What’s the difference between… a competition agency and an angel investor?

1 I sit and wait… does an angel contemplate my fate… 

Merger control is – at its core – prophecy. It calls on the authority to predict the future, or in fact, to predict and compare two different futures: one with, and one without, the merger under review. Much like early stage ‘angel’ investors, the CMA is in the business of assessing which nascent businesses or products will succeed, or at least be sufficiently successful to shake up the market.

Though predicting the future is inherently difficult, traditionally, a shortcut used in the majority of cases was to assume that the future without the merger – i.e. the counterfactual – is the status quo ex ante. A high standard of proof has always been applied (and continues to apply) to merging parties wanting to argue an alternative counterfactual in their defence (e.g. that the target is a failing firm, or that entry/expansion is likely and timely). 

The CMA historically applied the same high threshold to itself in defining counterfactuals in which the merging parties would be competing more aggressively with each other than in the status quo. Fair was fair – or was it? 

A quest for redemption after perceived under-enforcement has led the CMA to significantly expand the framework within which mergers in innovative markets are considered. 

Whilst once it might have been sufficient to show that merging parties are not close competitors now, and will not be close competitors in the future, Meta / GIPHY and most recently the provisional findings in Adobe / Figma (ultimately leading to termination of the deal) demonstrate this is not necessarily enough. Rather, the CMA can and in some cases will find a substantial lessening of competition (“SLC”) if one party’s efforts are imposing dynamic competitive pressure on the other, even if those efforts are unlikely to succeed. 

Assessing dynamic competition is inherently more difficult and speculative than in cases where the structure and modalities of competition are relatively static. Indeed, entire venture capital ecosystems exist attempts to do just this – to pick likely winners. 

In this post, we look at how the CMA has tackled this question in its recent decisional practice. We find an emerging but incomplete picture. The result is that it can be difficult for (potentially) merging parties to predict which deals are likely to get wings, and means that in practice, feasibility assessments in markets characterised by dynamic competition need to factor in this additional risk. 

2 And do they know, the places where [an innovative business] goes…

2.1 Dynamic competition: You’ll know it when you see it!

Prior versions of the Merger Assessment Guidelines (MAGs) were silent on dynamic competition, but the 2021 MAGs explain the CMA will assess whether a transaction is likely to disincentivise a competitor from entering or expanding in a market or result in an existing operator reducing “its efforts in the present to protect against the possible impact of a dynamic competitor”. For ease, we refer to this formulation of dynamic competition as a ‘dynamic SLC’ in the rest of this article.

Whilst the 2021 MAGs introduce the concept of a dynamic SLC, they provide little detail about how the CMA will assess this concept in practice. In the Meta judgment, the CAT says that while “one knows [dynamic competition] when one sees it”, it is difficult to define and apply in the abstract. The CAT further explains that where dynamic competition is present, this “involves a far greater consideration of innovation and invention… rather than analysis of an existing market of an assessment of future that lie within it”.

In essence, dynamic competition involves a more speculative / forward-looking assessment of how the market may evolve from its current state, based on the parties’ incentives to innovate. The most fundamental difference with traditional innovation theories is that a dynamic SLC requires no evidence that the innovation would succeed. It is sufficient that the process of dynamic competition is a constraint on the merging parties. This opens up the possibility that mergers that do not involve either material existing horizontal overlaps, or significant vertical links (nor any evidence that there is an imminent intent to create them), could be deemed problematic because they would lead to the loss of a dynamic competitive force.

2.2 Likely SLCs based on unlikely outcomes

No Platypus reader needs to be told that there is a clear consensus among authorities the world over that problematic deals were missed over the last two decades – especially in innovation-heavy markets. Institutional remorse over perceived historical under-enforcement has been particularly acute at the CMA, which unlike many other authorities, had jurisdiction to review (and did in fact review some of) these deals. The CMA has historically exercised caution in deals dealing with speculative theories of harm (eg. Paypal / iZettle and Amazon / Deliveroo, both cleared unconditionally). But recent years have seen a change in the approach to uncertainty and balancing of the risks of under- and over-enforcement. The CMA’s CEO has said that the CMA has “taken steps to ensure that the existence of some uncertainty about the future evolution of a market doesn’t itself create a presumption in favour of clearance”. 

In Meta / Giphy (2021), the CMA primarily focused on whether Giphy’s nascent paid alignment advertising service (adverts under GIFs) would become a close competitor of Facebook’s advertising business. The parties argued that there was little documentary evidence or data which indicated that Giphy would become a material competitive constraint on Facebook. Similar to the approach outlined in the MAGs, the CMA noted in its Final Report that “the elimination of a dynamic competitor that is making efforts towards entry or expansion may lead to an SLC even where entry by that entrant is unlikely and may ultimately be unsuccessful” (emphasis added).

The CMA’s dynamic SLC was subsequently considered and upheld by the CAT, which applied its own methodology for considering whether the transaction would result in a dynamic SLC. In particular, the CAT was critical of aspects of the CMA’s guidance to date, which it said do “very little to create certainty and remove dynamic competition and impairments to it from the realm of “it is what I say it is”. This is part of the reason why the CAT felt the need to articulate its own assessment.

However, while it is clear the CMA has both wings and a licence to fly when considering dynamic competition issues, the MAGs do not provide any examples of whether particular markets or competitive conditions make a dynamic SLC more likely. Third parties (including the CAT), have also noted that the MAGs provide no holistic framework for how the CMA will assess the likelihood of a dynamic SLC. Indeed, the MAGs note that there is “no special elevated evidential burden” with respect to theories of harm that “involve changes in future competitive conditions” (e.g. a dynamic SLC). Similarly, the MAGs state that “the presence of uncertainty… does not preclude the CMA from finding competition concerns” and the “absence of certain types of evidence… will not in itself preclude the CMA from concluding that the SLC test is met”.

This has led some critics (and indeed the opposing advocates in Meta / Giphy) to note that the MAGs provide the CMA with carte blanche to find a dynamic SLC decision whenever it chooses, handing some deals a one way ticket to Peter and Paul at the pearly gates. 

The PFs in Adobe / Figma (2023), prompting the merging parties to abandon the deal, are the most up to date reflection of the CMA’s practice on dynamic competition. The CMA provisionally found that Figma could become a close competitor to Adobe in the future in spite of limited third party support for the position that Adobe and Figma currently compete in image editing software, and Figma’s position that it is neither present in the image editing market, nor could it offer image editing software due to significant technological barriers to entry and a lack of commercial rationale to pivot its business. The decision leans heavily on the CMA’s review of the parties’ internal documents, in which the CMA has found evidence of:

• An internal review by Adobe of the threat from Figma in relation to its core image editing software (including Photoshop);
• Concerns by Adobe management over the threat from Figma in relation to professional users, including up to August 2022, a few weeks before the merger was announced in mid-September;
• Active steps taken by Adobe to mitigate the threat from Figma to image editing software;
• Stakeholder communications and strategy documents which indicate that Figma intended to develop image editing software;
• Employees at Figma with experience in image editing software who “could form a core around which further engineering resources could be allocated in future”.

If the key message from Meta / Giphy is that the CMA has wide discretion on when they find a dynamic SLC, the lesson from Adobe / Figma is that it has wide discretion on how it supports that SLC and internal documents are key - ‘you’ll know it when you see it’.

2.3 Entry or expansion by competitors: heads I win, tails you lose?

So it’s clear the CMA can intervene on the basis of dynamic considerations, provided they lead to a “substantial” lessening of competition. But flip the coin, what about third parties’ entry/expansion?

Dynamic considerations do not always point to an SLC. Indeed, when dynamic competition initially entered the merger control discourse, it was viewed as “a possible countervailing factor to a static SLC”. The MAGs set a higher evidential threshold when considering dynamic competition as a countervailing factor to an SLC. In particular, the MAGs note that entry and / or expansion will rarely be sufficient to prevent an SLC and that any evidence in support of that conclusion will need to be “robust”. In practice, this means that merger parties need to present a large body of documentary evidence and data to successfully argue that dynamic competition will mitigate an SLC.

Nonetheless, it is undeniable that over the last ten years, the CMA has cleared more mergers based on dynamic considerations than it has blocked. A poster-child for this is Just Eat / Hungry House, in which the CMA recognised that although the parties were at the time close competitors, they were responding to competitive pressure from new business models (the likes of Deliveroo and Uber Eats). But clearances on the basis of dynamic competitive forces are not a relic of the past: two of the CMA’s recent Phase 2 unconditional clearances (Viasat / Inmarsat and Norton Lifelock / Avast) could have been prohibitions had dynamic competitive forces not been given weight:

• Viasat / Inmarsat (2023): the CMA found that both parties were close competitors in the supply of broadband inflight connectivity (“IFC”) services to commercial and business aviation. However, the CMA was willing to clear the transaction as a consequence of the emerging presence of Starlink (a company owned by Elon Musk) which only started to offer IFC services to aircraft in the period between the CMA’s Phase I and Phase II decisions. Despite Starlink being a new operator in these markets, it had already won a number of tenders with airlines by the time of the CMA’s final report and featured prominently in the parties’ internal documents. The CMA considered that any reduction in competition brought about by the merger was offset by a combination of Starlink’s growing presence and the parties’ existing competitors. 

• Norton / Avast (2022): both parties were found to be close competitors in consumer cyber safety solutions for computer and mobile operating systems. Similarly, to Viasat / Inmarsat, the CMA was prepared to clear the transaction on the basis of the expanding presence of Microsoft, who had been developing its own security solution over a number of years. As with Starlink, Microsoft expanded its competing offering in the course of the CMA’s investigation and there was clear evidence that it would be a material competitive constraint on the parties going forward.

These decisions are rare examples of entry and expansion being a mitigating factor in a CMA clearance decision. However, it is notable that both Starlink and Microsoft featured in the parties internal documents pre-transaction and, perhaps crucially, progressed significantly in their entry / expansion plans during the course of the CMA’s investigations. This contrasts to the CMA’s approach in for instance Meta / Giphy and Adobe / Figma - where the CMA blocked (or in the case of the latter, put an end to the deal by negative PFs) deals on the basis of speculative entry or expansion by one of the merger parties. These decisions bring to life the existing asymmetry in the CMA’s approach when considering a dynamic SLC versus dynamic competition as a countervailing factor to an SLC. 

3 Whether I’m right or wrong… 

It can be no coincidence that the cases in which dynamic (or indeed ecosystem) considerations have featured most prominently involve “big tech” or “big pharma” (Roche / Spark, Illumina / PacBio) companies. Even in the two recent cases cleared on the basis of entry by third parties, that entry came from SpaceX (owned by Elon Musk) and Microsoft. To Platypus, this may hold the clue to a potential framework for analysis. 

While Parties would certainly be well advised to apply a “GAFAM” discount to any prospects analysis, Platypus believes that a closer analysis of the reasons why big tech / pharma have strong track-records of bringing innovations to market may assist in identifying the common themes from recent cases (and help applying a rigorous framework). Some of these are merger control bread and butter, but others are particular plus (or minus) factors when assessing dynamic theories. 

• Economies of scale / scope: these companies have scale advantages over smaller players, given the sprawling nature of Big Tech business models. These companies therefore benefit from economies of scale and scope, which can facilitate their ability to expand into new and adjacent markets, making disruption more likely.
• Identifiable and attractive target market: in all the cases to date, it was easy for the CMA to identify an attractive market that the disruptor was expanding into or capable of doing so. In the case of Meta / Giphy - the CMA particularly focused on Meta’s lucrative advertising business as an area where Giphy was capable of expanding whilst Adobe / Figma turns in part on whether Figma has a commercial rationale to pivot into image editing. 
• Access to capital: These same companies also had access to significant reserves of capital to fuel their expansion ambitions – they were their own angel investor. But the source of their funding and their financial structure can also be relevant to how easy or difficult it is in practice to fund significant investment.

As set out above, these market dynamics, supported by internal documents, have held the key for competition authorities when assessing potential impact on innovation. However, focus on these metrics alone may miss the more ephemeral but just as important "X factor”. In reality, many innovative businesses are not run in a way that a business textbook would predict. The defining feature of a visionary is that they will see opportunities and possibilities others do not. Such “visionaries” are often the people actually determining commercial strategy in innovative companies. But the reasoning for such “visionaries’” decision making is unlikely to be set out in a board pack. This is a huge challenge for authorities and an area where further face-to-face evidence or even something akin to a witness statement or US-style deposition might prove more insightful than internal documents prepared by their staff. 

Ultimately, like an angel investor considering whether to put money into an uncertain business proposition, there is always going to be a level of “gut feel” that influences decision making on whether a nascent business or product is considered sufficiently likely to succeed to be a true competitive threat. This is not an unequivocally bad thing – instincts and intuition exist for a reason and are often grounded in knowledge and experience. Being more explicit about what the knowledge and experience underlying such a “gut feel” is could – in Platypus’ view – be extremely instructive in predicting which deals should get wings.