800,000 in the first and second years, declining to $600,000 per year thereafter. IT costs total $500,000 in year 1 and $600,000 each year thereafter. Other general and administrative costs of $400,000 annually. Annual depreciation expense is estimated for simplicity to average $200,000/year for manufacturing and $100,000/year for non-manufacturing assets (an average straight-line depreciation over 10 years for manufacturing assets and 5 years for non-manufacturing assets). There are no investment or research and development tax credits.53 Depreciation in the continuing value column is increased to account for investment required to support growth in the years beyond the discreet forecast period (see, the last bullet of next section for explanation of this incremental investment). The tax rate is 39%. Tax losses are not used to offset future project net income in the model as their actual use depends on the tax circumstances of the entrant.54 Investment Assumptions Production equipment and leasehold improvements can be completed and put into service within one year and cost $2 million. The cash will be expended at time 0. Other equipment, including IT and distribution items, can also be obtained and put in service in one year and cost $500,000. Again, the cash will be expended at time 0. Incremental working capital, including inventory and other net short-term assets less short term liabilities, is estimated at $200,000 by time 0, growing at a rate of one-sixth of the change in annual sales thereafter. Introductory sales and marketing programs to be implemented later in the year during which the investments will take place will cost $800,000, expended at time 0. These are operating expenses under accounting rules. Applied R&D expenses to finalize commercialization will total $500,000 during the investment year. The cash will be expended at time 0. The model includes a total of $300,000 per year of capital expenditures in the continuing value column representing an amortization of the replacement of the initial capital investments.55 In addition, the continuing value column includes incremental capital expenditures to support sales growth. This capital is estimated as a percentage of incremental real revenue. Production investment is estimated to be 16 per cent of incremental revenues. Other investment is estimated at 4 per cent of incremental revenues. The basic model yields an NPV for the entry project of $14.4 million.56 As a positive NPV project, it increases the value of the firm and should be undertaken, assuming no financing constraints and, more importantly, that the information used in the model is seen as reliable. However, information is never perfect, rendering point estimates such as this one potentially problematic for making an informed business decision. Responsible management would want to understand the effect on entry’s NPV of changes in key value-driving factors, especially those over which the firm has little control. From an antitrust viewpoint, the most important factors to study are also difficult to predict: price and volume. The analyst uses Excel’s Data Table tool to analyse the NPV effects of the uncertain volumes and prices. The “Price/Volume Robustness Table, (see in Table A2)” shows the results of running the Data Table tool for the price and volume variables for this project.57 The basic model includes the expected real price of $75.00 for the duration of the analysis period, volumes of 33,000 and 67,000 for years 1 and 2, respectively and the steady-state volume of 100,000 units, achieved in Year 3. In order to use the Data Table tool, all annual prices and volumes in the basic model are keyed off single cells for each of these inputs (for price, the associated Year 1 figure and for volume, the Year 3, steady state figure). In Table A2, real prices as low as 70 per cent and as high as 110 per cent of that expected are assumed across the top in 5-percentage-point increments. These are the inputs for price. Steady-state volumes are assumed along the left side of the table, also ranging from 70 to 110 per cent of that expected and varying in 5-percentage-point increments. These are the volume inputs.58 The body of the table reports NPV for each corresponding price/volume combination. Negative NPVs are reported in Table A2 for 19 of the 81 price/quantity combinations, all located in the northwest region of the table where NPVs for significant price and volume shortfalls from expectations appear. At the lowest real price and volume combination tested, $52.50/unit and 70,000 Year 3 units (both constituting 30 per cent misses from expectations), the NPV is –$8.4 million. As expected, NPV increases as price and/or volume increase. Entry is a roughly zero-NPV proposition at about $53.25/unit and the expected steady-state yearly volume of 100,000. At the expected price of $75.00, NPV is positive at all tested volumes, showing that steady-state quantities could miss the planned volume by more than 30 per cent and still contribute to the value of the firm as long as price holds in the market. NPV is also positive at just under 75,000 steady-state units and a price of $67.50. That is a roughly 25 per cent volume miss and a 10 per cent price miss. At a price of $60.00, a shortfall of 20 per cent from expectations, NPV is still positive at a steady-state volume of a little more than 85,000 units, about 15 per cent below expectations. As long as management is confident that the odds of the shortfalls in price and/or volume that yield the negative NPVs in Table A2 are low, the project is clearly attractive. In this case, with few price/quantity combinations yielding negative NPVs and those that do representing very significant shortfalls from expectations, management should find this investment strongly attractive. There may be other factors whose estimated values are highly uncertain and could have a significant impact on the project’s NPV. The analyst can test the impact of these other factors for the various price and volume combinations in the expected case Table by making several copies of the entire worksheet and changing one key factor on each copy. The Table will update after the variable is changed. If only a small number of inputs in the base model, other than price and quantity, are materially uncertain and likely to have significant effects on NPV, the analyst can use separate worksheets to focus on the NPV effect of one or two of these variables and modify the Table for the highest and lowest expected values for the input(s) being tested. However, this approach will become increasingly cumbersome as the number of uncertain factors increases. The Scenario Manager tool in Excel is a versatile alternative that allows the analysis of changes to up to 32 base model cells. IV.SPECIAL ISSUES IN PROFITABILITY MODELLING The basic model presented above generates a reasonable approach to assessing likelihood of entry. It can be customized to specific environments. This may require consideration of either specialized theories that characterize the marketplace or unique fact patterns that affect the competitive process. Four examples are discussed below, although the list does not purport to be complete. In every case, careful analysis is needed to ensure that, when relevant, the special case considerations are included in the NPV modelling process. Impact of strategic entry deterrence Theorists believe that situations exist in which incumbents can strategically commit to post-entry strategies that affect the long run profitability of the entrant.59 To apply this type of model (which requires the use of a lower post-entry price in the NPV analysis), it is necessary to identify the commitment mechanism employed by the incumbents to lower price once the entrant has actually entered the market. To simply claim that incumbents will lower price in response to entry fails to recognize that the profit maximization calculus changes once the entrant actually sells into the market. By investing in sunk costs, the entrant commits to compete as long as it can cover avoidable costs and therefore must remain in the market even if the incumbents’ plan to materially lower price is implemented.60 Faced with an actual entry, incumbents must revisit their pricing decisions, unless the firms have also found some commitment mechanism. Therefore, profit maximization might require the incumbents to restrict output and accept entry. If prices remain too low, some plants will eventually exit, reducing capacity substantially, and returning the price to the competitive level. Here, it is possible that the merged firm may be the one required to close a plant.61 When strategic entry deterrence is credible, it needs to be considered by either adjusting the parameters of the model or reflecting on it as part of the robustness analysis. Consideration of customer-induced entry (‘Sponsored Entry’) Other situations exist in which customers may facilitate entry by agreeing to do business with an entrant prior to the introduction of the new product.62 Although customer interviews should always be used to aid in the estimation of the sales available to the entrant, the identification of specific customers with a clear willingness to switch a substantial share of business to an entrant serves to reduce the risk associated with the entrant failing to obtain material sales. For example, a customer with a history of dual sourcing would seem particularly likely to purchase some product from an entrant if its two sources of supply merged. The share of the smaller pre-merger source indicates the sales likely available to an entrant. Alternatively, a supplier might have a history of supporting new entry into a market and that history could quantify the level of sales available to the entrant. In any case, interview evidence associated with a specific level of baseline sales available to the entrant should be reflected in the model. Two approaches are possible to generalize the modelling process for customer support. When the evidence on customer support is strong, the parameters of the model can be adjusted to increase the entrant’s sales level. This would directly increase the likelihood of entry. If the evidence is weaker, customer support can be considered as a possible explanation for robustness analysis scenarios in which the volume exceeds that in the modelled case. Overall, customer support for entry should enhance the probability of a likelihood finding. Generalization of the model for repositioning In some cases, the basic entry model is applied to evaluate the potential for expansion of an existing competitor into the niche affected by the merger. The 1992 Merger Guidelines label this action ‘repositioning’, although the expansion may require the introduction of a new brand designed to serve the needs of the relevant customers. In modelling repositioning, a key issue involves the determination of the magnitude of the investments needed to produce the new product. Possibly, the firm only needs to invest to develop the new brand, along with the funds to cover the required working capital. Other assets, if underutilized in the firm’s current operations, may be able to be used to produce the closely related product. To the extent that the required investment is low, the firm is likely to earn a positive return from a relatively small position in the marketplace. Here a modeling analysis may require consideration of multiple modes of entry to determine if any approach is NPV-positive and likely to be implemented. Moreover, all NPV-positive modes of entry need to be considered simultaneously in a sufficiency analysis.63 Bias in the analytical process In some investigations, the staff may be forced to rely more on industry opinion than on actual business documents. This creates the potential for bias in the analysis.64 Parties to a transaction have incentive to provide opinions that imply entry is likely, while third parties, opposing the transaction, have the opposite incentive. Both problems can be minimized by careful analysis. The parties’ arguments are strengthened by increasing the agency’s confidence in the reasonableness and completeness of the model and by using knowledge of market conditions to quantify the probability of various price and quantity outcomes. Alternatively, third parties’ claims look more credible to the extent that they offer evidence that suggests the information provided by the merging firms is inaccurate. For example, customers may have experiences with past entrants and could offer insight on potential costs not noted by the merging parties.65 Finally, the information associated with market growth, economies of scale or scope, and excess capacities may aid in the analysis by verifying the credibility of the price and quantity projections. Issues with the objectivity of the evidence in a discounted cash flow analysis do not represent a problem unique to a study of the likelihood of entry. Similar problems exist for much of the evidence produced by the merging parties in a merger case. The best defence against this criticism is the reasonableness and completeness of both the analysis and the evaluation of the results. A model that is well-documented, put together with significant help from the parties’ documents and staff, evaluated with insights and guidance from the parties’ management and the economic experts, and informed by third-party input will go a long way toward addressing the concerns associated with over-reliance on any type of opinion evidence. In conclusion, it is possible to address a wide range of issues by generalizing the NPV model of profitability to consider the particular theoretical or factual concerns at issue. By conducting a detailed investigation into the competitive process within the relevant market, the merger analyst can design the likelihood model to focus on the key issues in the investigation. V.CONCLUSION In this paper, we argue that a good approach to evaluate the ‘likelihood’ of entry is to analyse the question the same way a responsible business manager would when considering any capital investment: using discounted cash flow analysis. If the financial analysis is reasonable, well informed, and complete and indicates a firm would profitably proceed with the investment, then there is a strong case for entry being ‘likely’ under the Merger Guidelines. We present an example of a comprehensive discounted cash flow model that addresses the major uncertainties of the project from an antitrust perspective. We also discuss how to evaluate the results of the model and explore various objections to and limitations of the modeling technique. Our approach provides antitrust practitioners and even business managers a useful tool for evaluating the likelihood of entry in the context of a merger. Timeliness and sufficiency considerations, introduced in the 1992 Merger Guidelines, remain relevant and are best undertaken prior to the likelihood analysis to allow the results of those studies to be integrated into the financial model. Likewise, the traditional analysis of Stiglerian barriers to entry should be part of the initial merger review and if significant barriers of that kind exist, likelihood modelling is obviated, because entry is, by construction, unlikely. Summing up, our proposed approach to entry analysis restructures the current methodology. Instead of following a sequential review of entry from timeliness to likelihood to sufficiency, the NPV approach focuses on the key issue of whether entry can be shown to generate sufficient return on the invested assets and addresses the entry-related considerations in a strategic order. The analysis would start with an overview of the classic understanding of Stiglerian barriers to entry to determine if incumbents hold a material cost advantage that would allow them to raise price. If no clear barrier to entry is found, the analysis would proceed to the evaluation of timeliness and sufficiency to obtain key information to be used in the model. This information includes a timeframe for the investment process from initial investment to sale of product to its effect on price and the details on the type of product to be sold to defeat the anticompetitive price increase. Finally, the likelihood of the entry question is addressed directly with a rigorously built financial model, coupled with a robustness review. Timely and sufficient entry that is significantly NPV-positive (ie entry that appears reasonably profitable from an economic standpoint and serves to protect competition in the market), should be considered likely and thus obviates antitrust intervention. The analyses and conclusions set forth in this article are those of the authors and do not necessarily represent the views of the Federal Trade Commission, any individual Commissioner or any Commission Bureau. We would like to thank H. Gabriel Dagen, Jeffrey H. Fischer, Jack B. Kirkwood, and Ronald L. Promboin for helpful comments on this article. Footnotes 1 See, US Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines (2010). <http://www.ftc.gov/os/2010/08/100819hmg.pdf> accessed 26 May 2017 at s 9 and European Union, Guidelines on the Assessment of Horizontal Mergers Under the Council Regulation on the Control of Concentrations Between Undertakings, 2004, <http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52004XC0205(02)&from=EN> accessed 26 May 2017 at para 68. Ease of entry is also relevant for all potential competition issues and some vertical transactions. 2 Predatory pricing in the EU does not formally require entry impediments, although the competitive concern is likely to be more substantial when entry is foreclosed. For a discussion, Miguel de la Mano and Benoit Durand, ‘A Three Step Structured Rule of Reason to Assess Predation under Article 82’ (2005) <http://ec.europa.eu/dgs/competition/economist/pred_art82.pdf> accessed 26 May 2017. 3 Even here, direct evidence of a price fixing agreement is needed to render evidence on ease of entry irrelevant. Evidence on easy entry remains a factor that can be used to reject an inference of a price fixing agreement, because it is illogical to conclude that a group of firms would enter into a price-fixing agreement when ease of entry dooms the agreement to failure. 4 Compare Bain (Barriers to New Competition (Harvard University Press 1956) 3) with Stigler (The Organization of Industry (Richard Irwin 1968) 67) or Schmalensee (‘Ease of Entry: Has the Concept Been Applied Too Readily?’ (1987) 55 ALJ 41–51, 44). Bain characterizes barriers as advantages that incumbents hold over entrants that allow those firms to price above the competitive level, while Stigler defines barriers as costs of producing product that must be borne by entrants, but not incumbents. Schmalensee notes barriers are factors that prevent entrants from eroding the economic profits of incumbents. 5 See Dennis W Carlton, ‘Why Barriers to Entry are Barriers to Understanding’ (2004) 94 Am Econ Rev 466–70 and Dennis W Carlton, ‘Barriers to Entry in Antitrust’ in Wayne Dale Collins (ed), Issues in Competition Law and Policy (ABA Book Publishing 2008). 6 Malcolm B Coate, ‘Theory Meets Practice: Barriers to Entry in Merger Analysis’ (2008) 4 Rev L & Econ 183–212. <https://www.degruyter.com/view/j/rle.2008.4.1/rle.2008.4.1.1240/rle.2008.4.1.1240.xml> accessed 26 May 2017. (hereinafter Entry Study). 7 Theorists were mired in their academic dispute throughout the 1980’s with Posner’s (Antitrust Law: An Economic Perspective (University of Chicago Press 1976) practical approach (with its focus on the time required for entry) appearing to have the most influence. By the 1990’s, Post-Chicago economists understood entry barriers could be endogenous while Chicago school economists highlighted the fragility of that theoretical result. The hopelessness of the dispute on the definition of barriers to entry was clear by 2004 (Carlton (n 5)). 8 US Department of Justice, Merger Guidelines, Antitrust Trade Regulation Report, No 1069, 1982 at s 3-B. Although the Merger Guidelines only applied to mergers, the general procedures could be applied to any antitrust case. 9 US Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines, Antitrust Trade Regulation Report, No 1559, 1992 at s 3. 10 Federal Trade Commission and US Department of Justice, Commentary on the Horizontal Merger Guidelines (2006) <https://www.ftc.gov/sites/default/files/attachments/merger-review/commentaryonthehorizontalmergerguidelinesmarch2006.pdf> accessed 26 May 2017. Examples of barriers to entry considered sufficient to preclude likely entry include access to key inputs (local nephrologists in DaVita-Gambro), intellectual property (broad patent portfolios in 3D Systems-DTM) and economies of scope (broad product lines of required tooling in Federal Mogul-T&N). Ibid at 43–45. 11 The entry studied identified seven examples of empirical analysis. Coate (n 6). 12 Merger Guidelines, (n 1) at s 9. 13 ibid. At least one revision appears potentially misleading. The 2010 discussion states that entry by a single firm replicating the ‘scale and strength’ of a merger partner is sufficient. Although obviously true, this should not be considered to set the standard for the magnitude of entry, because smaller levels of entry are often sufficient, as total cessation of operations on the part of one of the merged firms is extremely rare. 14 For instance, when the NPV of an investment equals $10 million, that means that the current value of the positive cash flows (eg the proceeds from the investment in the future) exceeds that of the negative cash flows (eg the initial investment cost) by an amount of $10 million. The discounting implies that cash flows occurring in the distant future carry less weight than cash flows received/used in the near future. 15 It is worth noting that NPV analysis is also used extensively as a prospective tool in other regulatory domains, such as access price regulation in network sectors. In the EU, it is further applied in the area of state aid control as a means to ascertain the likely impact of public subsidies on firms’ investment behaviour, cf Philipp Werner and Vincent Verouden, EU State aid Control: Law and Economics (Kluwer Law International 2017). 16 United States v Vons Grocery Co 384 US 270, 301 (1966). 17 This finding was supplemented by evidence of intense competition often driven by entry of new forms of food distribution, fringe competition, and expansion of existing rivals. 18 US v General Dynamics 415 US 486 (1974). Justice Stewart’s Von's Grocery dissent was also ignored by the Department of Justice (DOJ) in their 1968 Merger Guidelines, a document that memorialized the structuralist rules of the Supreme Court as the official policy of the government. Only special case situations of efficiencies were recognized as justifying an otherwise problematic merger. See, US Department of Justice, 1968 Merger Guidelines, <https://www.justice.gov/archives/atr/1968-merger-guidelines> accessed 26 May 2017. 19 US v Waste Management, 588 F.Supp. 498 rev'd 743 F.2d 976 (2d Cir 1984). 20 US v Calmar Inc, 612 F.Supp 1298, (DCNJ 1985); FTC v Promodes, 1989-2 Trade Cas (CCH) ¶ 68,688 (ND Ga 14 April 1989); US v Country Lake Foods, 1990-2 Trade Cas (CCH) ¶ 69,113 (D Minn 1 June 1990). See also, Echlin Manufacturing, where the plaintiff’s lack of evidence on both entry barriers (that block entry) and entry impediments (that delay entry) precluded finding a competitive concern. FTC v Echlin Manufacturing Co, 105 FTC 410 (1985). 21 US v Syufy Enterprises, 712 F.Supp 1386, aff'd, 903 F.2d 659, 667-669 (9th Cir 1990). 22 US v Baker Hughes Inc, 731 F.Supp 3 (DDC 1990), aff'd 908 F.2d 981 (DCCir 1990). 23 ibid at 987–88. The court also notes the very terms ‘quick’ and ‘effective’ appear ill-defined. 24 ibid at 989. 25 The 1984 revision of the Merger Guidelines tweaked the wording of the entry test. US Department of Justice, Merger Guidelines, Antitrust Trade Regulation Report, No 1169, 1984. 26 The FTC did not officially endorse either the 1982 or the 1984 Merger Guidelines, but noted the DOJ Guidelines would be given ‘considerable weight’ by the Commission. Staff generally adopted the spirit of the DOJ rules in their analysis (David T Scheffman, Malcolm B Coate and Louis Silvia, ‘Twenty Years of Merger Guidelines Enforcement and the FTC: An Economic Perspective’ (2003) 71 ALJ 277–318). 27 The details underlying the timely, likely, and sufficiency analysis were presented in papers such as William Blumenthal (‘Thirty-one Merger Policy Questions Still Lingering After the 1992 Guidelines’ (1993) 38 Antitrust Bull 593–642), Malcolm B Coate and James Langenfeld (‘Entry Under the Merger Guidelines 1982-1992’ (1983) 38 Antitrust Bull 557–92), and Jonathan Baker (‘The Problem with Baker Hughes and Syufy: On the Role of Entry in Merger Analysis’ (1997) 65 ALJ 353–74). 28 It is unlikely that mergers were regularly cleared when the entry was not ‘sufficient’ prior to the 1992 Guidelines. As the analysis focused on collecting evidence to ensure entry was able to ‘deter or counteract’ an anticompetitive price increase, issues with sufficiency were likely evaluated as special case complications within the basic algorithm. 29 Special case situations, such as scarce inputs could prevent the entry effect from scaling up in a market, while lumpiness of investment might lead to an integer problem affecting the ability of entrants to negate the full price effect. 30 The sufficiency analysis must combine the relevant effects from fringe expansion, repositioning, and de novo entry to evaluate sufficiency. 31 This approach economizes on time by not reviewing any style of entry that is not considered sufficient to protect competition. 32 Coate (n 6) at 202. 33 ibid. 34 Two mergers exhibited two barrier styles, so only 40 matters exhibited clear barriers to entry. 35 Net Present Value modelling had been applied years earlier in an FTC antitrust litigation. In the potential competition challenge of BAT Industries’ acquisition of Appleton Paper, the Administrative Law Judge (ALJ) (at 889–915) discussed a financial entry model, in the end concluding that the analysis did not meet the burden of proof standard required of the plaintiff. However, the staff was able to assemble a sophisticated entry analysis from the facts in the record. The Commission (at 942–46) reached similar conclusions to the ALJ, and implied the sensitivity of the model to the underlying assumptions would also be an issue. Although falling short of endorsing financial analysis, both the ALJ and the Commission showed an ability to deal with made-for-litigation studies. B A T Indus 104 FTC 852 (1984). <http://www.ftc.gov/sites/default/files/documents/commission_decision_volumes/volume-104/ftc_volume_decision_104__july-_december_1984pages_845_-_948.pdf> accessed 26 May 2017. 36 One matter involved both an identifiable entrant and financial analysis of entry, so only 17 of the 78 matters offered detailed evidence on the likelihood of entry and 61 matters did not. 37 The t-statistic is 2.9. The small sample of the current data (18) creates concern with the test and thus it should be repeated when more data becomes available. In two of these 18 matters with alleged, but not fully substantiated likelihood impediments, the investigation identified likely entrants, calling into question the finding that entry was actually unlikely. 38 The fiscal year 2006 break-point is artificial, but a review of the files shows limited activity prior to that point and more significant activity after that point. Possibly, this shift was caused by the publication of the Commentary on the Horizontal Merger Guidelines (above n 10). Although this document does not endorse the identification and evaluation of the most likely entrants, it does highlight the need for the likelihood analysis to consider evidence of failed attempts at entry. 39 From fiscal years 1993–2005, the staff found evidence of actual entry in 46 of 138 matters reviewed and identified impediments in 25 of these matters. The legal staff reported difficulty with the likelihood of entry in 16 of the 25 matters. For the 2006–12 period, 13 situations of expected future entry, and timeliness and/or sufficiency impediments were reported in 7. Staff never raised likelihood issues sufficient to preclude entry. The sample size is too small to support the finding of a significant difference. 40 If the entry is not sufficient in and of itself, then another analysis would presumably be undertaken to be sure it could be replicated and thus deter or defeat the price effect. 41 John Kirkwood and Richard Zerbe (‘The Path to Profitability: Reinvigorating the Neglected Phase of Merger Analysis’ (2009) 17 George Mason L Rev 39–117) do not find courts used ease of entry to dismiss a case. Possibly, this result is caused by the court’s broader consideration of supply-side flexibility, with some markets defined broadly when close rivals can quickly ‘enter’ the market in response to a price increase. (Regulators rarely find supply-side competition relevant to market definition and when they do, price discrimination arguments tend to negate the effect, leaving the issue to be addressed in ease of entry. See (Malcolm B Coate and Jeffrey H Fischer, ‘A Practical Guide to the Hypothetical Monopoly Test for Market Definition’ (2008) 4 J Competition L & Econ 1031–63). Alternatively, this result may be caused by sample selection. If the government never challenges a case without a colourable entry story, it would be unlikely to have a case dismissed solely on entry grounds. 42 ibid at 90–91. 43 The term ‘profit’ takes on different meanings in accounting and economics. Accountants define ‘profit’ as the excess of revenues over costs, with costs considered to include depreciation, amortization, and income taxes, but not a return to capital. Economists mirror the accountants’ definition, with theoretical proxies for costs used in place of accountants’ practical values. However, economists broaden the definition of costs to include the opportunity cost of the capital employed in the business. Thus, a business operation that earns an accounting profit less than that needed to cover the cost of capital is not profitable to the economist. Finance incorporates both concepts of “profit.’ Accounting profit is an input into an NPV analysis as the analyst is focused on incremental direct cash flows related to an investment in developing the NPV model. An NPV model is designed to guide management’s go/no-go decision by determining if the prospective project will likely increase the overall value of the firm. A project determined to be NPV-positive will increase firm value by generating more incremental revenue than is required to offset the investment, all incremental costs including income taxes, and the opportunity cost of the capital employed in the project, with all data stated in cash terms and the cash flows discounted back to the present. 44 FTC Second Requests include specifications asking for information on investments, revenues, and costs associated with entry into product lines of concern in the investigation. 45 We are using the term ‘intangible investments’ here in a corporate strategy sense, to represent upfront expenditures in activities that help ensure the success of the entry. Accountants (and tax authorities) may not treat the same outlays as investments, however. Such would be the case with the costs of ‘pre-launch marketing campaigns’. Accountants would treat them as expenses in the period in which the related services are rendered. Our example below shows how to model these expenditures. 46 Firms would be expected to estimate revenues using the same tools applied to any entry or expansion decision. The discussion below presents an overview of the issues that might be relevant for entry into either a differentiated or homogeneous goods market. We would not expect firms to use the tools of game theory applied to study theoretical industrial organization issues in their business decision-making. Of course, these tools and theories could guide the general analysis, as discussed below. 47 Strategic entry deterrence is discussed in the next section. 48 The information on long run volume is one of the most important inputs to the model, because the entrant’s volume interacts with cost conditions (minimum viable scale) to determine the profitability of the investment. 49 For this, the analyst can use input from the parties on the rate of return the firms would expect for projects of similar riskiness. In effect, the discount rate is an opportunity cost of capital, based on returns available to the firm from comparable investments. As a practical matter, the parties will likely suggest their weighted average costs of capital as the discount rate. The analyst may need to adjust these rates to suit the prospective conditions of the project being modelled (eg a start-up entrant may face a different discount rate than an established firm). 50 IRR analysis yields the same go/no-go recommendation as NPV in most cases (ie IRR > discount rate when NPV > 0). We recommend using NPV because it avoids various computational problems that arise in certain circumstances with IRR. 51 As noted above, economic theory generally expects the firm to choose either price and have the market set quantity (differentiated products) or quantity and have market conditions to determine price (homogeneous goods). 52 Although the parameters of the model may be defined to represent expected values, it is possible that some relevant considerations may be unknowable and thus best addressed as uncertainty in a robustness analysis. If the analyst feels confident in the parameterization, then the robustness results can be given little weight, although the results will still show the effects of possible error and imperfect knowledge of the future. 53 In real-world discounted cash flow analyses, tax considerations can contribute significantly to NPV estimates. Accelerated depreciation and various tax credits enable substantial reductions in cash outlays for taxes in the early operating years of a project, increasing estimated NPV, ceteris paribus. By ignoring these factors, our simplified model yields understated NPVs. 54 For example, a larger firm contemplating entry as a division may use the tax losses of the entering division immediately to offset gains elsewhere in the company. A start-up firm entering the business may use the losses in future years as it begins to have tax liabilities. The effect of this assumption is to understate NPV in the model. 55 This would likely lead to an understatement of the NPV of both the continuing value and the overall project. 56 Appendix A1 shows the model for the expected case. Cash flow is revenue less all incremental fixed and variable costs, including income taxes, plus depreciation, which is added back since it is non-cash. The NPV is the sum of the present values of the cash flows for the discretely forecasted years (1 through 5, calculated using Excel’s NPV function) and the continuing value less both the initial investments and Year 0 (pre-opening) operating results. 57 The results of the Data Table review are presented in the Appendix in Table A2. 58 The ranges used in the table can be as wide or narrow as the analyst desires, but should reveal a clear division between price and quantity combinations yielding positive NPVs and those yielding negative ones. 59 For a discussion of strategic entry deterrence, see Steven C Salop, ‘Strategic Entry Deterrence’ (1979) 69 Am Econ Rev 335–38 and Jean Tirole, The Theory of Industrial Organization (MIT Press 1988). 60 ‘Avoidable costs’ are the variable costs and fixed costs that will no longer be incurred upon exit from the market. Strategic entry deterrence may affect either differentiated or homogeneous goods markets. See for example, Inaki Aguirre, Maria Paz Espinosa, and Ines Macho-Stadler, ‘Strategic Entry Deterrence though Spatial Price Discrimination’ (1998) 28 Regional Sci & Urban Econ 297–314 for differentiated goods and Beth Allen, Raymond Deneckere, Tom Faith, and Dan Kovenock, ‘Capacity Precommitment as a Barrier to Entry: A Bertrand Edgeworth Approach’ (2000) 15 Econ Theory 501–30 for homogeneous goods. 61 For example, one of the incumbents may need to invest in new sunk costs to remain in the market and may choose to withdraw from the business if profits are low. 62 Economic theory models the possibility of customers facilitating entry into the market. See David T Scheffman and Pablo T Spillers, ‘Buyers’ Strategies, Entry Barriers and Competition’ (1992) 30 Econ Inquiry 418–436 and Andrew N Kleit and Malcolm B Coate, ‘Are Judges Leading Economic Theory? Sunk Costs, the Threat of Entry and the Competitive Process’ (1993) 60 Southern Econ J 103–18. 63 Here, great care must be taken to avoid the inference that the output provided by the smaller of the two merging parties must be replaced. In fact, only the output restriction associated with the price increase must be replaced to return the market to the competitive level. This point seems underappreciated in US v H&R Block, Inc, 833 F.Supp.2d 36 (DDC 2011) and US v Bazaarvoice, Case No 13-cv-00133-WHO (N D, Cal, 2014). In both cases, the reader gets the feeling that large scale entry comparable to the size of the acquired firm is needed, when in fact, the size of the entrant is related to the potential output reduction. Only economies of scale would mandate a larger entry and the large number of fringe firms in the relevant markets suggests scale is not an important factor. 64 Critics of the modelling process may object to an analysis that suggests entry is likely by asking how the model could be correct if entry has not already occurred. This objection ignores the role the merger itself plays in triggering entrepreneurial interest in entry. Pre-merger customers are often satisfied with their supply chains and not motivated to investigate alternatives. As a merger disrupts long-term purchase patterns, it often creates incentives for customers to consider new sources of supply. Customer search may stimulate entrepreneurs to look into entry and implies the marketing costs for entrants will be reduced as a result of the merger. Moreover, claims that the entrant will pick up sales are more credible in the post-merger world and much more credible if customers are faced with a post-merger price increase. Overall, the entry model is focused on a different post-merger marketplace and thus it is not surprising if the entry appears profitable. 65 Of course, the evidence provided by customers could also substantiate the information provided by the merging parties. Appendix A1. Entry model: data and analysis ENTRY MODEL EXAMPLE Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Continuing Value (CV) Present Value Investments Operating Production capital 2,000,000 224,970 Other capital 500,000 106,242 Working capital 200,000 412,500 425,000 412,500 25,000 25,500 26,010 Total Incremental Capital 2,700,000 412,500 425,000 412,500 25,000 25,500 357,222 Accumulated Capital 2,700,000 3,112,500 3,537,500 3,950,000 3,975,000 4,000,500 4,357,722 Revenues Units 33,000 67,000 100,000 102,000 104,040 106,121 Price 75 75 75 75 75 75 Revenue 2,475,000 5,025,000 7,500,000 7,650,000 7,803,000 7,959,060 Cost of Sales Unit variable cost 12 12 12 12 12 12 Variable costs 396,000 804,000 1,200,000 1,224,000 1,248,480 1,273,450 Comp & benefits 200,000 400,000 600,000 600,000 600,000 600,000 Other manufacturing costs 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 Mfg. depreciation 200,000 200,000 200,000 200,000 200,000 224,970 Total cost of sales 1,796,000 2,404,000 3,000,000 3,024,000 3,048,480 3,098,419 Gross margin 679,000 2,621,000 4,500,000 4,626,000 4,754,520 4,860,641 Sales & marketing 800,000 800,000 800,000 600,000 600,000 600,000 600,000 IT 500,000 600,000 600,000 600,000 600,000 600,000 R&D 500,000 Non-mfg. depreciation 100,000 100,000 100,000 100,000 100,000 106,242 Other G&A 400,000 400,000 400,000 400,000 400,000 400,000 Total SG&A expenses 1,300,000 1,800,000 1,900,000 1,700,000 1,700,000 1,700,000 1,706,242 Pre-tax operating income (1,300,000) (1,121,000) 721,000 2,800,000 2,926,000 3,054,520 3,154,398 Taxes (tax rate 39%) (455,000) (392,350) 252,350 980,000 1,024,100 1,069,082 1,104,039 Net income (845,000) (728,650) 468,650 1,820,000 1,901,900 1,985,438 2,050,359 Reverse depreciation - 300,000 300,000 300,000 300,000 300,000 331,212 Total cash flow (2,700,000) (845,000) (841,150) 343,650 1,707,500 2,176,900 2,259,938 2,024,349 NPV (discount rate 10%) Perpetuity value of CV 25,304,362 Present Value of CV 14,283,653 Present Value of Years 1–5 3,692,292 Present Value of Year 0 (3,545,000) Total NPV 14,430,945 ENTRY MODEL EXAMPLE Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Continuing Value (CV) Present Value Investments Operating Production capital 2,000,000 224,970 Other capital 500,000 106,242 Working capital 200,000 412,500 425,000 412,500 25,000 25,500 26,010 Total Incremental Capital 2,700,000 412,500 425,000 412,500 25,000 25,500 357,222 Accumulated Capital 2,700,000 3,112,500 3,537,500 3,950,000 3,975,000 4,000,500 4,357,722 Revenues Units 33,000 67,000 100,000 102,000 104,040 106,121 Price 75 75 75 75 75 75 Revenue 2,475,000 5,025,000 7,500,000 7,650,000 7,803,000 7,959,060 Cost of Sales Unit variable cost 12 12 12 12 12 12 Variable costs 396,000 804,000 1,200,000 1,224,000 1,248,480 1,273,450 Comp & benefits 200,000 400,000 600,000 600,000 600,000 600,000 Other manufacturing costs 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 Mfg. depreciation 200,000 200,000 200,000 200,000 200,000 224,970 Total cost of sales 1,796,000 2,404,000 3,000,000 3,024,000 3,048,480 3,098,419 Gross margin 679,000 2,621,000 4,500,000 4,626,000 4,754,520 4,860,641 Sales & marketing 800,000 800,000 800,000 600,000 600,000 600,000 600,000 IT 500,000 600,000 600,000 600,000 600,000 600,000 R&D 500,000 Non-mfg. depreciation 100,000 100,000 100,000 100,000 100,000 106,242 Other G&A 400,000 400,000 400,000 400,000 400,000 400,000 Total SG&A expenses 1,300,000 1,800,000 1,900,000 1,700,000 1,700,000 1,700,000 1,706,242 Pre-tax operating income (1,300,000) (1,121,000) 721,000 2,800,000 2,926,000 3,054,520 3,154,398 Taxes (tax rate 39%) (455,000) (392,350) 252,350 980,000 1,024,100 1,069,082 1,104,039 Net income (845,000) (728,650) 468,650 1,820,000 1,901,900 1,985,438 2,050,359 Reverse depreciation - 300,000 300,000 300,000 300,000 300,000 331,212 Total cash flow (2,700,000) (845,000) (841,150) 343,650 1,707,500 2,176,900 2,259,938 2,024,349 NPV (discount rate 10%) Perpetuity value of CV 25,304,362 Present Value of CV 14,283,653 Present Value of Years 1–5 3,692,292 Present Value of Year 0 (3,545,000) Total NPV 14,430,945 Assumptions are growth rate = 2%; production capital investment rate in continuing value = 16% of sales change in CV; and other capital investment rate in continuing value = 4% of sales change in CV. Table A2. Net present value by price and volume estimates Price/volume robustness table Initial real prices 14,430,945 52.50 56.25 60.00 63.75 67.50 71.25 75.00 78.75 82.50 Steady-state volume (Yr. 3) 70,000 (8,360,622) (6,628,912) (4,897,203) (3,165,494) (1,433,785) 297,925 2,029,634 3,761,343 5,493,052 75,000 (7,035,898) (5,180,495) (3,325,092) (1,469,689) 385,713 2,241,116 4,096,519 5,951,922 7,807,324 80,000 (5,711,173) (3,732,077) (1,752,981) 226,115 2,205,212 4,184,308 6,163,404 8,142,500 10,121,597 85,000 (4,386,449) (2,283,660) (180,870) 1,921,920 4,024,710 6,127,500 8,230,289 10,333,079 12,435,869 90,000 (3,061,725) (835,242) 1,391,241 3,617,725 5,844,208 8,070,691 10,297,174 12,523,658 14,750,141 95,000 (1,737,001) 613,176 2,963,352 5,313,529 7,663,706 10,013,883 12,364,060 14,714,236 17,064,413 100,000 (412,277) 2,061,593 4,535,463 7,009,334 9,483,204 11,957,074 14,430,945 16,904,815 19,378,685 105,000 912,447 3,510,011 6,107,575 8,705,138 11,302,702 13,900,266 16,497,830 19,095,394 21,692,958 110,000 2,237,171 4,958,428 7,679,686 10,400,943 13,122,200 15,843,458 18,564,715 21,285,973 24,007,230 Price/volume robustness table Initial real prices 14,430,945 52.50 56.25 60.00 63.75 67.50 71.25 75.00 78.75 82.50 Steady-state volume (Yr. 3) 70,000 (8,360,622) (6,628,912) (4,897,203) (3,165,494) (1,433,785) 297,925 2,029,634 3,761,343 5,493,052 75,000 (7,035,898) (5,180,495) (3,325,092) (1,469,689) 385,713 2,241,116 4,096,519 5,951,922 7,807,324 80,000 (5,711,173) (3,732,077) (1,752,981) 226,115 2,205,212 4,184,308 6,163,404 8,142,500 10,121,597 85,000 (4,386,449) (2,283,660) (180,870) 1,921,920 4,024,710 6,127,500 8,230,289 10,333,079 12,435,869 90,000 (3,061,725) (835,242) 1,391,241 3,617,725 5,844,208 8,070,691 10,297,174 12,523,658 14,750,141 95,000 (1,737,001) 613,176 2,963,352 5,313,529 7,663,706 10,013,883 12,364,060 14,714,236 17,064,413 100,000 (412,277) 2,061,593 4,535,463 7,009,334 9,483,204 11,957,074 14,430,945 16,904,815 19,378,685 105,000 912,447 3,510,011 6,107,575 8,705,138 11,302,702 13,900,266 16,497,830 19,095,394 21,692,958 110,000 2,237,171 4,958,428 7,679,686 10,400,943 13,122,200 15,843,458 18,564,715 21,285,973 24,007,230 Published by Oxford University Press 2017. This work is written by US Government employees and is in the public domain in the US.

http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Journal of Antitrust Enforcement Oxford University Press

http://www.deepdyve.com/lp/oxford-university-press/modelling-the-ease-of-entry-in-merger-analysis-can-financial-analysis-EwhcRrZNbZ

Modelling the ease of entry in merger analysis: can financial analysis move the ball?

Loading next page...

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher: Oxford University Press
Copyright: Published by Oxford University Press 2017. This work is written by US Government employees and is in the public domain in the US.
ISSN: 2050-0688
eISSN: 2050-0696
DOI: 10.1093/jaenfo/jnx013
Publisher site: See Article on Publisher Site

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Modelling the ease of entry in merger analysis: can financial analysis move the ball?

Modelling the ease of entry in merger analysis: can financial analysis move the ball?

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Modelling the ease of entry in merger analysis: can financial analysis move the ball?

Modelling the ease of entry in merger analysis: can financial analysis move the ball?

References (0)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies