Exclusion clauses and limitation of liability clauses (aka damages caps) are important features of many IT contracts. Drafted correctly, they allow parties at the outset of a venture to balance risk against potential benefits, to procure appropriate insurance cover, to control and predict financial exposure and, ultimately, to manage their businesses in a commercially sensible way. Drafted incorrectly, they will not achieve these ends – indeed they may achieve nothing at all.
In the context of contracts between commercial parties in the IT field (the subject of this article), drafting an exclusion clause or liability cap ‘correctly’ essentially amounts to three things, namely ensuring that the clause:
· is incorporated into the agreement;
· is enforceable;
· will successfully limit or exclude losses (or any other remedies which might otherwise be claimed) as intended.
Whilst purely commercial aspects of liability caps (most notably the level of the cap) are frequently the subject of intense negotiation between parties to IT contracts, the operative wording of many exclusion clauses and liability caps tends to consist of ‘tried and tested’ boilerplate that is known (or at least hoped) to meet the three requirements of incorporation, enforceability and successful limitation or exclusion. Standardised clauses contain phrases and mannerisms that have evolved over time, having been moulded and re-shaped in the light of judicial scrutiny, and display a number of recognisable characteristics (some sensible, some not). The purpose of this article is to review the basics of these clauses and to remind ourselves of why some of the common drafting oddities are there. It does so by reference to the exclusion clauses, liability cap and, for the sake of variety, the entire agreement clause in the standard written terms of business of the (fictitious) purveyor of cloud computing infrastructure and services, Clouds Unlimited.
The Clauses
The clauses under scrutiny are as follows:
10. Clouds Unlimited does not exclude or restrict liability for fraud, or for death or personal injury caused by its negligence.
11. CLOUDS UNLIMITED LIMITED SHALL NOT BE LIABLE TO CUSTOMER FOR LOSS OF PROFIT, REVENUE, CONTRACTS, ANTICIPATED SAVINGS OR ANY OTHER INDIRECT OR CONSEQUENTIAL LOSS.
12. In any event, Clouds Unlimited shall not be liable to Customer for more than £1m.
13. These terms and conditions constitute the entire agreement between the parties. Neither party has relied upon any statement or representation made by the other in agreeing to enter this contract.
The draftsman has included many of the elements generally found in the ‘exclusions and limitations’ sections of standard terms of business; he has also invoked a few of the wording oddities that are peculiar to exclusion clauses, and has fallen into at least one of the traps.
We look at each clause in turn below.
Clause 10: The Carve-out
It is important to limit the ambitions of an exclusion clause, as failure to do so can render the clause unenforceable. In particular:
· Any clause which, on a proper construction, excludes or restricts liability for negligence resulting in death or personal injury is unenforceable by virtue of s 2(1) of the Unfair Contract Terms Act 1977.
· Any exclusion or restriction of contractual liability in a set of written standard terms and conditions must be reasonable (UCTA, s 3). If the clause is unreasonable, it all falls away. The court will not re-write it or enforce it only to the extent that it complies with UCTA.
· It is always unreasonable to exclude or restrict liability for a contracting party’s own fraud.
The draftsman in our case has sought to limit the ambitions of all of his exclusions and limitations in one fell swoop by means of clause 10. The intention, clearly, is for clauses 11 and 12 to be read subject to clause 10 so that (for example) the exclusion of liability for loss of profits in clause 11 does not apply if the cause of the loss is fraud by Clouds Unlimited. A more cautious draftsman might have begun clause 10 with the words ‘Notwithstanding clauses 11 and 12, Clouds Unlimited does not exclude or restrict…‘ or perhaps ‘Nothing in this agreement shall exclude or restrict…‘. A third, commonly used alternative would be to make clauses 11 and 12 expressly ‘Subject to clause 10‘.
A carve-out such as that in clause 10 is not the only means of complying with the fraud, death and personal injury exclusion rules; indeed, even express reference to words such as ‘fraud’ may not be strictly necessary. For example, the Court of Appeal has held as a matter of construction that a clause excluding liability for various kinds of loss ‘in any circumstances’ did not purport to exclude liability for fraud or wilful damage because it would not naturally be read as doing so, and was therefore not unreasonable despite the lack of an express carve out (see Regus (UK) Limited v Epcot Solutions Limited [2008] EWCA Civ 361). That said, clients are unlikely to thank their draftsmen for using exclusion clauses to test the judicial boundaries! Including a carve-out definitely reduces the risk of a court construing the clause as a whole as unreasonable and is the most sensible course in most circumstances.
Clause 11: The Exclusion
Clause 11 is the exclusion clause proper, by which the draftsman hopes to limit the scope of Cloud Unlimited’s potential liability substantially by the exclusion of various categories of economic loss. A number of aspects of this clause deserve comment.
The Capital Letters
This odd practice of using capital letters for exclusion clauses probably arose in response to judicial suggestion that the more onerous or unusual a term, the more must be done to bring it to the attention of the other party. Draftsman have probably had in mind the typically graphic observation of Lord Denning in J. Spurling Ltd v Bradshaw [1956] 1 WLR 461 that ‘some clauses which I have seen would need to be printed in red ink on the face of the document with a red hand pointing to it before the notice could be held to be sufficient’.
In the modern era, it seems to us that capital letters are an entirely unnecessary oddity. Since 1977 courts have been able to regulate exclusion clauses by the operation of UCTA: if the clause is unreasonable it will simply not be given effect. Thus there is no injustice in the standard doctrine that if the clause is printed within standard terms of business and those standard terms are incorporated within the agreement, the exclusion clause will also be incorporated. There is thus no need for the courts to invoke a special doctrine of incorporation to avoid the consequences of an unreasonable exclusion of liability.
It is true that ‘whether the customer knew or ought reasonably to have known of the existence and the extent of the term‘ is one of the relevant factors under sch 2 to UCTA. This will no doubt be enough to persuade the most risk-averse draftsman to keep using capital letters. But, in our view, no court would think that businessmen ought reasonably to read only those standard terms of business which are in capital letters.
Indirect and consequential loss
Clause 11 excludes liability for (amongst other things) ‘indirect’ and ‘consequential’ loss. These somewhat opaque terms are usually construed to refer to losses falling within the second limb of the rule in Hadley v Baxendale (1854) 156 ER 145, ie to such loss as ‘may reasonably be supposed to have been in the contemplation of both parties, at the time they made the contract‘ as opposed to ‘direct’ losses which arise naturally, ‘in the ordinary course of events’, from the breach (see Watford Electronics Ltd v Sanderson CFL Ltd [2001] EWCA Civ 317, per Chadwick LJ at [36]).
References to indirect or consequential loss are a standard feature of exclusion clauses. There are probably two main reasons for this. First, as the exclusions or limitations must be reasonable, they are generally drafted to allow some recovery as opposed to no recovery at all. Second, because indirect and consequential losses are often less obvious losses than those arising directly, they are harder to predict and quantify when the contract is drawn up. Thus a party wishing to keep his liability within predictable confines will seek to exclude indirect or consequential losses but be more sanguine about direct losses.
In practice, there can be considerable difficulties in trying to determine whether a particular kind of loss claimed is ‘direct’ (typically not excluded) or ‘indirect’ (typically excluded), unless the kind of loss that is under scrutiny is specifically cited by the clause (see Rix J in BHP Petroleum Ltd v British Steel Plc [1999] 2 Lloyd’s Rep 583). This is exacerbated by the use of the second limb of Hadley v Baxendale as the means of identifying indirect and consequential loss: whether a particular kind of loss may have been contemplated by the parties (and so whether it is covered by the exclusion clause) depends on their state of knowledge, which may be imputed by the court (the first limb) or actual (the second limb).
A simple strategy to avoid these difficulties is to spell out within the exclusion clause examples of the categories of loss which are intended to be excluded. In relation to loss of profit, anticipated savings, contracts and revenue, our draftsman has done this in clause 11. However, even here, care is necessary: in GB Gas Holdings Ltd v Accenture (UK) Ltd [2010] EWCA Civ 912 Centrica alleged that Accenture’s breaches caused it to over-pay gas distribution charges by almost £19m; the trial judge had concluded that those alleged losses were not caught by a provision excluding ‘loss of revenues‘, and the Court of Appeal agreed.
A common mistake
Our draftsmen may have thought to single out loss of profit, anticipated savings, contracts and revenue as excluded losses, but in concluding his list ‘or any other indirect or consequential loss‘ he has made a potentially calamitous (and very common) mistake. Problems frequently occur when the word ‘other’ is used in this context because, often, the preceding examples are instances of direct loss rather than indirect loss. Loss of profits (cited in our clause 11) is the classic example: it is usually considered to be a form of direct loss in the context of a commercial contract.
In such cases the court is in a dilemma. Did the parties really intend to exclude the items in the list only insofar as they are consequential or indirect loss, or should the word ‘other’ simply be ignored as a mistake? The solution, of course, is to avoid using the word ‘other’ in this context. If clause 11 read ‘… or any indirect or consequential loss‘ no difficulty would arise.
Clause 12: The Liability Cap
The imposition of a cap on liability (as in our clause 12) is also a common feature of exclusion and limitation clauses. It is a useful device. It reduces potential liability further by capping liability for those types of loss which are not excluded. It also salvages something from the wreckage if the preceding clauses are found to be unenforceable under UCTA, as liability caps are not regarded by the courts with the same hostility as exclusion clauses.
The principal trick to drafting an enforceable liability cap is to find a limit which is reasonable in the context of the particular agreement in question. Simply providing a figure within standard terms of business, as the Clouds Unlimited draftsman has done, runs the risk that the limit will be hopelessly low in the context of a particular transaction using those standard terms and therefore unenforceable; conversely, there is a risk that the limit will be unnecessarily high and therefore useless in the context of a particular deal.
The more successful caps try to provide for a limit which is sensitive to the particular agreement in question. This can be achieved, for example, by setting the cap with reference to the price of the goods or services. Another approach with a proven track record in the courts is for a supplier to provide the option of an increased liability cap for a higher price. A figure which might otherwise have been considered ‘derisory’ may be held reasonable on the basis that the terms made clear that a higher cap could have been purchased on request.
The prohibition against excluding or restricting liability for fraud or negligence resulting in death or personal injury applies to liability caps as much as it does to exclusion clauses. As currently drafted it is not clear that clause 12 does this, notwithstanding the carve-out at clause 10, because clause 12 is said to apply ‘in any event‘. Thus there is a danger of it being held to be unreasonable. The best solution is to make clause 12 ‘subject to‘ clause 10, or to amend clause 10 as suggested above.
Clause 13: Entire Agreement
A party involved in a dispute will often try to rely on statements other than those set out in the written contract, either statements contained in other documents or statements which the party alleges were made orally. The party may seek to show that:
· other documents are incorporated as terms of the written contract
· some terms of the contract were agreed orally so the contract is only partly written
· one or more written or oral terms had effect as a separate ‘collateral’ contract or warranty
· the party was induced to enter the contract by reliance upon a particular statement.
The thorough draftsman will try to close down all of these avenues with an ‘entire agreement’ clause, which deals with the first three points above, or a ‘no reliance’ clause, which deals with the last. Clause 13 of the Clouds Unlimited standard terms seeks to do both.
The first sentence is the entire agreement provision. Its inclusion prevents arguments to the effect that the agreement was in part embodied in other documents or in conversations they had. In short, it prevents a party from ‘threshing through the undergrowth‘ of negotiations to find a remark that is useful to them that did not make it into the final written agreement (see Inntrepreneur Pub Co Ltd v East Crown Ltd [2000] 2 Lloyd’s Rep 611).
Technically, an entire agreement clause acts either to deprive collateral warranties or contracts of any effect, or to estop the parties from relying on extrinsic evidence to prove terms other than those recorded in the written contract. It is important to note that the clause does not render extrinsic evidence inadmissible for all purposes: the parties are still free to rely on other documents or conversations to assist (for example) in the interpretation or meaning of the written agreement or to resolve a dispute as to the circumstances or validity of the formation of the contract. It is also important to note that an entire agreement clause will probably not, in itself, succeed in excluding liability for misrepresentation even if worded in broad terms. A clause asserting that the agreement in question ‘shall together represent the entire understanding and constitute the whole agreement between the parties … and supersede any previous discussions, correspondence, representations or agreement between the parties with respect thereto‘ was not sufficient to ‘withdraw’ misrepresentations and thereby exclude liability for them in BSkyB v HP Enterprise Services UK Ltd [2010] EWHC 86 (TCC).
Achieving that goal requires the second sentence of clause 13, namely the ‘no reliance’ machinery. The aim of this machinery is squarely to prevent a party from mounting a misrepresentation claim on the basis that it was induced to enter into the contract by false pre-contractual representations. Technically the clause achieves its goal by raising an ‘evidential estoppel’, which will be effective even if a party did in fact rely upon the representation. However, the clause will generally have this effect only if the party who made the representation entered into the contract in the genuine belief that the other had not relied upon his representation: this kind of drafting is not intended to provide a way of deliberately misleading those with whom one is doing business.
In Watford Electronics Ltd v Sanderson CFL Ltd [2001] EWCA Civ 317 the Court of Appeal gave important guidance on the enforceability of ‘entire agreement’ and ‘no reliance’ clauses, and in particular clauses which, like our clause 13, contain provision for both. The courts will generally seek to give effect to this common form of drafting, recognising the commercial reality that both parties will generally want the certainty of knowing that the complete bargain between them can be found in the written agreement.
Conclusion
A principal goal when drafting exclusion clauses, liability caps and entire agreement clauses is certainty. The various clauses and strategies considered above all arise from attempts by contracting parties to distribute risk contractually in order to bring a degree of certainty to their dealings: certainty about the circumstances in which one party may be held liable to the other; certainty about the kinds of damage for which they may be liable; certainty about the maximum amount for which they risk being liable; and (in the case of entire agreement clauses) certainty about precisely what it is that they have agreed.
To succeed in this objective, a clause should be drafted to take account of the various issues discussed in this article. It must also be clear and unambiguous, well-defined in scope, and precise in application. In many circumstances it must also satisfy UCTA and its requirement of reasonableness, a subject which we have touched on briefly but that must await another Back to Basics article for more detailed treatment.
Michael Taylor is a barrister at 4 Pump Court. He is recommended as a leading IT junior and sits on the SCL London Group Committee.
Matthew Lavy is a barrister and is a member of the IT Group at 4 Pump Court. Prior to being called to the Bar, he did time as a technical writer, system administrator and software developer.
