Introduction
It's great to be here today and to have the opportunity to speak directly to this audience. The tech and life sciences sectors are important industries, which bring many benefits to UK consumers and the wider UK economy. The Competition and Markets Authority (CMA) recognises the importance of these sectors and today provides the opportunity to engage in a two-way dialogue about the process and impact of UK merger control. So thank you all for having me.
International reach of UK merger control
1 . Could you please first explain the international reach of UK merger control. For example, we've seen the CMA take jurisdiction over the last few years on deals that are perceived to only impact the US and/or where the target has hardly any activities in the UK, or does not generate revenue in the UK - what's the basis for this and why has the CMA intervened in these deals?
When we look at the jurisdictional reach of the UK's merger control regime, it's helpful to consider 2 different aspects: what deals does the CMA have jurisdiction to review and when does the CMA have the power to block a deal.
Our jurisdictional test includes a turnover test - similar to turnover tests used for notification requirements in many countries - and also has an alternative "share of supply" test. That may be less familiar to non-UK businesses and advisors. So, broadly speaking, if the merger parties together supply UK customers goods/services above the threshold of 25%, with an increment from the merger, we are able to review the deal. The test was deliberately designed to be applied broadly and flexibly, and the UK courts have endorsed this reading of the legislation. The principal aim of the test, as articulated by the Competition Appeal Tribunal in Sabre v CMA is to identify a merger which does not meet the turnover test, but in respect of which there is a sufficient prospect of a competition concern arising from an overlap in a relevant commercial activity as to render it worthy of investigation. [footnote 1] I'd like to highlight a number of points about the share of supply test.
Firstly, in terms of identifying relevant goods/services, the share of supply test is different to a "market share" threshold that some merger control regimes have. In particular, and as explained in our Guidance on the CMA's jurisdiction and procedure (the Jurisdictional Guidelines), it does not require an assessment of the relevant economic market and there is no automatic read across to the substantive competition assessment. We will consider the commercial reality of merging parties' activities; this could, for example, include taking account of the life cycle of the supplies in question, as we did in Roche/Spark, where we considered the parties had a material presence in the UK market by virtue of pipeline products. And when assessing if the share of supply threshold is met, the share doesn't need to be measured in revenue terms. The legislation provides a non-exhaustive list of criteria that can be considered as part of the CMA's assessment - of which revenue is only one factor - for example, cost, quantity, capacity, and workers employed. [footnote 2] It makes sense not to focus solely on revenue, given there can be competitive interactions between players even without revenue generating activities. For example, in digital markets users are often offered services free of charge and companies can therefore have a presence in the UK - and serve UK consumers - without necessarily generating revenue in the UK, as we saw for example in Meta/GIPHY.
Secondly, there is no de minimis threshold when considering the increment needed to satisfy the share of supply test. In some of the cases we have looked at - for example Sabre/Farelogix - there was a significant existing share of supply on the part of the acquirer combined with a small increment. This approach should not take advisors by surprise - it is founded on a precautionary gating principle that acquisitions that may lead to increments to pre-existing high UK shares of supply (which may often reflect positions of UK market power) merit investigation and careful scrutiny by the CMA.
Finally, to satisfy the test, there needs to be a UK nexus. This is broader than just company location; otherwise, simply situating a business outside the country would allow you to avoid merger control over a deal affecting domestic consumers. So, we look at the arrangements in practice, for example, if services are being provided to customers located in the UK, or where goods/services are ultimately delivered, supplied, accessed, or used. We aren't limited by whether there is a direct contractual relationship between suppliers and customers, and the legislation allows us to look at making services available to potential users. [footnote 3]
So, in answer to the question, we've never asserted jurisdiction in relation to a merger that only affects the US - nor would we have any reason to want to do that. But we have looked at cases where the merger parties are serving UK customers but are based elsewhere, including the US, as explained. The UK is an important market for global businesses, so it is not surprising that in many cases a global business has activities in the UK.
But of course, just because we have jurisdiction to review, does not mean that there is necessarily a substantive competition concern and as mentioned the share of supply test does not have an automatic read across to the competition question. We have cleared mergers where jurisdiction was established on the basis of the share of supply test, for example, Roche/Spark and Google/Looker.
And it is worth recalling that the UK has a voluntary merger regime: merging parties can choose whether or not to notify a deal to us - and although we also have the power to "call in" deals that meet the jurisdictional test but are not notified, the CMA does not call-in every merger that meets the jurisdictional test. We use our review power proportionately. In addition to mergers that are proactively notified to the CMA, each year our mergers intelligence committee reviews between 500 and 700 transactions and considers whether to "call in" those transactions for phase 1 review. We only call in those mergers that have a reasonable chance of both raising competition concerns and meeting the jurisdictional test. In practice, therefore, very few of the mergers the CMA's intelligence committee looks at are "called in"; for example, in 2020 to 2021, only 7 of the 550 mergers reviewed were called in.
As I mentioned, it is important to note that there is also a jurisdictional element to our substantive assessment: where we have jurisdiction to review a deal, we can only take action where we identify a substantial lessening of competition in the UK. So there is in effect a requirement for a UK nexus at this point too; we have to show that the deal creates competition concerns in the UK (reflecting our duty to protect UK consumers).
Finally, the UK government announced last week a broad set of reforms to the UK competition and consumer regime following a public consultation last year. The changes are subject to legislation and therefore will take some time and are not anticipated to come into effect this year. But taking a forward look, the changes to the merger regime include introducing an additional test to determine jurisdiction aimed at "killer acquisitions" and other mergers that do not involve direct competitors, such as vertical and conglomerate mergers. The test would remove the need for an increment if one of the merger parties has a significant UK presence, established where the party has both a share of supply of 33% and a UK turnover of at least £350 million. There will be a specific UK nexus criterion, details of which are to be set out.
These reforms are distinct from the proposed reforms to establish a separate merger regime for firms with 'strategic market status' - the large digital players - whereby these firms would be subject to closer scrutiny. We are waiting to see which proposals the UK government decides to take forward. But in any event our current powers enable us to review digital mergers closely where we have the jurisdiction to do so under the current regime and where potential competition concerns arise.
Innovation competition in the tech and life science sectors
2 . How does the CMA analyse a tech or life sciences merger's effects on innovation competition? Does the CMA see its approach as more expansive than other leading authorities around the world? In particular, what is the CMA's position on acquisitions by digital platforms?
It's worth noting at the outset that the importance of innovation efforts as a parameter of competition (and the critical role that merger control plays in protecting innovation) is well-established in the CMA's decisional practice, the practice of other competition / antitrust authorities and the economic literature. There has been a sharper focus on assessing innovation competition in recent years given the nature of markets in which innovation competition is a feature.
The CMA has assessed the loss of innovation efforts in previous mergers, including in relation to digital markets and the life sciences sector. We have reflected the importance of innovation competition in our updated Merger Assessment Guidelines (MAGs). Our guidelines draw together decisional practice over the years reflecting our improved understanding of dynamic markets and recent contributions in the economic literature.
But the CMA isn't an outlier in this regard. There is broad consensus across leading competition authorities on the policy approach on innovation competition. The European Commission's horizontal merger guidelines (which date from 2004) discuss the importance of innovation as a parameter of competition. The FTC/DOJ horizontal merger guidelines (published in 2010) provide for the assessment of innovation, and there is a public inquiry underway to update the guidelines in light of the unique characteristics of digital markets. And in 2021 the CMA, Australian Competition and Consumer Commission (ACCC) and Bundeskartellamt issued a joint statement on merger control enforcement explaining the potential competitive harms that can arise in dynamic markets where innovation is a feature. The joint statement emphasised the importance of robust scrutiny of mergers where dynamic competition is a feature, explaining that the inherent uncertainty in these forward-looking assessments does not mean that a potentially anticompetitive merger should be cleared. The new leadership of the DOJ and FTC have expressed very similar policy positions in a number of recent speeches. And this is all grounded in the economic literature. [footnote 4]
So we don't see our approach as more expansive but rather it is built on established principles and applied similarly to other leading authorities.
Looking at acquisitions in digital markets in particular, innovation competition is particularly important in these markets. The importance of innovation in digital markets in particular was highlighted in the Furman Review, a report prepared by a digital competition expert panel for the UK government. Markets with digital platforms are typically highly concentrated markets with features such as high barriers to entry due to network effects. This can result in high market concentration, such that market power is easily created or entrenched, and is likely long-lived - and that we are therefore concerned even about an incremental loss of competition. Therefore, the major digital platforms should expect close scrutiny when undertaking acquisitions. And we would encourage those companies to engage early with the CMA in respect of their proposed mergers.
Of course, none of this means that the UK - or CMA - is anti-tech or anti-innovation. Quite the contrary. We think it is appropriate to apply a high degree of scrutiny to acquisitions of nascent competitors by large digital players so that digital markets continue to develop in a way that fosters innovation and competition, stimulating growth and benefiting UK consumers. This is advantageous for the many companies seeking to grow their businesses in these markets (and for investors in these companies, including those considering exit strategies). And close scrutiny does not amount to a policy of blocking competitively benign mergers; we have recently cleared several mergers involving major platforms that did not give rise to competition concerns, including Google/Looker, Amazon/Deliveroo and Facebook/Kustomer.
The CMA's approach to future and dynamic competition
3 . Can we just laser in on the CMA's distinction between potential/future competition on the one hand and current/dynamic competition on the other, as it has very important practical significance. On tech and life sciences deals in particular, we often hear acquirers tell us that the company they're planning to buy is an innovator but isn't active in the relevant product market yet and may well not have successfully entered in the future absent the transaction. Does that uncertainty mean the CMA will be fine with the deal?
When looking at mergers involving a potential entrant, we can look at the loss of competition in 2 different ways.
One is to look at the loss of future competition which involves an assessment of the likelihood and impact of that future competitive entry.
But there can also be a more immediate loss of competition because existing firms and potential competitors can interact in an ongoing dynamic competitive process even before entry actually occurs, driven by efforts to enter or expand in a market. This competition in innovation may result in improved competitive offerings from the potential entrant, the acquirer or other market participants, for example as incumbents may be incentivised to respond to these efforts to prevent a future loss of profits. This process of dynamic competition is particularly important in innovation markets.
With regard to dynamic competition in particular, there are several key principles underlying how we assess dynamic competition, as set out in the MAGs and applied in our recent practice:
- Firstly, as explained in our MAGs, we consider the importance of dynamic competition in the context of the structure and functioning of the specific market. The MAGs note that: "Where there are few existing suppliers, the merger firms enjoy a strong position or exert a strong constraint on each other, or the remaining constraints on the merger firms are weak, competition concerns are likely. Furthermore, in markets with a limited likelihood of entry or expansion, any given lessening of competition will give rise to greater competition concerns." [footnote 5] Context is crucial. For example, we are naturally and quite properly concerned where the incumbent acquirer already has substantial market power; efforts to enter and expand may be particularly important in this context. We also may be concerned where entry involves significant costs or risks, or where key aspects of the final product are determined during the investment phase, as dynamic competition may be critical in these markets. These issues were assessed, for example, in Meta/GIPHY where we identified the important innovation efforts of GIPHY and the impact that these could have on Meta and other market participants in UK display advertising even before GIPHY had launched its novel advertising product in the UK. It was this loss of dynamic competition that underpinned our horizontal SLC finding, particularly in the context of Meta's existing and significant market power.
- Secondly, you mentioned the role of uncertainty. There will always be uncertainty about the outcome of innovation efforts absent the merger, for example, if new products will ultimately be made available to consumers. As explained in the CMA/ACCC/Bundeskartellamt joint statement, competition agencies face a difficult challenge when taking a view on future market positions and company actions, and the forward-looking nature of merger control review will always mean competition agencies face some uncertainty when making such decisions. Indeed, it is the uncertainty of outcome that can drive dynamic competition and is inherent in driving dynamic competition. As explained in our MAGs, a process of dynamic competition can be valuable even when the outcome is uncertain because it increases the likelihood of new products or innovations, and that is in a sense what we are looking to preserve. That's why it's particularly important for us to preserve the scope for incremental competition in dynamic markets and to protect the process of dynamic competition that's driven by innovative independent entry. And, as explained in the joint statement, uncertainty as to the future should not necessarily mean that potentially anticompetitive mergers are cleared because of that uncertainty: a seemingly small transaction can cause a competitive market to tip in an anticompetitive direction.
- Notwithstanding any uncertainty, our approach is grounded in evidence and when assessing the impact of a merger on dynamic competition, there is a range of evidence we can consider, including business plans and other internal documents of the merging parties, evidence of steps taken towards entry/expansion, evidence on the extent to which the new product would compete with existing products (including evidence from customers), and evidence on the incentives or actions of incumbent firms to respond to these efforts.
- We'll also look at other factors that play a role in assessing how important a loss in dynamic competition is, for example, the ability and incentive for other potential competitors to enter or expand. We've cleared mergers on this basis too. For example in Roche/Spark we found that there were other suppliers developing a gene therapy treatment and concluded that therefore Spark did not offer particular clinical or commercial advantages over others. We took a similar approach in Amazon/Deliveroo in respect of one of the theories of harm - although the future development of the market was uncertain, and we considered the parties were likely to have expanded and competed more closely, the evidence showed us that other market participants would be well-placed to compete. And since our investigation, we have, in fact, seen significant expansion by a number of players.
Global mergers and international cooperation
4 . Let's assume the CMA and one or more other agencies are reviewing a global merger. How important is it to the CMA to be seen to reach a coordinated, consistent and synchronised outcome with other authorities? Is there a risk of divergence and what is the CMA's view on these divergence risks given the impact this can have on deal certainty, particularly for worldwide deals? How is the CMA's relationship with Brussels post-EU Exit and how does that relationship compare with other leading authorities such as the FTC/DOJ?
We recognise that in global mergers coordinating on merger review and remedies where applicable is desirable and efficient for both merging parties and authorities, and we will strive to achieve this as far as possible. Indeed, we note this in our Jurisdictional Guidelines, which we updated following EU Exit. But it's important to note that consistency is not an end in itself. The CMA does not seek, or seek to avoid, divergent outcomes but will seek the outcome guided by the evidence that is right for UK consumers. That's our priority, consistent with our statutory duties.
The CMA has always worked closely with other competition authorities around the world including the FTC/DOJ and European Commission. There are great examples of cooperative working in Illumina/PacBio, ThermoFisher/Gatan and Sabre/Farelogix. Following EU Exit, the scope, scale and intensity of that cooperation has increased, and we have a close, indeed, closer relationship with the European Commission and other leading authorities such as the FTC/DOJ post-EU Exit. We have cooperated both with the European Commission and the FTC/DOJ very well in recent cases, for example, Nvidia/Arm, which was abandoned earlier this year and Cargotec/Konecranes where we collaborated extensively with the DOJ, ACCC, the European Commission and others. All of these authorities have highlighted the close and constructive collaboration between the CMA/DOJ/ACCC/European Commission throughout the review until the merger was abandoned. A good example in relation to remedies specifically is Stryker/Wright, where the parties offered a divestment package to address competition concerns in both the US and the UK, we extended our timetable for considering remedies in order to align with the FTC's timetable and ensure that any remedy accepted by the CMA was also acceptable in the US (and vice versa). Another more recent example is S&P/IHS where we worked closely with the US and the European Commission on a material divestment package.
Remedy outcomes in global mergers
5 . Thanks for mentioning Cargotec/Konecranes. The case reminds dealmakers that we have to recognise though that on complex deals involving negotiating potential global remedies with multiple authorities to resolve concerns in global markets, there's always a risk that even if the authorities agree on the diagnosis of the problem, they may diverge on the remedy they require. Although not a tech or life sciences case, Cargotec/Konecranes is a very recent example. What can you tell us about the case that produced this divergent outcome on remedies?
As you note, Cargotec/Konecranes is an example where the CMA, DOJ, ACCC and European Commission all raised similar concerns regarding the merger. All authorities applied a broadly similar analytical framework, and it is relatively rare for there to be divergence based on analytical approach.
But differences can arise in evidence and reflect rational differences in context or effect, for example, differing market conditions, or regulatory regimes. In Cargotec/Konecranes, the CMA rejected the proposed remedies. Although the DOJ's and ACCC's investigation did not finish before the merger was abandoned, both authorities also indicated that they had concerns with the remedy and the DOJ told the parties that it was preparing a challenge to the transaction. The European Commission came to a different conclusion on remedies based on its evidence from players on the market, reflecting to some extent differences in the underlying evidence base. Importantly, we also had concerns about the composition of the remedy package, which involved carving out assets from the merging businesses' existing operations and knitting them together into a new combined business. We saw that as complex and risky.
As mentioned, our priority is to get the right outcome for UK consumers not to avoid divergence per se. And it is worth re-iterating that divergence on the basis of the same global facts where conditions of competition are global, and there are no UK-specificities to consider is relatively rare. None of this affects the close collaborative relationships between competition agencies, conducted with a view to each achieving the right outcome for consumers in the markets they protect.
The CMA's approach to behavioural remedies
6 . Mergers in the tech or life sciences space can create ecosystem concerns, in particular, when the merger gives the acquirer privileged access to a set of commercially valuable data or an ability to leverage from one product/service area into another. Where does the CMA stand on the use of behavioural remedies such as promises to grant fair and equal access, rather than business divestitures, to resolve competition concerns?
The CMA has conducted a significant amount of analysis on the success of merger remedies, conducting a detailed evaluation of over 18 case studies on an ongoing basis over the last 15 years, spanning structural remedies such as divestiture, behavioural remedies such as price controls and vertical separation, as well as intellectual property and licensing remedies. We are quite unique globally in the amount of ex post analysis we have done on merger remedies to learn from previous experience and refine our approach. This gives us proper standing globally in the discussions about appropriate remedies in merger control.
In relation to behavioural remedies, the CMA has stated together with the ACCC and Bundeskartellamt in the joint statement on merger control that in dynamic markets - like in the tech or life sciences space - it favours structural remedies over behavioural remedies. This is not a change to our standard for assessing remedies but reflects our long-held view, as articulated in our Mergers Remedies Guidance, that behavioural remedies are less likely to effectively address competition concerns, supported by the ex post research we have done as mentioned earlier. The case against behavioural remedies is empirical, and pragmatic. Our starting point is to consider what we have learned across many proceedings in implementing complex and behavioural remedies. That experience shows that the complexity of some markets and transactions renders behavioural remedies less suitable in a number of ways. Behavioural remedies create continuing economic links and are unlikely to create the same level of pre-merger competitive intensity between the merging firms. Behavioural remedies can become quickly outdated or unsuited to remedying issues as markets, products and customer desires change. Structural remedies are more likely to avoid these pitfalls and preserve competition. In our view, where no divestment is available, agencies should not be afraid of prohibiting a merger. Preserving competition is in the best interests of consumers. For example, in Meta/GIPHY, we rejected behavioural remedies proposed by Meta that essentially involved a time-limited commitment to provide continued access to GIPHY. In light of the dynamic and fast-changing nature of the relevant markets, a static behavioural remedy would not have been effective in addressing the competition concerns. A number of problematic deals have been abandoned in the last few years where we did not accept behavioural remedies, for example, Illumina/PacBio, Crowdcube/Seedrs, TopCashback/Quidco. So, advisors should consider the inherent challenges of behavioural remedies before proposing remedies of this nature to resolve competition concerns.
Best practice for engaging with the CMA
7 . What can parties do to make it more likely that their engagement with the agency will be constructive? What are some of the thing's parties should avoid when engaging with the CMA?
I'd like to highlight a couple of points for parties engaging in global mergers, for digital companies and some general practical tips about the CMA's processes.
In terms of global mergers, we encourage merging parties to discuss the process and timing of multijurisdictional mergers as early as possible to allow for the alignment of timetables across agencies. Sometimes parties hold back waivers, or choose not to align engagement across agencies, procedurally and substantively. Inevitably this merely slows down our review, rather than achieving any benefit for parties, and we think it is in the interests of parties and authorities for the waivers to be provided to enable authorities to share information. This coordination aids consistency of outcomes and the speed of assessment so factual differences can be understood, tested, and ironed out quickly.
Thinking about digital companies in particular, we anticipate that all those firms are considering how to engage with regulators and competition authorities, as part of their broader regulatory strategy. Where a transaction has a nexus to the UK, but you do not believe it raises competition questions, we would encourage you to submit a briefing note to our mergers intelligence committee (this is our team that identifies which mergers should be 'called in'). Proactively engaging with us and, where appropriate, notifying these mergers can help parties manage deal execution risk. For acquisitions by major digital platforms in particular, we would encourage companies to proactively notify mergers that raise jurisdictional and/or substantive questions that are likely to be subject to review in the UK.
In terms of the CMA's process itself, the CMA runs an administrative process rather than prosecutorial process. There is significant transparency in our process with opportunities for merger parties to see CMA provisional concerns/conclusions and respond in writing and orally.
Decision makers (at phase 1 and phase 2) are very keen to hear directly from the company executives. We have had instances of companies attending main party hearings without the key executives and with external lawyers essentially repeating the main talking points about the merger. What is lost when that happens is the chance for us to hear directly about the way that competition operates in each market, and the commercial perspective. So, it can take longer, and the parties have to work harder to make their points, which is in no one's interest. It's important to note in this context that where a merger is referred for a phase 2 investigation there is independent decision-making performed by members of a CMA Panel. Panel members are non-political appointees, selected by way of an open competition based on their experience, ability, and diversity of skills in competition economics, law, finance, and business. The CMA Panel acts as fresh decision-makers between the 2 phases of mergers cases.
Comparative timelines
8 . Can you comment on the reasons for the apparent increased length and scope of merger investigations?
The first key point to highlight is that the UK has a voluntary merger notification regime and typically subjects only about 50 to 60 cases per year to any kind of formal investigation - a number that's far lower than jurisdictions with mandatory merger control regimes. For example, the FTC reported 3,644 HSR filings in 2021. The CMA issued 45 phase 1 merger decisions in its 2021 to 2022 financial year. Of the CMA's phase 1 decisions in 2021 to 2022, 8 were referred to phase 2, 5 were resolved by way of remedies at phase 1, and 1 was abandoned.
The second key point to highlight is that merger control is, by its nature, complex, particularly where a deal merits an in-depth Phase 2 investigation. And complex and fact-intensive investigations tend, by their nature, to take some time.
The third key point to highlight is that merging parties share the responsibility for how quickly a case proceeds.
For our 2020 to 2021 fiscal year, the average length of significant investigations by the CMA (Phase 2 or Phase 1 with remedies) was 11.9 months compared with 2020 averages for the US of 11.4 months and the EU of 14.9 months. Although the overall duration of merger investigations in the UK is similar to other jurisdictions, there are parts of our timeline that are less flexible, in particular than the timelines of the US agencies. We encourage merging parties to discuss the process and timing of multijurisdictional mergers as early as possible to allow for the alignment of timetables across agencies where feasible and more generally, to engage with us constructively at an early stage to optimise the timeframe for review.
Footnotes
[footnote 1] Sabre v CMA [2021] CAT 11, paragraph 144.
[footnote 2] Enterprise Act 2002, section 23(5).
[footnote 3] Enterprise Act 2002, section 128.
[footnote 4] For example, economic literature has pointed to a reverse "U-shaped" relationship between competition and innovation efforts (Aghion, Bloom, Blundell and Griffith (2005), "Competition and Innovation: An Inverted-U relationship, Quarterly Journal of Economics, 120(2), 701-28). As many problematic mergers reviewed by the authorities involve markets where competition is limited (or might be limited post-merger), an increase in competition would lead to an increase in innovation efforts.
[footnote 5] MAGs, paragraph 4.3.
With thanks to Neha Dhaun-Brewer, Chris Prevett and other CMA colleagues who assisted in preparing these remarks.
