There was a (slightly under the radar, but very useful for practitioners and clients) judgment handed down by the Competition Appeal Tribunal (CAT) just before Christmas in Keltbray v CMA [2024] CAT 79, in an appeal of a penalty imposed by the Competition and Markets Authority (CMA) in relation to a Chapter I, Competition Act 1998 infringement decision. The judgment provides useful instruction on the CMA’s current thinking and approach to penalty calculation and settlement (particularly to ‘by object’ infringements).

This was never going to be an easy one to win. That pivotal word ‘discretion’ features heavily in what regulators can lawfully consider in relation to penalty, where the appeal court or Tribunal is reluctant to interfere. Spoiler alert: Keltbray lost on grounds 1 and 2 which were essentially both market definition grounds, argued differently and in the alternative. The CMA criticised Keltbray for bringing these as a ‘collateral’ challenge against the ‘by object’ infringement decision itself, which the CAT said it was ‘sympathetic’ to. Never a good start. With the knowledge of a fair chance of losing on the first two grounds, securing a net-net win based on the third ground, namely the  ‘penalty calculation’ alone was always going to be tough, knowing the CMA’s starting position would almost certainly be to revoke its 20% settlement discount. Adding in costs meant Keltbray would need to prove the CMA’s penalty was miscalculated by more than 20% just to come out even, and well above that to make the appeal worth the risk and effort.

The CMA’s approach and judgment all sound like a very strict application of the CMA’s Penalty Guidance (Keltbray claimed it was ‘mechanistic’), but is that good or bad? For legal certainty, this may be a good thing. But is that a bad for thing for discouraging appeals of penalty decisions, holding the CMA to account? That may be true, but only to the extent, perhaps of Keltbray’s grounds. The CAT also provided some useful indications on where Keltbray simply didn’t do enough to prove its case, or went about it in the wrong manner, which may help the parties (and the CMA) during the investigation phase negotiating settlement in future.

There is plenty more to dissect in the detail of this judgment, but below is a brief summary of some key points.

Keltbray was one of ten demolition companies that was found by the CMA to have colluded on tenders for contracts. Essentially, the companies decided between themselves who would put in the low, winning bid and that the others would submit higher bids, so called “cover bidding”, which is considered by the CMA to be a serious ‘by object’ infringement i.e., so no need to show effects on competition/harm.

Keltbray admitted liability, presumably in good part on the basis that the CMA would entertain a settlement agreement. A settlement was reached and the £20m penalty was thereby discounted by 20% to £16m. Keltbray appealed and while it framed its appeal to the CAT as limited to the CMA’s decision on penalty calculation, grounds 1 and 2 tackled this on scope of market definition and ‘effects’. The CMA thereby argued those grounds included a collateral attack on substantive elements of liability.

Keltbray claimed the CMA was wrong to calculate Keltbray’s relevant turnover (for the purpose of penalty) by reference to revenue across a broadly defined market (£820m) instead of just the eight contract tenders under consideration (worth £58m), which would have resulted in a much lower penalty calculation, around £12m.

The CAT rejected Keltbray’s arguments on the basis that it was reasonable to include the wider market revenue given:

• the type of infringement is considered to be ‘by object,’ the CMA is required to have regard to potential wider harmful effects of the infringement (and no evidence was provided to the contrary);
• the approach that Keltbray submits should be undertaken, to carry out an investigation of the actual effects for the purpose of penalty is inconsistent with the exercise the CMA is required to carry out where there is a ‘by object’ infringement i.e., it does not have to investigate or show actual effects; and
• it is inconsistent with the CMA’s guidance on the calculation of penalties in such cases.

Keltbray argued that the CMA’s decision was wrong because while the CMA recognised there were separate sub-markets for demolition services i.e., highly complex demolition services (HCDS) and general demolition services (GDS), the CMA failed to take these into account in its calculation of penalty by including both in one market. Economic evidence was submitted from both parties.

The CAT said the CMA’s purpose in the circumstances was to define the market by taking a broad view of the trade potentially affected by the conduct. The CAT rejected Keltbray’s arguments essentially on the basis that, on the evidence (which it noted was ‘crafted for the appeal’), no stark division was established between the resources required for HCDS or GDS projects, and that Keltbray itself takes a joined-up approach to the use of its resources when it is addressing a HCDS or GDS. There was no evidence that market participants could not simply shift resources from one type of service to the other.

Keltbray argued, in essence, that the penalty was disproportionate, and the CMA therefore failed to have regard to its statutory objectives. Without going into the detail of the penalty calculation (of which there is a lot that is helpful guidance), the CAT agreed to an extent and decided that the £20m penalty should be reduced to £18m.

However, by virtue of appealing against the penalty, Keltbray was departing from the settlement agreement not to appeal, giving the CMA the option to revoke the settlement discount, which it did i.e., the 20% off £20m was added back. Keltbray nevertheless claimed it would be unfair for the CMA to revoke the discount given that it had ‘won’ on a reduction of penalty and the CMA should therefore exercise discretion not to revoke the discount. The Tribunal disagreed on the basis that by bringing the appeal on penalty in the first place, Keltbray had essentially undermined the policy rationale for the CMA agreeing to settlement discounts – to save further resources for finalising the investigation – and that it was sympathetic to the CMA’s submission that the appeal amounted to a collateral attack on the infringement decision itself, which caused the CMA to expend significant resources to defend the appeal.

Overall negotiating settlements with the CMA, early, thorough engagement and robustly supported economic and legal arguments are more critical than ever. The lessons of Keltbray—especially around market definition disputes, penalty appeals, and the revocation of settlement discounts—underscore the careful cost-benefit analysis that should be performed before challenging a penalty in the CAT.

Keltbray demonstrates the CMA will stick as close as possible to its Penalty Guidance unless it has good reason to do otherwise. By reference to the Guidance, the CMA will do what it can to come up with an accurate and proportionate, but not rigid, penalty assessment based on the particular evidence it has and, by extension, it is up to the party settling to help the CMA to help them. If there is any chance of convincing the CMA to take an approach different from its initial view (e.g., the evidence supporting the draft Statement of Objections), it must be given good, defensible, reasons, preferably supported by solid and robust evidence.

Challenges to administrative decisions of an Authority permitted to apply significant discretion to an assessment are notoriously difficult to win to begin with. Appeals of penalties that could be argued by the CMA to amount to a collateral challenge to the substantive decision makes it even tougher. Does that dissuade such challenges? To a degree yes, but that’s the idea – you have to give something to get something.

And perhaps, one more important reminder – in settlement negotiations, a deal is a deal.