Is breaking up that hard to do?: taking stock on parallel EU/UK merger reviews since Brexit "freedom"
Introduction: small target, meet big ticket
Not everything – for richer or poorer – can be attributed to Brexit. Landmark CMA new-era cases such as Facebook / Giphy, Sabre / Farelogix, Illumina / PacBio are all examples of the CMA taking an active prosecutorial interest in “small-target” deals (small globally and certainly in the UK) it might not have looked at in the past, and using the all-you-can-eat buffet that is the share of supply test to do so.
This radical shift in case prioritisation (and in some cases jurisdictional ingenuity) has nothing to do with Brexit and these cases were not reviewed by the European Commission (“EC”) – note they predate the Illumina / Grail interpretative turn on Article 22. However, there must be more to life than small cases. The end of January marked two years since the UK proverbially “took back control”. The end of the application of the EU’s One Stop Shop in the UK has given the CMA power to get its hands on “big-target”, big-ticket global deals that trigger EU filings – previously, the only mergers that could be guaranteed off-limits for the CMA.
As we wrote in a Platypus post last year, the UK flexing its big-ticket muscles gives rise to the possibility of UK and EU outcomes diverging not only notionally, when looking at different national or sub-national markets (for which divergence is of trivial interest) but also truly when looking at the same market, say, one that is at Europe-wide (known in the trade as EEA-wide and now EEA-UK wide, post-Brexit) or global. On this anniversary of two years of “freedom”, Platypus thought it appropriate to do a methodical review of parallel reviews to classify substantive divergence in practice (for more on procedural considerations, see our recent post here).
Consulting our Platypus case register reveals that since 1 February 2021, there have been 20 “full-fat Phase 1” or beyond completed parallel reviews by the EC and the CMA. This is significantly lower than the 60-100+ predicted when the CMA’s budget was boosted. However, it is important to remember that the true number of cases considered by both the EC and CMA is significantly higher than those listed here, given that cases considered on an informal briefing paper by the CMA’s Mergers Intelligence Committee are not public. The CMA has effectively “blessed” many more mergers than appear on this list by choosing not to call them in. This is a welcome development that saves private resources and, more importantly, allows public resources to be focused on cases more likely to warrant possible intervention and remedial action.
In this post, we analyse the statistics and offer some reflections on what the numbers can tell us about the risk of divergence in practice. We find that true divergence – where the CMA and EC have considered the same markets and reached different conclusions – has so far been infrequent. But even for cases both authorities cleared unconditionally at Phase 1, there are some interesting lessons for deal-doers.
In this article we take a closer look in turn at: (i) cases with the same outcome based on similar competition analysis; (ii) cases with the same ultimate outcome, but where there were material differences in the scope of either concerns or remedies offered; and (iii) cases with different outcomes. We examine the reasons for differences – in many cases explained by different facts in the different geographies, but in others driven by divergent substantive or policy approaches. Finally, we offer some reflections on what the first two years can tell us about what to expect in parallel reviews going forward.
Cases we can agree on
In half of the cases subject to parallel review, the EC and CMA were dancing in time and unconditionally cleared these at Phase 1 by both authorities. The outcome is unremarkable, but for merger control junkies, Platypus can still point the flipper at a number of interesting procedural observations about the “aligned” cases.
First, half received their unconditional clearance decisions from the EC and CMA within a month of each other and two further cases within two months. In most but not all cases, the EC clearance came first. Given the CMA’s longer statutory timetable for Phase 1 and the discretionary and idiosyncratic vagaries of the prenotification process on both sides of the Channel, Platypus would say that these parallel processes look – at least from a timetabling perspective – fairly smooth.
Three cases had more significant timing gaps and, in two cases, the CMA clearance came after the EC clearance. In Macquarie / National Grid and Brookfield / Scotia Gas the CMA clearances came more than six months and three months respectively after the EC clearances. In both cases however, this is clearly explicable on the basis the assets in question were in the UK and the EC’s review was on a simplified procedure. More interesting is Microsoft / Nuance in which the CMA’s clearance came two and a half months after the EC. Platypus is not privy to the reason: one hypothesis is that the CMA held the parties in pre-notification for longer in order to do more forensics (internal documents and otherwise) to get comfortable with the facts and consider if the deal did not raise competition concerns.
Another noteworthy case of relative alignment – albeit not on unconditional clearance – is that of IAG / Air Europa in which a CMA investigation was launched late in the Commission’s Phase 2 process. Platypus’ hypothesis here is that the CMA’s intervention was intended to ensure the CMA had a seat at the table with respect to any remedy package offered by the Parties, but the deal, that had originally signed in November 2019, was abandoned by IAG, potentially due to a mix of COVID and the scope of the EC’s concerns. The EU filing was withdrawn, and the CMA issued a “found not to qualify” decision after its late intervention. However, that deal is now back on the table.
Remedies: I never wanted that dog, but the clear-cut sound of that vinyl collection is all mine
There have been three amicable splits where the CMA and EC both cleared the deals at Phase 1 with remedies. In Bouygues / Equans, the CMA’s review found problems in a single UK-specific market relating to HS2, and accepted remedies on this basis, and the EC did the same for its own review – so the same outcome, but with different considerations and remedy packages. In Parker-Hannifin Corporation / Meggitt, a single global remedy package was accepted by both the EC and the CMA (though the CMA UILs were accepted by the Secretary of State as there was a public interest intervention notice in the UK in this case).
Perhaps the most interesting parallel Phase 1 remedies case is S&P Global / IHS Markit. In this merger of financial data and services providers, the merging parties conceded a number of competition concerns (SLCs in UK parlance). Each of the CMA, EC and US Department of Justice found identified concerns, but their number and scope differed for each authority despite the markets being global such that one might expect commonality of diagnoses. Ultimately, all four of the commodity price assessment SLCs identified by the CMA were addressed by the sale of IHSM’s overlapping businesses to single purchaser as an upfront buyer (negotiated in parallel with other authorities).
In Sika / MBCC, both authorities found problems and the case was ultimately cleared with remedies by both the CMA and the EC but the processes diverged. The CMA process ran first, and the CMA rejected the merging parties’ remedy offer at Phase 1. In one of the first fast-track Phase 2 processes, the CMA ultimately accepted an enhanced remedy package in Phase 2. The same remedy package was accepted by the EC in Phase 1, with the EC clock not starting until well into the CMA’s Phase 2 process.
We agree to disagree: divergent outcomes
That leaves four cases in which the EC and CMA (consciously) uncoupled and reached different substantive conclusions: ALD / LeasePlan, Veolia / Suez, Facebook / Kustomer and Cargotec / Konecranes.
In ALD / LeasePlan and Veolia / Suez, the divergence is at most notional and easily explained by the different competitive dynamics in the sub-national markets in question. For ALD / LeasePlan, the CMA found the merged entity would continue to be constrained by a large number of competitors in the UK leasing services market, whereas the EC found that there would be limited competitive constraints facing the merged entity in markets for operational leasing in Czechia, Finland, Luxembourg, Ireland, Norway and Portugal, so divestment remedies were offered and accepted. Similarly, Veolia / Suez was cleared by the EC with Phase 1 remedies but was effectively unwound in the UK after the parties appeared not to be willing or able to offer satisfactory remedies to address the CMA’s concerns regarding the overlap in the parties’ competing operations in the UK.
The same notional divergence point will be able to be said for another ongoing case – Korean Air / Asiana – which the CMA has recently cleared in Phase 1 with UILs, but where there is an ongoing Phase 2 EC process. In this airline merger, the CMA and EC were considering different routes (aka geographic markets) and in the case of the CMA, only a single city-pair (London-Seoul).
Most interesting are the cases where the CMA and EC considered the same supra-national markets, but reached different conclusions. To date, this has only happened in two cases where the issue of remedies was potentially determinative.
First, in Facebook / Kustomer, the CMA cleared the transaction unconditionally at Phase 1 despite, among others, two input foreclosure theories of harm. The EC had referred the transaction to Phase 2 a month earlier, and ultimately, the EC conditionally approved the deal after accepting a behavioural remedy for similar input foreclosure concerns. Given the CMA’s policy stance on behavioural remedies including in Facebook / Giphy, this deal would have presumptively been viewed by the CMA as effectively binary in outcome, as detailed in our earlier Platypus publication.
Second, in Cargotec / Konecranes, the CMA effectively blocked the deal at Phase 2, requiring full divestiture as the CMA was not satisfied with the remedy package offered by the parties, and accepted by the EC. Since our previous post on this topic, the CMA has issued its Final Report and a number of CMA officials have spoken about the case. The CMA’s view is clear that notwithstanding the markets in question being supra-national, its duty was to consider specifically its impact in the UK. The CMA’s Final Report suggests that third-party feedback from customers and potential purchasers regarding the remedy package given to the CMA was different to that given to the EC.
Two years and 20 cases (and counting) on, truly divergent outcomes of EC / CMA review have been – for merging parties – mercifully rare. To date, where outcomes have diverged, such divergence has generally been driven by different competitive conditions in the markets under consideration. However, the unpredictable risk of different feedback being given to different authorities about the same product / geographic markets is clearly worrying for merging parties. This is particularly acute because although the CMA and EC do so far cooperate (to an extent) on parallel reviews, they are not able to share their evidence base with one another without waivers from the relevant third parties. Such waivers are generally readily given by merging parties, but are more complex to obtain from third parties, whose evidence is frequently outcome-critical. This makes it critical for parties considering a particularly difficult merger to understand how customers in different geographies could have differing views on the deal, even if markets are global in scope.
The second systemic factor shaping up to prompt divergence is the EC and CMA’s markedly different remedy policies. The CMA has had a long-held stance against behavioural remedies, exemplified by its former CEO Andrea Coscelli’s vocal criticism of remedies accepted by the EC e.g. in Google / Fitbit. Sarah Cardell, the CMA’s current CEO, recently gave a speech in which she reiterated the CMA’s scepticism about behavioural remedies and – while not ruling them out entirely – made clear they would not be acceptable in most cases. The EC on the other hand has shown they consider themselves legally obliged to accept behavioural remedies, even where they fall short of perfection, where they are deemed to proportionately address the identified harm. It is clear that the CMA’s remedy policy drives binary outcomes, especially in tech markets. The ongoing Microsoft / Activision case is shaping up as a key test case on this point.
Finally, how exactly the CMA will choose to prioritise cases is hard to predict with precision. The CMA’s guidance suggests the CMA may in some cases choose not to open a parallel investigation where other authorities are already investigating a merger and any concerns / remedies would be common to the UK. There are some notable omissions from the CMA’s case docket: to give just two examples, Illumina / Grail and Meta / Within, which are cases in which the EC and FTC respectively have found serious competition concerns in putatively global markets. But this doesn’t tell the whole story: the CMA has opened investigations in other cases with similarly global profiles.
Part of the explanation could be that, while the CMA’s jurisdiction is extremely broad, it is not – legally speaking – unlimited. The CMA must still jump through the hoops of showing that either the turnover or share of supply test are met in the present, not in future, albeit that following Sabre / Farelogix it is clear that the UK courts will accept some fairly innovative calculation methodologies. By contrast, if the ECJ upholds the General Court's decision in Illumina / Grail, jurisdiction for the EC will be purely a matter of NCA / EU discretion. But case selection may also be down to policy factors such as case prioritisation and comity, rather than bottoming out a jurisdictional no-can-do or “FNTQ”.
How each of the EC and CMA use their – for all practical purposes – unlimited discretion in both parallel and separate reviews is a key item to watch in coming years, especially as the DMA and its proposed UK equivalent will oblige tech “gatekeepers” / “strategic market status” firms to tell the EC and / or CMA about every acquisition they make.
Like any separating couple, the relationship between the CMA and the EC may have its ups and downs but ultimately, agree or disagree, neither party can ignore the other. For these two leading authorities, Platypus hopes that politics and other factors allow for a productive and amicable relationship of respect.
Tables for Reference
|EU/UK||P1 clearance||P1 UILs||P2 clearance||P2 remedies||Prohibition / full divestiture||Withdrawn / FNTQ|
|P1 clearance||All other cases (10)|
|P1 remedies||ALD / LeasePlan||Bouygues / Equans; S&P Global / IHS Markit; Parker-Hannifin Corporation / Meggitt||Sika / MBCC||Veolia / Suez|
|P2 remedies||Facebook / Kustomer||Cargotec / Konecranes|
|Withdrawn||IAG / AirEuropa; NVIDIA / ARM|
*Grey indicates deal did not proceed
2 Unconditional clearances – dates
|No.||Case||UK Date of Clearance||EU Date of Clearance|
|1||Macquarie and BCI / National Grid gas transmission and metering||22 November 2022||3 June 2022|
|2||Deutsche Post DHL Group / J.F. Hillebrand Group AG||15 March 2022||25 January 2022|
|3||Microsoft Corporation / Nuance Communications, Inc.||2 March 2022||21 December 2021|
|4||Brookfield Asset Management Inc. / Scotia Gas Networks Limited||1 March 2022||22 November 2021|
|5||Thermo Fisher Scientific Inc / PPD Inc||3 December 2021||7 December 2021|
|6||Graphic Packaging Holding Company / AR Packaging Group AB||6 October 2021||12 October 2021|
|7||IHS Markit Ltd / CME Global Inc. JV||27 July 2021||20 July 2021|
|8||AstraZeneca / Alexion Pharmaceuticals||14 July 2021||5 July 2021|
|9||Advanced Micro Devices, Inc / Xilinx, Inc||29 June 2021||30 June 2021|
|10||SK hynix / Intel's NAND and SSD business||28 June 2021||20 May 2021|