If the cap fits: Pitching a liability cap

IF THERE IS ONE much-debated clause in an accountant’s engagement letter, it is very often the liability cap. But where does one aim to pitch the level of a limitation clause?

Exclusion clauses in accountancy firms’ engagement letters will, to the extent they seek to restrict liability for negligence, be subject to a reasonableness test under the Unfair Contract Terms Act 1977 (UCTA). If a cap falls on the wrong side of UCTA, it is struck out completely. Courts will, in any event, interpret a limitation clause strictly. Too swingeing a clause may be found not to cover the alleged breach at all.

For audit engagements, there is the separate regime under the Companies Act 2006. Clauses can be rewritten to what is “fair and reasonable” (see box below), although very few agreements limiting the auditor’s exposure appear to have been approved in practice. We have seen none tested in court.


Accountants’ engagement letters reach the courts relatively rarely, but the current theme appears to be that robust caps will be upheld, at least on transactions with sophisticated commercial parties. A number of the recent cases relate to contract disputes between commercial parties on bespoke terms, where UCTA does not apply, but they do provide a pointer as to how the courts read limitation provisions.

Contractual interpretation

One such case is the recent decision on AstraZeneca v Albemarle (21 June 2011), which involved claims by AstraZeneca for non-delivery of a pharmaceutical product and a counter-claim by Albemarle for breach of a right of first refusal.

AstraZeneca’s claims in that case were capped at the purchase price of the goods (less than 10% of the claimed losses), similar to the commonly seen limitations based on fees incurred. While the judge was not looking at the clause in UCTA terms, there was, he said, “nothing unusual or unreasonable” about such a cap.

Second, did the seriousness of the alleged breach affect the enforceability of the limitation clause? A 2009 case – Internet BC v Marhedge – had held that there was a “strong presumption” against exemption clauses being construed so as to cover a deliberate breach. The judge in the AstraZeneca case, however, was highly critical of the earlier decision. All limitation clauses should be interpreted strictly, he said, but there are no presumptions for a case of deliberate breach. In other words, if contracting parties wish to exclude deliberate breach, or wilful default, from the scope of a limitation clause, they need to say so clearly.

Third, the courts will interpret limitation clauses with an eye to the commercial purpose of the contract. The judge refused to allow AstraZeneca to rely on a “no loss of profits” limitation clause, apparently in its favour, since to do so would have deprived Albemarle of any meaningful remedy for its claims under the “right of first refusal” clause. That, he said, cannot have been the intention of the clause.

UCTA reasonableness

Aside from any issues on the construction of a limitation clause, a firm will need to show that any cap excluding liability for negligence is fair and reasonable under UCTA. The reasonableness test is judged at the time the engagement was entered into, not after the event. That entails looking at various factors, including the strength of the bargaining position of the parties, any inducement to agree to the liability cap, whether the client knew of the existence of the term and to what extent the accountant could cover the liability risk through insurance.

Recent cases have generally been slow to strike down limitation clauses under UCTA. In Marplace v Chaffe Street (2006), a £20m cap on a claim against solicitors was upheld as reasonable. The judge referred to the equal bargaining position of client and firm; the fact that the clause had been discussed and had not been imposed as “non-negotiable”; and the commercial logic of the cap in light of insurance cover and the nature of the transaction.

Regus v Epcot (2008) was a contractual dispute about air-conditioning, which turned on the enforceability of limitations on liability. The Court of Appeal found that a clause in Regus’ standard terms set at 125% of fees or £50,000, (whichever is higher) was reasonable. Epcot’s counterclaim was for around £626m.

In Dennard v PricewaterhouseCoopers (2010), Mr Justice Vos looked at a liability cap of £1m, which – while the point was irrelevant because of his earlier findings – he ruled was reasonable. He found that the individual claimants in that case were experienced business people with considerable commercial influence. They knew and understood that accountancy firms customarily limited their liability and had a choice of advisers, but they had chosen not to negotiate the limitation clause. They could not, he said, be regarded as “innocents abroad”.

Most recently, in Camerata v Credit Suisse (9 March 2011), a claim against the bank for, among other things, negligent investment advice was successfully limited to “gross negligence, fraud or wilful default”. The client was sophisticated and well-advised; the investments were risky; and it was reasonable for Credit Suisse to wish to protect itself by excluding liability for “mere negligence”.

A robust approach?

As ever, one needs to be careful in taking too much from any one case. As the courts invariably point out, each contractual dispute will turn on the terms in question.

To repeat the lawyer’s (hindsight-driven) mantra, it is obviously critical that limitation clauses are clearly drafted. To satisfy the UCTA test, one needs to show that the cap was fair, given the particular circumstances at the time of the engagement letter. Ideally, one would be able to back up the commercial logic behind the cap by reference to contemporaneous documentation showing how this was communicated to the client. If not, one will at least be on safer ground relying on a limitation based on the particular engagement, say by reference to the fees or value of the transaction, rather than a “catch-all” cap applied universally across all service lines.

If there is one factor which appears to weigh more heavily, however, it is the bargaining strength and sophistication of the client. The court’s approach will be much stricter on an engagement with a less experienced client, but where terms are agreed with an entity well-versed in dealing with the profession and with limitations of liability (often since the client relies liberally on such clauses itself), it seems there will be a reluctance to interfere.

Limitation clauses are, after all, a now-familiar aspect of professional service engagement terms and, much as a balance needs to be struck, it appears that the instinctive default will be to hold commercial parties to their deal. All of which perhaps points to firms pitching caps at a more aggressive level where possible. How easy it is for the professional to negotiate those limitation terms in the first place is, of course, another matter entirely.

Andrew Howell is a partner in the accountants liability team at Taylor Wessing LLP

Related reading

aidan-brennan kpmg
The Practitioner