REUTERS | John Kolesidis

Collateral warranties and the Construction Act – a nasty surprise

After nearly 30 years in the construction law game, I should no longer be surprised when an unexpected decision comes along. But Akenhead J’s judgment in Parkwood Leisure Ltd v Laing O’Rourke Wales & West Ltd made me sit up and draw breath.

Last week’s Practical Law legal update notes that the decision will be a “surprise to most practitioners”. With great respect to the learned judge, I would go further and say that it is simply wrong. It is also likely to have highly undesirable ramifications for the negotiation of collateral warranties in future.

Rationale for the decision

In holding that the collateral warranty in the Parkwood case was a “construction contract” for the purposes of the Housing Grants, Construction and Regeneration Act 1996 (Construction Act 1996), Akenhead J appears to rely on three overlapping justifications:

  • Statutory interpretation.
  • Construction of the warranty itself.
  • Timing (the fact that the warranty was entered into before practical completion).

I will deal with each of these in turn.

Statutory interpretation

As Akenhead J points out, a collateral warranty is clearly a contract, and moreover one that relates to construction operations. However, in my view it is stretching a point to describe it as a contract “for the carrying out of” construction operations. For example:

  • The beneficiary has no (direct) power to instruct how the works are carried out, to order variations or to terminate or suspend the contractor’s obligations.
  • If the primary contract is terminated, the beneficiary can no longer insist that the works are completed, unless the warranty includes a step-in right (which is a different matter).
  • Nor is the beneficiary obliged to make payment, other than perhaps a nominal consideration (in this case £1). Indeed, it is difficult to see how the payment provisions of the Construction Act 1996 could sensibly apply to a collateral warranty. On the contrary, the beneficiary is simply the recipient of a unilateral undertaking, collateral (the clue’s in the name) to the primary contract under which the works are carried out.

It is worth remembering the mischief at which the Act is directed, namely payment abuse by main contractors. I would suggest that it was never the intention to give third parties a fast track means of redress against a defaulting contractor. Had Parliament addressed its collective mind to this question (which it undoubtedly did not), it would have made clear that collateral warranties were not covered by the Act. Strictly speaking, this is probably irrelevant to the question of interpretation, but it is nevertheless instructive.

Construction of the warranty

Akenhead J appears to have placed much store by the words “warrants, acknowledges and undertakes”, to imply that the draftsman meant more than a “simple” warranty. I would venture to suggest that he intended nothing of the sort. This is torrential drafting, pure and simple. If nothing else, the case should serve as a stern warning that in contracts (unlike in oratory) one should never use three words when one will do.

Akenhead J also seems to have been heavily swayed by the phrase “has carried out and shall carry out and complete the Works in accordance with the Contract”. But this simply reflects the concept that the contractor will owe the same (but no greater) obligations to the beneficiary as it owes to its employer. If anything, they reinforce the classification of the warranty as a derivative obligation, rather than a separate construction contract.

Finally, Akenhead J seems to have ignored (or misinterpreted) the important proviso “the Contractor shall have no liability… in respect of any delay in the progress and/or completion of the Works”. In short, this means that the beneficiary has no remedy if the works are finished late; or, to put it another way, it cannot enforce timely performance, which is surely one of the critical elements of a “true” construction contract. I’m not clear how he has come to the conclusion that this proviso “does not exclude liability otherwise for non-completion”.


At this point the judge’s reasoning is even harder to follow. Akenhead J was persuaded that, because the collateral warranty was entered into before practical completion, it should be treated as a contract to carry out the works, rather than simply a warranty in respect of work carried out. But what about work done before the warranty was given? How can the contractor be said to have “contracted” with the beneficiary to carry out this work, when (in the absence of a warranty) no contractual relationship will have existed between them?

This situation is different from one where a formal contract is entered into between the employer and the contractor after work starts. In that case a contract of sorts (whether oral or based on an exchange of correspondence) will invariably exist and will give rise to mutual rights and obligations (to carry out and to pay for the works), even in the absence of a formal written contract. No such link will exist between the contractor and a beneficiary unless and until it is documented in a collateral warranty.

More importantly, this reasoning implies that the time of execution of the warranty is critical. But that simply does not reflect reality. Take a typical multi-let development under which the contractor agrees to provide warranties to tenants of individual units in an agreed form as and when called on to do so. Will those warranties given before practical completion be covered by the Construction Act 1996, and those given afterwards not? The warranty wording will be exactly the same, and – at least as far as the contractor is concerned – the timing will be entirely a matter of chance. In short, the supposed logic behind the distinction does not really exist.


This is not simply a matter of academic pin-dancing. I fear that it will have real implications for the negotiation of collateral warranties going forward. I envisage that contractors will wish to amend future warranties in order to escape the effect of the decision, for the reasons outlined in Practical Law’s legal update. Beneficiaries (and their advisers) will understandably seek to resist such amendments, and employers will do likewise out of concern that a warranty that is not caught by the Construction Act 1996 may be seen as “defective”. The last thing we need is another excuse to complicate the drafting, and prolong the negotiation, of collateral warranties in this way.

Akenhead J suggests in his judgment that the particular warranty used in the Parkwood case was not in standard form, and that accordingly he should construe it as a one-off document. That may in one sense be true, but there was nothing unusual or non-standard about it. On the contrary, there must be literally thousands of warranties “out there” drafted in very similar terms. We simply do not need the uncertainty to which this decision gives rise.

4 thoughts on “Collateral warranties and the Construction Act – a nasty surprise

  1. For more information and comment on Akenhead J’s judgment in this case, see:

    James Ladner’s blog post, which explains some of the defendant’s arguments that are not set out in the judgment.

    Matt Molloy’s blog post, which gives an adjudicator’s perspective.

  2. The decision does not sit easily with s.109 either.

    In the absence of an agreement that no payment is due in accordance with s.109(2), then the contractor would be entitled to payment under the warranty pursuant to s.109(1). In the absence of a compliant payment mechanism the Scheme would apply – s.109(3).

    In turn, Reg 2 of Part II of the Scheme would entitle the Contractor to be paid for the value of work done, less sums paid.

    The above scenario is pretty absurd, but follows a literal interpretation of the Act if Parkway is correct. Whilst it is clearly not what was intended, I cannot help wondering whether a canny contractor may try his luck to use this to argue that the Beneficiary should be liable where the Employer does not pay.

    Perhaps we should start expressly excluding liability on the part of the Beneficiary?

  3. Interesting – the phrase “has carried out and shall carry out and complete the Works…..” is arguably clear on the facts and has been construed logically given the warranty was executed after the date of the building contract and before PC. Presumably (and perhaps supporting the logic) when arranging for warranties to be executed at the same time as the building contract is entered into or on or after PC this phrase, if used, would then be unclear, because in the former case there wont be any past works and in the latter case there wont be any future works for the phrase to apply to. Presumably therefore the intention of the parties here, given the wording, can only have been to create an effective warranty at some point between entering the building contract and PC. It is perhaps more logical to interpret these words in relation to the “timing” aspect rather than in relation to the aspect of controlling the level of obligations confirmed to the beneficiary which John favours. I think different and very clear wording should have been used to expressly achieve control of the level of obligations to be gien under the warranty if this is what the parties intended. I agree with John though, the ramifications are undesirable because the timing and the words used has created a “construction contract” governed by the Construction Acts on the facts.

Comments are closed.

Share this post on: