The draft Merger Guidelines seek to reduce mergers and acquisitions, especially those that remove potential entrants. However, precluding acquisitions in those settings ignores what incentivizes startups and investors to take initial risks, as well as the advantages that large incumbents have to parlay acquisitions into further innovation and an array of widely commercialized consumer products. The overall effect may dampen innovation, write Ginger Zhe Jin, Mario Leccese, and Liad Wagman.
The draft Merger Guidelines released by the United States Department of Justice and the Federal Trade Commission (the Agencies) on July 19 feature many significant changes from earlier Guidelines. Of the 13 guidelines highlighted in the draft, two are particularly new and important for tech acquisitions. One is Guideline #4, which states that “mergers should not eliminate a potential entrant in a concentrated market.” The other is Guideline #9, which says that “when a merger is part of a series of multiple acquisitions, the agencies may examine the whole series” (emphases added).
While the draft Guidelines provide few details on #9, they do offer a list of evidence that the Agencies would consider in support of #4. For example, the Guidelines state that a firm’s “sufficient size and resources to enter,” expansion “into other markets in the past,” current participation “in adjacent or related markets,” being considered by industry participants as “a potential entrant,” as well as “subjective evidence that the company considered entering absent the merger” can all constitute evidence for the firm’s reasonable probability of entry. More importantly, a reasonable probability of entry is presumed to result in deconcentration or other significant benefits for competition, unless there is substantial direct evidence that the competitive effect would be de minimis. Simply put, a merger that is deemed to reduce a reasonable probability of entry is presumed to harm market competition.
Guideline #4 appears to hinge on the implicit assumption that, if not for mergers and acquisitions (M&A), all entities with a reasonable probability of entry would likely enter the market, vigorously compete with each other, and significantly promote market competition. Thus, mergers and acquisitions are superfluous methods of market entry that necessarily reduce competition. To avoid a linguistic debate on “reasonable,” “likely” and “significant,” it may be worthwhile to examine this assumption in a simple illustrative example.
A Stylized Example
Suppose a large incumbent A and a small startup B are currently the only two firms competing in a given “focal market.” Let us assume B is at some competitive disadvantage because A is more efficient in business functions such as marketing, finance, management, and customer acquisition; as a result, A has a dominant market share. Outside the market, there are three types of entities: Company C is a team of founders who may enter the focal market if the founders anticipate sufficient future returns from the potential entry. Company D is an incumbent operating in an adjacent or related market—for example, as a customer of or a supplier to A or B. Company E is an incumbent in an unrelated market but is large and resourceful. According to Guideline #4, incumbents such as C, D and E are all potential entrants with a reasonable probability to enter the concentrated focal market. Given that A and B are currently the only competitors, if A proposes to acquire B, it will create a monopoly in the focal market, at least temporarily, as it takes time and luck for C, D, and E to potentially enter, survive, and effectively compete with A. But that is not what Guideline #4 is about; rather, Guideline #4 challenges such M&A deals as A acquiring C, D acquiring B, D acquiring C, E acquiring B, E acquiring C, or even B acquiring C.
|A||Dominant incumbent in the focal market|
|B||Startup active in the focal market but less efficient than A|
|C||Team of founders considering entry into the focal market|
|D||Incumbent active in a market adjacent to the focal one|
|E||Large incumbent active in a market unrelated to the focal one|
According to Guideline #4, all of the aforementioned potential acquisitions would harm competition in the focal market, as compared to the implicit scenario where C, D, and/or E enter organically and all firms in the market exert significant competitive pressure on each other. This hopeful scenario is often less likely in light of the following five facts demonstrated by existing research.
Fact 1: Big Tech and Tech Startups are Disproportionately Responsible for Innovation and Economic Growth
Since the 1970s, large U.S. corporations such as AT&T, Xerox, IBM, and DuPont have gradually moved away from scientific research and towards the commercial development of this research, although investment in science has increased substantially in terms of public funding, the number of high-degree workers, and research articles published. The growing gap between basic research and commercial applications is in part filled by past and present venture capital-funded enterprises.
Since Guidelines #4 and #9 appear largely motivated by tech innovations and related acquisitions, especially those made by the five largest tech companies—Google/Alphabet, Amazon, Apple, Facebook/Meta, and Microsoft (collectively known as GAFAM)—we should first seek to understand the size and importance of tech, tech startups, and venture capital in research and development.
According to the National Center for Science and Engineering Statistics (NCSES), research and development (R&D) in the U.S. reached $717 billion in 2020, of which 76% came from the business sector. As shown by a 2017 Report by the Information Technology & Innovation Foundation (ITIF), the tech sector accounted for 79.1% of business R&D investment and 58.7% of R&D jobs between 2007 and 2017, where the report defines “tech” as a set of industries with sufficiently large R&D-to-sales ratio and a share of STEM workers that is twice the national average. These statistics suggest that a large fraction of R&D investment comes from tech. Ryan Decker, John Haltiwanger, Ron Jarmin and Javier Miranda show that tech or non-tech business startups contribute about 20% of U.S. gross job creation, and high growth startups account for as many as 50% of gross jobs created annually.
As for the role of venture capital, the 2023 National Venture Capital Association (NCVA) Yearbook indicates that total assets under VC management reached $1.12 trillion in 2022. In 2021 alone, $345 billion of venture capital was invested into 18,521 deals. Admittedly, not all VC investments are in R&D or tech, but software as a category is the largest recipient of VC investment and comprised 40% of all VC-backed deals in 2022. As of the end of 2022, companies that received VC as startups account for the seven largest publicly traded companies by market capitalization in the U.S., namely Apple, Microsoft, Alphabet, Amazon, Tesla, Meta, and NVIDIA. Out of these seven companies, five are ranked by Fortune in the top 20 of America’s “Most Innovative Companies” in 2023. In a more systematic study of publicly traded companies conducted by Will Gornall and Ilya Strebulaev in 2021, they find that those that were VC-backed account for 41% of total U.S. market capitalization and 62% of R&D spending within publicly traded companies.
In short, while it is true that VC investments cannot be compared apples-to-apples in relation to total R&D activities in the U.S., it is demonstrable that VC-backed startups, past and present, especially tech startups, play a substantial and crucial role in driving forward innovation, job creation, and overall economic growth.
Fact 2: M&A is one of the most important forms of capital liquidity, driving the funding, creation and growth of VC-funded tech startups.
For VC-backed startups, initial public offering (IPO) and M&A are the two most common means of successful exits. A survey conducted in 2020 finds that 58% of U.S. startups view being acquired as the long-term goal. According to the 2023 NCVA yearbook, 22% of U.S. IPOs from 2012 to 2022 were VC-backed. During the same period, the number of U.S. VC-backed M&As is 12 times that of VC-backed IPOs. These numbers suggest that, for aspiring entrepreneurs and their investors, M&A is an important, if not the most important means to reach capital liquidity and generates the incentives to enter and invest in the first place.
Fact 3. Tech M&As are not concentrated among a handful of firms or in a single sector.
Our own peer-reviewed research demonstrates that technology companies are acquired by a wide spectrum of public companies across the economy. In particular, among all public firms listed in North American stock exchanges, we find that 13.1% engage in tech acquisitions in a dataset compiled by Standard and Poor’s (S&P). It is common to observe firms operating in finance, health care, supply chain, trade, or services acquiring targets that specialize in internet content and commerce, software, mobility, or information management. Using Refinitiv’s classification regarding whether an acquirer’s core businesses can be regarded as “high-tech,” we find that a quarter of tech M&As recorded by S&P have a non-high-tech acquirer, supporting the argument that M&A is an effective way for entities that do not focus on technological innovation themselves to expand into new technology categories.
Furthermore, out of the 41,796 majority-control tech acquisitions that S&P recorded during 2010-2020, GAFAM accounts for 595, which is about 1.4%. On a per-firm basis, GAFAM firms are relatively more acquisitive, but some top technology acquirers, including private equity companies and other non-GAFAM firms, have matched or exceeded GAFAM in the volume of majority-control acquisitions per year since 2018.
Fact 4: Technology acquirers increasingly overlap with each other through M&A.
It is challenging to define precisely who is competing against whom in the tech space among all public and private firms. Instead of relying on ad-hoc market definitions, our research utilizes a technology taxonomy developed by S&P. Because the M&A data that S&P collects under this taxonomy have been widely used by investors in financial markets, the business areas identified by the S&P taxonomy can help identify potential and/or nascent competition that may take place in antitrust markets in or related to those business areas. We find that a GAFAM acquisition in a technology area is positively correlated with other firms also entering the area via tech M&A. If we examine M&A within GAFAM over the same 2010-2020 period, the five giants increasingly overlap in the extent to which they acquire tech targets within the same business areas. These data patterns point to the possibility that acquirers may increasingly compete against each other in the same business areas entered through M&A.
Fact 5: Most acquired firms in tech M&As fall outside the acquirer’s core area of business.
In the S&P taxonomy categorizing majority-control tech M&As, every firm is assigned to a level-1 parent category and a level-2 child ”business area,” which enables researchers to classify whether the acquirer and the target are in the “same” business area (same level-2), “adjacent” areas (same level-1 but different level-2s) or “unrelated” areas (different level-1s).
Based on S&P’s merger data during 2010-2020, we find that GAFAM and other top acquirers primarily acquire tech companies in order to expand into unrelated areas, although GAFAM acquisitions are less concentrated across level-1 tech categories than other top acquirer groups, due, in part, to an “acquire-adjacent-and-then-expand” strategy. Focusing on publicly traded companies, we find a similar pattern: the majority of targets in tech M&As fall outside the acquirer’s core area of business (defined by level-2 in the S&P taxonomy); and firms are, in part, driven to acquire tech companies because they face increased competition from other publicly traded companies (as defined by “product market fluidity,” a firm-specific dynamic measure of competition developed by Gerard Hoberg, Gordon Phillips, and Nagpurnanand Prabhala).
What do the aforementioned facts imply for the draft Merger Guidelines?
Let us first consider potential entrants D and E in our initial stylized example. By definition, E is a large resourceful company operating in a market unrelated to the focal market, while D operates in an adjacent or related market, and could well be a publicly listed company aiming to differentiate its products and/or diversify away from their core business area by acquiring startup B or C, as we have studied. To be responsible to their shareholders, D or E should compare the pros and cons between entry via organic growth and entry via M&A. According to Guideline #4, this consideration alone would qualify them as a potential entrant even if they are not large and resourceful. However, Facts 3/4/5 suggest that many public firms prefer to enter an unrelated business area via M&A rather than through organic growth. Thus, D or E may not necessarily enter via organic growth if M&A is disallowed, as per Guideline #4.
If D or E cannot enter the market via M&A, the financial returns that startup C could expect from entering the focal market would be much lower based on Fact 2, as it can only hope to survive via IPO or staying private. However, Fact 2 also implies that venture capital (or other private) investors would have the same discounted expectations of future profits and thus be reluctant to fund C prior to an IPO, which further reduces C’s chances of survival. All the above suggests that Guideline #4 may deter the potential entry of C, D and E, which is exactly opposite to the hopeful scenario the draft Guidelines implicitly assume and presumably aim to foster.
More alarming is the effect of Guideline #4 on the existing players in the focal market in our example. By definition, startup B is already in the market. However, if B cannot expect a successful exit through M&A with either D or E, it may have a lower chance of survival because it cannot leverage the expertise and resources of D or E. As a result, Guideline #4 does not only weaken B’s incentive to continue competing against A but also precludes a potentially more vigorous competition between A and the acquirer of B.
The new incentive that Guideline #4 introduces for the incumbent A is even more unfortunate. It enables A to claim to antitrust agencies that C, D, E are potential entrants, and therefore potentially deter their entry altogether, especially if entry via M&A is much more efficient for those firms than entry through organic growth.
We provide five empirical facts that together imply that Guideline #4 in the draft Merger Guidelines, as it is currently proposed, could profoundly distort the incentives of founders to create new firms, of venture capitalists to fund these new firms, and for established companies to leverage firm advantages to expand in new markets. We provide an example with five firms where all firms face new, distorted incentives as a result of Guideline #4, and these new incentives could deter potential entry and considerably diminish competition in the market under consideration.
Such unintended anticompetitive effects occur not just because the hopeful counterfactual scenario behind Guideline #4 tends to overestimate the potential entrants’ probabilities of entry through organic growth. Guideline #4 specifically provides a tool for a dominant incumbent to request the assistance of the antitrust agencies in deterring and handicapping current and potential rivals. This cannot be acceptable as part of a guideline intended to promote competition.
We recognize that acquisitions can lead to anticompetitive effects if they result in killer acquisitions, kill zones that effectively deter future entries, or complete foreclosures of competitors’ access to key inputs. But these possibilities should be carefully examined in light of empirical facts in each particular case and in comparison with alternative theories of harm as well as potential pro-competitive benefits and efficiencies. Merger Guidelines should not delineate a short list of circumstantial evidence with the understanding that the evidence covered in that list would automatically lead to a presumption of substantial harms to competition.
Related recent court decisions reinforce our concerns. In a recent 2022 case where the FTC challenged Meta’s acquisition of Within Unlimited (a startup that develops a fitness virtual reality app), the FTC argued that Meta was a potential entrant because it had sufficient size and resources to enter the dedicated fitness virtual reality market (a reflection of the proposed Guideline #4), but the court rejected this argument because it found that Meta had considered its own entry through organic growth but concluded that it did not have all the relevant expertise. The FTC subsequently withdrew its case. In another case, the FTC challenged Microsoft’s acquisition of Activision. The complaint alleged that as a result of the merger, the acquirer could gain control of top video game franchises, thus harming competition in high-performance gaming consoles and subscription services by denying or degrading rivals’ access to its content. In contrast, the court allowed Microsoft to proceed with the acquisition, arguing that the merger may in fact enhance consumer access to Activision’s content. The court also questions the FTC’s argument regarding a trend toward further concentration in the industry (a reflection of proposed Guideline #9), asserting that the FTC fails to explain how this trend is anticompetitive.
Assuming the Agencies can revise Guideline #4 to address the issues we delineate above, a systematic consideration of potential entrants will effectively require DOJ and FTC staff to function as a venture capitalist, predicting future market structure, future product development and future consumer preferences. The extent of resources that the Agencies would require to match the capabilities of the VC industry (which manages over $1 trillion in assets) is unclear.
The draft Merger Guidelines do not elaborate on Guideline #9, so it is difficult to ponder its potential unintended consequences and impact. One fact is worth considering: based on S&P’s tech merger data from 2010 to 2020, we find that the vast majority of tech acquisitions (81.56%) are consummated by firms that have completed prior tech M&As, and that the average time period between any two same-acquirer tech acquisitions is relatively short (525 days). This implies that a systematic evaluation of serial acquisitions would require substantial resources, even if Guideline #9 is free of any incentive loopholes and antitrust staff at the Agencies can replicate the due diligence functions of venture capitalists.
Authors’ Disclosures: Jin and Wagman worked full time at the U.S. Federal Trade Commission in 2015-2017 and 2020-2022, respectively. Wagman is an academic affiliate at the International Center for Law & Economics (ICLE). Jin was on academic leave to work full time at Amazon from 2019 to 2020. Each of them has provided consulting services to a few companies covered by the studies they authored and cited in this article. These companies may have a financial interest in the output of the Merger Guidelines. Part of these studies are supported by the Washington Center for Equitable Growth. The content and views expressed herein do not relate to any institution or organization with whom the authors are or have been affiliated.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.