KNect365 Law is part of the Informa Connect Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.


Innocent Until Proven Guilty – Five Things You Need to Know About Killer Acquisi-tions

So-called “killer acquisitions” have been a focus of recent and intensive debate. The question relates to the impact of mergers on innovation, and how to deal with transactions in digital industries in particular. Based on feelings of unease but no empirical evidence, advocates of a more stringent review of deals in the tech-sector are already suggesting changes of the law, such as a reversal of the burden of proof. But it remains unclear which transactions may qualify as killer acquisitions, and in which circumstances one should, according to these voices, assume that they are really anti-competitive. Here are five things people should know about the current debate on killer acquisitions to form a view.

What are killer acquisitions?

There are at least two different types of transactions discussed under the headline “killer acquisitions”.

In the first scenario, a company that has an actual product on the market acquires a target business with a pipeline product that will likely soon compete with the acquirer’s product. In those cases, the transaction would qualify as a killer acquisition if the acquirer intends to discontinue (i.e., “kill”) the target’s pipeline product in order to avoid cannibalizing the acquirer’s own product’s sales post-transaction. This scenario has been the subject of several European Commission merger control decisions in the pharma and other sectors, and the focus of a recent influential economic paper on killer acquisitions in the pharma sector by Colleen Cunningham et al.[1]

The second scenario could be called “zombie acquisitions”: The acquirer acquires a target company that has an early stage product or service and intends to continue it (i.e., not “kill” it but “keep it alive”). The consequence is, however, the same as in any other acquisition of sole control, namely that the target will be controlled by the acquirer, and ceases to be an independent player, and possibly ceases to evolve into a major competitor that may “tip” the market. This scenario seems to be at the core of the current debate on whether antitrust agencies need new tools to deal with transactions in digital markets.

What is the issue?

The common theme around both types of killer acquisitions is the idea that there is a significant impediment to effective competition if incumbents buy up potential future rivals or rival products. It seems that this perception is particularly prominent in cases where an incumbent large tech company acquires numerous companies in its broader ecosystem. The pace of acquisitions in the tech sector led Commission Chief Economist Tommaso Valletti to state: “One acquisition every three weeks for the past 18/17 years – only one was vetted by DG Comp […] there is something there that doesn’t strike right.”[2]

In essence, the theory of harm is that incumbents prevent future competition by buying up potential rivals – (1) before they can become a threat, (2) before the acquisition becomes reportable under the current (typically revenue-based) filing thresholds, and/or (3) before reliable data becomes available to assess the transaction.[3]

Is there really a problem?

So apart from the unease perceived by some contributors to the debate: is there really a coherent theory of harm and a problem of such magnitude to form a basis for a fundamentally new application or change of the law?

The authors of the one paper that is widely cited by Commission officials, Cunningham et al., have analyzed more than 35,000 pharmaceutical drug projects in order to assess the magnitude of killer acquisitions in the pharma sector. Their findings suggest that 6.4% of all acquisitions were killer acquisitions. In particular, the authors find that discontinuation of a pipeline product (a “project”) is more likely when (1) an acquired project overlaps with a product in the acquirer’s existing product portfolio; (2) the acquirer does not face significant competition; or (3) the acquirer’s overlapping product will remain patent-protected for some time. Further, the authors found that killer acquisitions often happen to occur slightly below the merger control thresholds. The authors conclude that their “results caution against interpreting acquisitions of nascent technologies solely as incumbents’ efforts to integrate and foster entrepreneurial innovation. Instead, a significant source fueling this trend may actually be killer acquisitions that harm innovation.”[4]

Commission officials have quoted these findings in their contributions to the debate about killer acquisitions – even regarding digital markets, which are, of course, entirely different. However, for several reasons, it seems, to say the least, premature to call for new tools.

First, what conclusion should be drawn from the finding that 6.4% of pharma acquisitions are killer acquisitions? How should this number be interpreted? Comparing the rate of killer acquisitions to the Commission’s intervention rate in reported acquisitions (approximately 8.5% if simplified procedure cases are counted as well[5]), Tommaso Valletti noted that “6/7% – these are the things we should look into – it is not a small number.”[6] But why should the Commission’s intervention rate be the reference point?

Second, is 6.4% really the relevant number? According to the study, 49 out of a total of 758 acquisitions each year are killer acquisitions. This results in the quoted 6.4%. However, the authors of the study also note that a total ban of acquisitions of overlapping projects (25.5% of all acquired projects or 193 acquisitions per year in the sample; rate of development = 5.7%) would yield only seven additional projects for which development would have continued each year.[7] Seven “saved” projects out of a total number of 193 acquisitions of overlapping projects each year results in an “effective rate” of only 3.6%.[8] Taking the idea of banning acquisitions of projects even further (to include non-overlapping projects, which account for 74.5% of all acquired projects or 565 acquisitions per year in the sample; rate of development = 7.6%), this would result in an additional 10 projects for which for which development would have continued each year.[9] A total ban of acquisitions of pipeline projects would therefore result in 17 “saved” projects each year. “Saving” 17 projects per year yields an “effective rate” of only 2.2% of all acquisitions (44.9% of all projects or 758 per year[10]) and less than 1% of all projects (1,727 per year). Are those numbers that justify a whole scale revision of the rules, with the potential downside of stifling innovation? In our view, the jury is out.

Third, the innovation-stimulating effects of (even killer) acquisitions need to be considered in more detail. In fact, the authors note “[b]ecause killer acquisitions may motivate ex-ante innovation the overall effect of such acquisitions on social welfare remains unclear.” Venture capital and private equity investment into early-stage projects is critical. For this investment to remain attractive, venture capitalists and private equity firms must be able to exit their investment by selling to a strategic buyer.[11]

Fourth, the paper and its findings are limited to the pharma sector,[12] where the Commission has dealt with a number of mergers in which it assessed the likelihood of pipeline products being discontinued. No similar set of data exists for, e.g., the digital sector, where, as mentioned above, the concern is not only that products are discontinued, but also that they are continued under the acquirer’s ownership, and where some people seem to have a feeling of unease about this level of concentration. The advocates of new antitrust tools to deal with these concerns are well aware of these shortcomings.[13] The appropriate approach in this scenario should be caution; to take a step back and get the facts in order before pondering major changes to well-established principles under the current system.

Do we need new tools?

In essence, the advocates of a stricter stance towards potential killer acquisitions suggest two changes to the current merger control regime: (1) a revision of the merger control thresholds to capture acquisitions of early stage products/firms; and (2) shifting to the acquirer the burden of proof that the acquisition is competitively neutral or pro-competitive.

New thresholds? The debate about revising merger control thresholds to capture transactions involving nascent companies/products that do not generate sufficient turnover at the time of acquisition to trigger a filing has been going on for some time now. The obvious option is the introduction of a transaction value-based threshold.[14] The idea is that the purchase price better reflects the target’s (future) market potential than its current turnover. Austria and Germany have recently introduced such a transaction value threshold. The debate is still ongoing at the EU level.[15] Another proposal is to make a comparison between the objective value of the target and the agreed purchase price. The idea is that only an incumbent with lots to lose would be willing to pay a significant premium in order to preserve its position.[16]

However, much is still open to debate. At what level should a value threshold be set? Austria set its threshold at € 200 million, while Germany opted for double that with € 400 million. And if the threshold were to reflect not only the purchase price but also the target’s “objective value”: who would set the objective value of a target? And what premium over such value would trigger a filing obligation?

The most fundamental question, however, is whether any such revision would really enable the regulatory agencies to capture more problematic acquisitions and result in a more thorough review of potentially problematic transactions in the pharma or digital sector. Initial experience from Germany indicates that the introduction of the transaction value threshold has not led to more problematic tech deals being reported. According to the FCO, between June 2017 (when the transaction value threshold was introduced) and the end of 2018, there were only 10 filings based on the transaction value threshold, and all cases were cleared (unconditionally) in Phase I, and there were another roughly 20 consultation cases that did not end up in filings or where filings were withdrawn because the thresholds were considered not met.[17] These numbers are dwarfed by the total number of around 1,200 merger control filings in Germany each year. The situation is similar in Austria: almost one year following the introduction of the new thresholds, 13 filings have been submitted based on the transaction value threshold, and 20 informal consultations have been held. In the same time, the Austrian FCA received approximately 400 notifications based on the traditional thresholds.[18]

Reverse burden of proof? Typically, competition authorities must establish that a transaction will result in a significant impediment to effective competition if they are to be entitled to block or enact conditions in clearing a transaction. This is a good thing and in line with the general economic freedom of merging parties. Only in clear-cut cases based on sound and undisputed economic research should rebuttable presumptions be an option. Perhaps such a situation might be the rebuttable presumption of single dominance where a company’s market share exceeds a certain threshold, as, e.g., in Germany.[19]

However, some commentators consider the current approach to be a disadvantage for competition authorities in cases of acquisitions in an early stage of the target’s development. In particular, the lack of reliable market data is a concern. Jean Tirole asked recently whether past tech deals reduced competition: “My own gut feeling is: yes. But am I able to prove it? No. I have absolutely zero evidence on that. There is no way you can prove it because we don’t have any data.”[20] That is quite a remarkable statement, it neither suggests what the anti-competitive effects are supposed to be, nor disputes the fact that there is simply no data on which to base such a finding.

Nevertheless, in order to overcome this informational disadvantage and with the aim of capturing more problematic “killer acquisitions”, the second proposal by advocates of a more stringent review is a presumption that acquisitions of small targets by large incumbents are anti-competitive.[21] It would be for the acquirer to show the benefits and efficiencies of the transaction.

The details of this proposal, however, are far from clear. What is the necessary standard of proof? Should the presumption be limited to deals in the tech sector? And if so, how would this sector be defined? Or should the presumption also apply in other sectors, e.g., pharma?[22] Should it be limited to dominant acquirers? Or should some sort of “super dominance” – however that would be measured – be required? Should the presumption only apply for the acquisition of nascent or small companies? And how would those be defined? Are licensing deals covered (and if so, when)? Or are we only concerned with acquisitions?

Regardless of the details, this proposal must be rejected for a number of legal and factual reasons.

First, as mentioned above, presumptions need to be limited to the most evident and clear-cut cases. As such, the acquisition of a nascent company by a larger player is not such a clear-cut case. Indeed quite the opposite. The idea of large companies buying up potential competitors with the intention to shut them down seems paradoxical. In particular in situations where the target has already invested significant money into the project, and the purchase price is significant, common sense would indicate that it seems more likely than not that the product will be launched. Both in pharma and tech, there is usually room for differentiation, or even to aim to “expand the pie”. The finding that in <1% (or even 2.2%) of cases in the pharma sector an acquisition that may have the objective of shutting a potential future competitor down could (potentially) be prevented is simply not enough to justify a reversal of the burden of proof.[23] And even more so, the counterfactual – what would have happened but for the acquisition – is far from clear when a startup that is not yet generating significant revenues is in need of external investment.

Second, the acquirer may not have the necessary data to show the transaction will be pro-competitive, and it should not be required to do so. Experience from “regular” merger control cases where data is in fact available teaches us that merging parties’ efficiency arguments are rarely successful in practice. In essence, reversing the burden of proof would mean the incumbent acquirer would be tasked with proving a negative, i.e., proving that it is not intending to kill off a significant future constraint.

In a situation where there is a lack of reliable data, there is a significant risk that the assessment would turn on the parties’ internal documents alone. While always a tricky subject in merger control proceedings, this is particularly problematic in the case of acquisitions of upstream pipeline projects or nascent companies: sales projections in internal documents will typically be overly optimistic. On the one hand, the often substantial expense of R&D needs to be justified. On the other, a purchase price above the current objective value of the target requires some explanation. In essence, the target’s documents will typically focus on how they will “dethrone the buyer”.[24] Taken at face value, such statements could be used as evidence that the transaction will eliminate a successful future rival, and hence result in a significant impediment of effective competition. Further, proving that the target might not become as successful as predicted in internal documents will typically require a comparison to competitor pipeline projects. However, these are highly confidential. And even if such confidential information is available through a data room, the acquirer cannot reasonably report on it or argue with the content (the “data room trap”).

Third, the Commission is already today substantially stretching the boundaries of its existing tools. In pharma transactions, in an evolution starting from a traditional pipeline analysis focusing on late-stage pipeline products and current products, the Commission has gradually moved further upstream to now assess even early pipeline products and “overlapping lines of research”[25]. In addition, the Commission is also looking at innovation competition at industry level, without any reference to individual products. Both developments have been heavily criticized by lawyers and economists alike, and neither one has been tested in court.

Against this background, a call for a further change in a presumption of an SIEC is based on shaky foundations – and clearly goes too far.


A degree of competition authority paranoia regarding potential killer acquisitions is likely to remain on the agenda for the foreseeable future. Where a transaction meets the current filing thresholds, acquisitions by incumbents in tech, but also in pharma and possibly real estate, will likely be subject to stricter scrutiny under the existing merger control framework. Particular care should be taken by corporates when assessing potential transactions and in reports to senior management. Anything that could be perceived as a strategy of eliminating competition, including future competition, should be considered with great care. Speculations within the company to this end should not be endorsed, but avoided.

De lege ferenda, the introduction of a transaction value-based filing threshold appears to be the proposal with the greatest prospects of implementation. However, despite consultations several years ago, no concrete next steps are in sight at EU level. The proposed reversal of the burden of proof is still a nascent idea, and must be rejected. A cause for optimism is that the current debate will likely lead to further empirical studies before any steps will be considered. It is time to get involved!


Jacquelyn MacLennan
Partner, White & Case, Brussels, London

Jacquelyn MacLennan offers a wealth of experience for clients worldwide, representing leading multinational corporations, governments and trade associations.

She focuses on EU competition law, and clients benefit from her experience and insight in developing strategy and running major international cartel investigations, as well as advising on complex distribution problems and licensing arrangements, abuse of dominance cases, mergers and joint ventures.

Dr. Tilman Kuhn
Partner, White & Case, Düsseldorf, Brussels

Tilman's international experience across various locations includes merger control reviews, cartel matters, vertical conduct investigations, dominance matters, civil and appellate litigation and advisory work, as well as foreign investment control work. He represents clients across key industries, from oil & gas to chemicals, pharmaceuticals, consumer products, automotive and tech.

Prior to joining White & Case, Tilman worked in the Cologne office for a renowned US law firm, with a particular focus on German and European antitrust law. During this time, he spent one year at the Brussels office.

Juve states competitors are describing him as "good and well experienced, especially regarding merger control cases". He is a prolific author on antitrust law and is an active member of Studienvereinigung Kartellrecht.

Thilo-Maximilian Wienke
Associate, White & Case, Düsseldorf

Thilo advises on German and EU competition law, including merger control, vertical conduct, and dominance matters.

Prior to joining White & Case in 2019, Thilo worked for another renowned US law firm in Cologne and Brussels for four years. During this time, he was involved in a number of high-profile merger control proceedings, including The Dow Chemical Company's merger of equals with DuPont.

Thilo has particular experience in distribution law matters. He was involved in the recent revamp of a leading German branded goods manufacturer's selective distribution network, and advises regularly on vertical conduct.


[1]              Colleen Cunningham, Florian Ederer, Song Ma, “Killer Acquisitions”, 2018, available at

[2]             Tommaso Valletti, CRA Conference, 5 December 2018.

[3]              In the first scenario described above, the additional element of less consumer choice by “blocking paths that deliver innovation to consumers” comes into play; see Margrethe Vestager, Shaping competition policy in the era of digitization conference, 17 January 2019.

[4]              Colleen Cunningham, Florian Ederer, Song Ma, “Killer Acquisitions”, page 38.

[5]              The Commission’s intervention rate was approximately 8.5% during 2015-01/2019. However, the intervention rate is a noteworthy 28% when excluding cases dealt with by way of simplified procedure from the equation. This marks a significant rise from the previous intervention rates of approximately 19 and 18% in the period 2005-2010 and 2011-2014, respectively.

[6]              Tommaso Valletti, CRA Conference, 5 December 2018.

[7]             See Cunningham et al., pages 35/36: ((9.3%-5.7%)/8.3%) x (1-56.1%) x 25.5% = 4.8%. 4.8% x 8.3% x 1,727 = approximately 7 projects per year.

[8]             One would have expected that all killer acquisitions would have been avoided by a total ban of acquisitions of overlap projects. Applying the suggested rate of 6.4% killer acquisitions, the total number of “saved” projects would be expected to be at least 12 (i.e., 6.4% of a total number of 193 acquired projects with overlap).

[9]              Applying the same formula used by Cunningham et al. on pages 35/36: ((9.3%-7.6%)/8.3%) x (1-56.1%) x 74.5% = 6.7%. 6.7% x 8.3% x 1,727 = approximately 10 projects per year. Again, one would have expected the total number of “saved” projects to be at least 36 (i.e., 6.4% of a total number of 565 acquisitions without overlap).

[10]            See Cunningham et al page 35.

[11]            In 2016, more than 97% of all tech exits happened through acquisitions, see

[12]           The pharma sector is particularly well documented; see also Haucap/Stibale, DICE Discussion Paper 218 (2016) “How Mergers Affect Innovation: Theory and Evidence from the Pharmaceutical Industry”.

[13]            Jean Tirole, Shaping competition policy in the era of digitization conference, 17 January 2019: “That is something we economists must think more about.” Tommaso Valletti, CRA Conference, 5 December 2018: “The great work that has been done for pharma can be done also for these industries.” (Slides: “Disunion/proposals: 1. Systematically examine data for acquisitions, price paid, nature of business acquired, internal documents giving reasons for transactions (academia)”.

[14]           However, this is not the only way of capturing transactions involving targets with low turnover; see, e.g., Sabre’s acquisition of Farelogix, where the UK’s CMA is determining whether it has jurisdiction to review the deal under the UK’s market share threshold.

[15]           The recent expert report by Jacques Crémer, Yves-Alexandre de Montjoye, and Heike Schweitzer on “Competition Policy for the digital era” (4 April, 2019), commissioned by the European Commission, advocated against changes to the current EU thresholds for the time being: “While it is important to ensure that potentially anti-competitive transactions are duly scrutinised by competition authorities, one also has to consider the market need for legal certainty, as well as the need to minimise the additional administrative burden and transaction costs which an extension of jurisdiction would trigger. We therefore conclude that it is too early to change the EUMR’s jurisdictional thresholds; it is better for the time being to monitor the performance of the transaction value-based thresholds recently introduced by certain Member States, as well as the functioning of the referral system.” (Page 10.) The report is available at

[16]             Tommaso Valletti, CRA Conference, 5 December 2018.

[17]            We understand that approximately 1/3 related to pharma, 1/3 to digital industries, and 1/3 other industries.

[18]            Federal Competition Authority, Digitisation, Transaction Value Thresholds in Merger Control and Associated Challenges, available at

[19]            See, e.g., Paragraph 18 (4) of the German GWB: 40% market share.

[20]            Jean Tirole, Shaping competition policy in the era of digitization conference, 17 January 2019.

[21]           Differentvariations of this proposal exist. For example, the EU expert report on “Competition Policy for the digital era” (see above) notes that its proposal does not “does not create a presumption against the legality of such mergers” but “should help to minimise ‘false negatives’”: “The test proposed here would imply a heightened degree of control of acquisitions of small start-ups by dominant platforms and/or ecosystems, as they would be analysed as a possible defensive strategy against partial user defection from the ecosystem as a whole. Where an acquisition plausibly is part of such a strategy, the burden of proof is on the notifying parties to show that the adverse effects on competition are offset by merger-specific efficiencies.” (Page 124.)

While also expressly resisting calls for a reverse burden of proof as “not a proportionate response to the challenges posed by the digital economy” (para. 3.101 et seq.), the recent “Furman report” (Unlocking digital Competition, Report of the Digital Competition Expert Panel, March 2019) advocates a “balance of harms approach” (para. 3.88 et seq.) to be applied by the UK’s CMA. According to the report, “both the likelihood and the magnitude of the impact of the merger” should be weighed up. Mergers should then be blocked “when they are expected to do more harm than good.” The report is available at
However, the CMA’s Chief Executive Andrea Coscelli has already asked the Permanent Secretary for the Department for Business, Energy and Industrial Strategy and the Second Permanent Secretary to the Treasury “to discuss our concerns with you regarding this recommendation”. In particular, Coscelli warns of “practical challenges in applying this kind of test in a transparent and robust way” and “unintended consequences” of the “fundamental shift in merger policy” that would be brought about by the proposed changes. Coscelli’s letter is available at

[22]            According to Tommaso Valletti, CRA Conference, 5 December 2018, “people also mention real-estate as a potential source of problems.

[23]            However, advocates of stricter enforcement argue in favor of: “err on the side of competition”, see Jean Tirole, Shaping competition policy in the era of digitization conference, 17 January 2019. See also Heike Schweitzer, ibid.: “If we cannot resolve this uncertainty, maybe then the question becomes one of standard of proof. Can we then work with a standard of ‘more likely than not’, or should we then rather shift over to a situation where we consider error costs? […] Error costs is an important point to look at in the innovation context. […] Maybe, in this context, in this very dynamic context […] in a context where we have actors with a strong position, an entrenched position of market power, we should be less concerned with […] false positives and we should be very concerned with false negatives.” And as expressed by Tommaso Valletti, CRA Conference, 5 December 2018: “A small probability of future (intense) competition can be sufficient to make it optimal to block mergers”.

[24]            See Gil Ohana, Senior Director, Antitrust and Competition at Cisco Systems, Knect365, 12 February 2019.

[25]             See, e.g., Commission Decision M.7932 – Dow/DuPont, para. 3024.

Get articles like this by email