Florence, 6 September
Thank you for the opportunity to speak at this year’s conference.
This event is one of the highlights on the antitrust calendar. And to hold it as every year in Florence is a guilty pleasure.
I would like to talk about our efforts to improve competition enforcement. The focus of the day is on abuses of dominance and, of course, mergers. Let’s start with mergers. A couple of days ago, the Court of Justice gave an interesting judgment in Illumina/Grail. It said we could not assert jurisdiction over mergers that are referred to us by Member States who, themselves, are not competent to review them. Except when those referrals come from countries that do not have a merger control regime at all.
So that puts an end to our acceptance of those referrals, which was a policy that we initiated three years ago. This marks the beginning of an important discussion. There needs to be a debate about what we should do now about killer acquisitions, I use the term “killer acquisition” as shorthand.
But as you know there are different types of possible anticompetitive effects when large players take-over innovative targets with low turnover. We will get to that in a minute. And then abuses of dominance are also a hot topic. Our first ever draft Guidelines are out! A consultation is ongoing. I hope you have all enjoyed the summer reading and that many of you will contribute to the consultation.
This is of course a major development in our policy work. It is an important step for the future enforcement of Article 102. These are very technical topics. But they raise a broad question. Abusive conduct and killer acquisitions are among the most critical competition challenges that we face today. So the larger question to be asked is, simply, as we face these issues, what competition enforcement do we think will most further Europe’s competitiveness? And how can our antitrust and merger regime achieve that? These are the important questions. To answer them, we first need to take the measure of the challenges that we face. Only then can we discuss the adequate response.
The state of competition has changed in the past 20 years. We have undergone waves of profound innovation. Not trivial innovation that improves everyday products. But innovations that have changed the way business is done and the way we live. Innovation has become the key factor of competitiveness in many sectors, that range from digital to biotech, pharma, chemicals and industrial products.
In an innovation economy, the focus of enforcement must be on creating and protecting opportunities for innovative firms. That means preserving a fair and open competitive process. A process in which innovators can challenge incumbents and disrupt established positions. But in many sectors, these conditions are not in place. Last June, DG COMP issued a lengthy report on this. The report provides evidence on the evolution of competition in the past 25 years. Its results are consistent with findings by other agencies and academics. Across industries, we observe increasing concentration, markups, profits and barriers to entry. Market power is often entrenched and durable. These trends are global. And they have costs. These costs are macro-economic: the study estimates a 5 to 7 percent reduction in GDP and lower productivity growth. They are also societal, as they translate into higher wage inequality. And they are strategic, because dependence on entrenched powerful firms is a risk for our resilience. This is harming small and medium firms that work in markets dominated by giants. The profit and productivity gap between industry leaders and other market players has increased. Smaller competitors and new entrants are not getting a fair shot. This is true in digital markets where players depend on large platforms to compete. But also in other manufacturing and services sectors. The main problem is not necessarily higher prices. The problem is reduced contestability, dynamism and innovation.
As competition enforcers, we have a responsibility to tackle this. How do we do it? By targeting the strategies and conduct by powerful firms that distort competition.
One such strategy is the acquisition by large firms of targets that may have no or little turnover. If such targets carry innovations, the issue is that they may be acquired by very large companies that want to acquire them to protect their market power. For instance, acquisitions might be done to kill an innovation that would otherwise threaten the large company’s core market. And this is why, in 2021, we broadened our policy to accept referrals of transactions from Member States whose notification thresholds were not met. Our interpretation has now been quashed by the Court. So we now need to reflect on the way forward. But before we do, we should review the experience of the past 3 years. In those three years, we have examined 100 below-threshold mergers. Those are transactions that we picked up as we monitored mergers and acquisitions across sectors. 45% of those cases were in pharma and biotech, 19% in digital markets. The rest took place in a wide variety of sectors.
That experience showed us a few things:
- In the pharmaceutical industry, we clearly see a trend for the acquisitions of targets that are highly valued despite the fact that they do not yet generate any turnover. Between 2023 and 2024, we identified 14 transactions with a reported value of more than 4 billion dollars that did not meet our notification thresholds, because of the target’s low turnover;
- In digital and tech, our monitoring also confirms that large sums of money are being paid by companies to acquire targets with little or no turnover. These targets are often active in nascent markets. That strategy is pursed by companies like Google, Apple, Amazon and Microsoft. But we also see large hardware and software companies, like NVIDIA, IBM and Qualcomm, that also acquire start-ups or small targets. Does that mean all these mergers are problematic? No: most of them raise no issue. In fact, the high value of these transactions alone is not related to competition concerns. On the contrary, the value of mergers that were referred to us by Member States and that did raise concerns was often less than 1 billion euros. Rather, what this shows is that turnover is not always the right measure to determine whether a transaction is economically significant.
Still, the number of problematic deals escaping our turnover-based jurisdiction is limited. This is no surprise. As we all know, many start-ups invest in innovation with the objective of being bought by large companies later on. So there is a positive incentive to preserve here. What we are after, is the specific set of circumstances where an acquisition might threaten innovation and distort competition. As is often the case in merger control, only a small minority of cases are problematic. Which is why, out of the 100 below-threshold mergers that could have qualified, we only reviewed three mergers in the last three years. These cases are Illumina/Grail in biotech, a case in semiconductors (Qualcomm/Autotalks) and another in financial services (EEX/Nasdaq).
Yet, our experience shows that this minority of cases can raise serious issues that need to be addressed. Judging from what other authorities have done during the same period, we don’t seem to have missed other important global mergers. Let’s take this in. There is a general trend towards high-value merger and acquisition that focuses on strategic promising targets that have little turnover. This is true across sectors, but it is particularly widespread in innovation-intensive pharmaceutical, biotech and digital markets. At the same time, the truly problematic cases are few. In the EU, over the past three years, we considered that 97% of the below-threshold transactions that we looked at more closely did not require a review. We need to consider our policy choices with these facts in mind. The first thing to say, is that a surgical, focused intervention may be the right way to go. The point is that, if most of these transactions are fine, and some might even be procompetitive, they should be left alone. What is needed is the ability to find that one fish in the pond that could be problematic, and to catch it in a way that is proportionate to the problem.
There are several ways to allow the Commission to find the right fish. None of them, however, seems optimal at this stage:
- we could lower our thresholds. But given the numbers that I have just discussed, doing so would overwhelm us with harmless mergers. And we would dramatically increase these companies’ administrative burden;
- we could assert jurisdiction based on the value of transactions. Certain Member States have adopted this solution. But again you risk catching a lot of mergers that raise no competition issues. And you might still miss the cases that really matter;
- we could also imagine introducing the power to call-in the notification of transactions when we have indications of competition risks. But that would require some tolerance in terms of legal certainty, which is exactly what companies and their advisors were complaining about in relation to our proposed interpretation of Article 22.
So what should we do?
The Court pointed in two directions:
- One direction could be for the legislator to change the Merger Regulation and introduce a “safeguard mechanism” to enable us to review below-threshold transactions. Such “safeguard mechanism” could even be a revision of Article 22. That revision could allow for the referral of sub-threshold mergers by Member States without jurisdiction in defined circumstances;
- Another direction is to rely on Member States to expand their own jurisdictions, thus allowing for more referrals under the “traditional approach” to Article 22. This is already happening. Eight Member States [Denmark, Hungary, Ireland, Italy, Latvia, Lithuania, Slovenia and Sweden] have introduced such powers in their toolbox. The way forward will not be mine to walk.
But I have full confidence in the future. Ursula von der Leyen has drawn up Political Guidelines. They advocate a new approach to competition policy. That approach focuses on the need to support companies to scale up, to ensure a level playing field and to protect innovation. The protection against killer acquisitions is a clear and explicit objective for the next Commission. So the work that we did over the past years will help the next Competition Commissioner chart this path.
Enforcement against abusive conduct is equally crucial to preserve competition and fairness in the marketplace. In the past 10 years, we have taken a series of enforcement actions in a novel setting. Today’s digital world is nothing like what anyone in 1957 could have imagined. It is quite remarkable that rules that were introduced more than 60 years ago proved fit to tackle things like digital ecosystems, two-sided platforms and the role of data.
In doing this work, we did not just keep up with market change. We often led the way, globally. We have constantly modernized our approach to antitrust issues. We have pushed the boundaries to make sure markets stay fair and open. In digital markets, we tackled self-preferencing and digital ecosystems. At times, we have explored and applied new theories of harm. At other times, we have used “classic” theories of harm, but in a new setting. And we have done so under intense judicial review. The current state of our practice against abusive conduct owes a lot to the input of the Union Courts. In the last two decades a lot has happened. With the Commission and National Competition Agencies enforcing Article 102, the Union Courts have delivered a large number of judgments on exclusionary practices. And yet Article 102 is one of the few areas in EU competition law that still does not have Guidelines. It is time to change that.
Our goal is for the new Guidelines to encapsulate the rich case law and our enforcement practice in this area. Guidelines will promote legal certainty and predictability. This is very important because our enforcement system is decentralized. Because of this, Guidelines would foster certainty and predictability at the national level as well. And of course, Guidelines should facilitate compliance by dominant firms.
We propose to achieve this with a few principles and objectives in mind. Let me underline some of them. First, let me be perfectly clear upfront: the Guidelines are not meant to mark a change of policy. They are meant to map-out the state of the law and the learnings from the Commission’s enforcement practice. We will not depart from the effects-based approach. On the contrary, the effects-based approach remains front and centre in the current draft.
At the same time, sticking to an effects-based approach does not mean you cannot learn from experience. And it certainly should not lead us to overlook the case law. The draft Guidelines lay out a coherent framework of assessment. That framework is based on the case law and our practice. This is what we call a “workable” effects-based approach. It reflects the existing body of law. And this body of aw allows enforcement to act to protect innovation, to foster consumer choice, to allow competitors to contest established positions. This is what we need to preserve going forward.
Let me give you two examples. First, there are certain types of conduct that are particularly harmful for competition. The Courts acknowledge that. The case law suggests that these conducts are subject to legal presumptions of exclusionary effects. On this basis, the draft Guidelines distinguish between different types of conduct:
- First, you have so-called “naked restrictions”. This is conduct that has no other economic rationale than to exclude. Such conduct can be presumed to have exclusionary effects. It is highly unlikely that it can be justified. That would be, for instance, a dominant undertaking destroying an infrastructure used by a competitor;
- Then, you have conduct that may not be as blatant, but still has a high potential to lead to exclusion. Those would be subject to a presumption, often with the need to meet a legal test. For example, imposing exclusivity conditions or offering predatory prices to entrench a dominant position.
When we design those categories of conduct, we try to be consistent. We propose to give a similar treatment to categories of conduct that threaten competition with a similar degree of intensity.
Let me give you a second example. The Union Courts have told us that quantitative tests may or may not be useful in our analysis, depending on the practice at stake. This means that these tests should not be required in every single case. And we think it is useful to clarify this guidance from the Union Courts in the Guidelines. Indeed, we do not intend to subject ourselves to a harsher burden than the one our judges consider adequate. Who would benefit from that?
But of course, creating these Guidelines is a complex process that takes time, and we need to get it right. Preserving an effective and workable antitrust enforcement framework is in everyone’s interest. That’s why the public consultation is so important. So I call on you and all stakeholders to provide feedback and comments to the text. It will help us develop these Guidelines, and shape the enforcement of Article 102 for the years to come. We will carefully consider all the feedback and discuss it at a workshop that will be organised in early 2025.
Let me close with a quick remark on procedure. Having clear substantive rules is crucial. How we enforce our rules—the toolbox that we have at our disposal—matters just as much. Yesterday we published the results of the evaluation of Regulation 1/2003. Regulation 1/2003 was a revolutionary piece of legislation when it came into force 20 years ago. It abolished a cumbersome notification system under Article 101 of the Treaty. And it effectively decentralized competition enforcement.
“Regulation 1” ushered the modern form of EU enforcement. The evaluation highlights this success. It also identifies areas where we can do better. Some will be very familiar to you. Antitrust investigations take a long time. Our investigative powers could be updated. The procedure for access to the file could be made more efficient. Our decision-making powers should also be considered. There is notable support for increasing the use of interim measures.
But I guess, what you really want to hear from me today is what is going to happen next. Will we revise Regulation 1/2003 and when? I am afraid these are questions that I cannot answer. Whether or not Regulation 1/2003 will be amended will be for the new Commissioner for Competition to decide. But I assure you that there will be opportunities for you to engage with the process.
To round up, I think it is fair to say that a profound updating of European competition rules is in motion. The work is not finished, but competition policy is a continuum. It is like jazz. You know that Miles Davis quote? “When I’m playing, I’m never through. It’s unfinished. I like to find a place to leave for someone else to finish it. That’s where the high comes in.”
So you see, unfinished work can be a gift. Thank you
Source – EU Commission