Axa Insurance Ltd v Akther & Darby Solicitors & Ors

Under s 2 of the Limitation Act 1980 (“the 1980 Act”) claims in tort are time-barred six years after the cause of action accrued. Where the claims are in negligence, damage is an essential ingredient of the tort and so damage must be suffered before the six-year period starts to run.

The question on this appeal is: when is damage suffered for this purpose when the damage is alleged to flow from an unsecured contingent liability incurred by the claimant? Put another way, when does the clock start ticking? To answer this question, the judge, Flaux J, had to consider number of authorities but in particular the recent decision of the House of Lords in Law Society v Sephton [2006] 2 AC 543. That decision lies at the heart of this appeal and we too will have to consider the effect of that decision in some detail.

The contingent liability in question in this case is the contingent liability of the insurer under policies of “after the event” (“ATE”) insurance, that is, insurance which is taken out after the occurrence of the event which is the subject of a claim in legal proceedings to cover some or all of the potential costs and liabilities in those proceedings. The essential background is this. National Insurance and Guarantee Corporation (“NI”) issued ATE policies to clients of certain solicitors at the request of those solicitors. It is contended that those solicitors were negligent in the initiation of those policies (“vetting claims”) and in the conduct of the litigation on the client’s behalf, or the failure to notify NI when the prospects of success fell below 50% (“conduct claims”) pursuant to the terms of a funded solicitors’ agreement entered into with NI’s agent. The book of business turned out to be disastrous. NI has assigned its rights against those solicitors to Axa Insurance Ltd (“Axa”). Axa wishes to pursue the vetting and conduct claims against some 89 firms of solicitors (known in these proceedings as “the panel solicitors”). The time of the occurrence of the damage is important because some of the relevant policies were taken out more than six years before this action was commenced and claims against panel solicitors arising out those policies amount to some £19m.

The panel solicitors contend that time started to run for the vetting claims at the inception of the policies, and, for the conduct claims, when the failure to notify resulted in the prospects for success being reduced to below 50% or (where the breach consisted of the failure to take some step in the management of the claim) when the claim became liable to be struck out. Axa, relying on a line of authority culminating in, or following, the recent decision of the House of Lords in Sephton, contends that time only began to run in the case of both the vetting breaches and the conduct breaches at the time when the claim could have been made under the policy. NI’s liability under the ATE insurance was no more than an unsecured contingent liability of NI until that occurred.

Flaux J ordered a preliminary issue with respect to the parties’ contentions about limitation. This issue was tried on the basis of a statement of assumed facts which the judge summarises in his judgment. For the purpose of this judgment, the background, additional to that which I have already given, can be stated quite simply. NI was the insurer of ATE insurance under a scheme administered by Composite Legal Expenses (“CLE”), which was a coverholder for NI and provided claims management functions. It is not necessary to explain how the policies worked, save that the rights conferred on NI included the right to receive and retain the premium and the right to discontinue cover if it considered that reasonable prospects for the recovery of damages had ceased to exist.

The obligations of the solicitors and the alleged breaches as described by the judge in his judgment are as follows:

“6.

For the purposes of the preliminary issues, the court is asked to assume that the relationship is governed by a Funded Solicitors Agreement (“FSA”). Part of the role performed by the panel solicitors under the FSA was the vetting of claims which members of the public wished to pursue. Claims accepted under the Scheme had to have prospects of success of at least 51% and be for a minimum amount of £1,000. In respect of claims accepted, CLE acting on behalf of NIG would issue an ATE policy of insurance insuring the Scheme Claimant. In conjunction with the policy, a Funder (either First National Bank or Bank of Scotland as appropriate) would grant a loan to the Scheme Claimant which would be the source for payment of (i) the premium for the ATE policy and (ii) funded disbursements. NIG also entered into agreements with the Funders under which it agreed to indemnify the Funder in respect of the loan in certain circumstances, for example where NIG avoided the ATE policy for misrepresentation.

7.

The allegations made by Axa which are relevant to the preliminary issues are that, pursuant to the FSA and/or at common law, panel solicitors owed duties to NIG, as insurer, to:

(1)

Vet and only take on Scheme Claims that had (i) greater than a 50% prospect of success and (ii) a likelihood of damages of £1,000.

(2)

Conduct cases with reasonable care and skill thereafter. This duty falls into two relevant categories. First a duty to notify to NIG Scheme Claims for withdrawal of indemnity where (i) the prospects of success fell below 50% and/or (ii) it became clear that damages would not exceed £1,000. Second, a duty to conduct claims with due care and diligence (where appropriate to a successful conclusion).

8.

The requirement that a claim should not only have greater than 50% prospects of success but should also be likely to result in a damages award of at least £1,000 resulted from the costs regime on the small claims track. If the value of the claim was less than £1,000 then it was extremely unlikely that the Scheme Claimant would be able to recover his costs and disbursements, including the sums insured by NIG under the NIG Policy, (which NIG would have to pay out instead). These requirements that the claim should be worth at least £1,000 and should have greater than 50% prospects of success are alleged by Axa in para 1.3 of the Amended Generic Particulars of Claim to be “vital to the success of the [CLE] Scheme (and to the financial interests of the Insurer thereunder)”.

9.

Axa contends that by reason of panel solicitors' breaches of those duties, NIG suffered loss and damage, consisting in essence of the following:

(1)

In relation to breach of the duties set out at para 7(1) above (“vetting breaches”) the amount paid by NIG to the relevant Funder to discharge the Scheme Claimant's loan account and/or other amounts paid pursuant to the NIG Policy (such as the successful party's costs) when a Scheme Claim failed; and

(2)

In relation to breach of the duties set out at para 7(2) above ('conduct breaches'), either:

i.

The extra interest and/or disbursements incurred during the period where a Scheme Claim was wrongly continued; or

ii.

The lost opportunity of securing a successful outcome at trial or settlement and so avoiding a call on the NIG Policy, where a Scheme Claim should have been brought to a successful conclusion.”

The judge held in favour of the panel solicitors. He held that, in the case of vetting breaches, damage occurred when each ATE policy incepted and not when each insured claim ultimately failed. In a thorough and lucid judgment, he held that there was nothing in the House of Lords’ decision in Sephton that precluded the conclusion that damage occurred when the policy incepted. He therefore rejected Axa’s argument that this was a case of “purely contingent liabilities standing alone”, as had been the case in Sephton. The panel solicitors were under a tortious duty to bring about a transaction with a particular feature. This feature was that the prospects of success in a particular were assessed by a competent lawyer as being greater than 50%. As such, this case fell within the line of Court of Appeal decisions running from Forster v Outred & Co [1982] 1 WLR 86 to Shore v Sedgwick Financial Services [2008] PNLR 37. When the professional’s duty is to procure that a transaction has a particular characteristic or feature and the professional breaches that duty, the cause of action accrues on entering the flawed transaction.

Before the judge, Axa submitted that the cases in this area of law could be divided into three categories and it then argued that its case fell into the category of purely contingent liability. The judge rejected the categorisation approach as inflexible. Axa has not relied on that categorisation as such on this appeal although it continues to rely on the propositions of law which it contended were demonstrated by its categorisation. Much of the reasoning in the judge’s judgment is concerned with Axa’s categorisation, but in the circumstances it is not necessary for me to consider that part of his reasoning.

In view of the centrality of Sephton, I shall unusually examine that decision and the subsequent case law before I examine the arguments on this appeal.

Sephton concerned the accrual of a cause of action when a breach of the duty of care has resulted in a party being subject to a contingent liability. The contingent liability was unusual. The claimant was the Law Society and the liability in question was its liability to pay claims out of its compensation fund. I take the description of its obligations from Lord Hoffmann’s speech:

“3.

By section 36 of the Solicitors Act 1974 the society is required to maintain and administer a compensation fund for the purpose of making grants for, among other things, the relief of loss caused by dishonesty on the part of a solicitor. The society has power to make rules about the fund and the Solicitors' Compensation Fund Rules 1995 contain "guidelines" which explain the circumstances in which grants will ordinarily be made. General principle (a) says that the "basic object of the fund is to replace 'client's money' misappropriated by a solicitor". General principle (b) emphasises that grants are wholly at the discretion of the council and that "no person has a right to a grant enforceable at law" but that the intention of the council is to "seek to administer the fund in an even-handed and consistent manner". Claims must be made in a form prescribed by the society (rule 5) and delivered to the society within six months after the loss has come to the knowledge of the applicant (rule 6).”

The defendant accountant had certified to the Law Society that certain accounts of his practice complied with the Solicitors’ Accounts Rules. The defendant had negligently failed to identify that the solicitor had misappropriated monies belonging to his clients. The solicitor's clients in due course made claims against the Solicitors' Compensation Fund. The Law Society sued the accountant, who raised a limitation defence. The question that had to be decided was whether the loss suffered by the Law Society occurred when the solicitor had originally committed the fraud, thereby exposing the fund to potential claims, or when the claims were actually made against the fund. The House of Lords unanimously held that, until a claim was actually made, no loss or damage had been sustained by the fund and no cause of action had accrued.

The leading speech was given by Lord Hoffmann. The other members of the House agreed with his speech, Lord Walker and Lord Mance also gave speeches with which all the members of the House agreed, including Lord Hoffmann.

Lord Hoffmann held that it was unnecessary to go further back in the authorities than Forster v Outred [1982] 1 WLR 86. In that case Mrs Forster had executed a mortgage over her own property as security for her son’s bank borrowing on the basis of her solicitors’ advice. There was an issue as to when time began to run and this court held that it began to run when she executed the mortgage as the value of her property was then diminished. There is an ambiguity in the judgments of this court because on one reading the court accepted that time would also begin to run if an unsecured contingent liability was incurred.

As Lord Hoffmann points out, this interpretation was unanimously rejected by the High Court of Australia in Wardley Australia Ltd v State of Western Australia (1992) 175 CLR 514. In that case, the State of Western Australia had agreed to grant an indemnity to the bankers of a bank (R) in financial trouble in consideration of those bankers extending further loan facilities to R. The indemnity created merely a contingent liability. The question of limitation of actions arose in connection with a claim for misrepresentation made by the State against another bank, W, which had made representations to the State which led to its giving the indemnity. The High Court held that time did not begin to run until the liability of the State under the indemnity had ceased to be contingent. Lord Hoffmann approved the analysis by the majority of the High Court in their judgment of Forster, which the majority held was explicable:

“by reference to the immediate effect of the execution of the mortgage on the value of the plaintiff's equity of redemption … It has been contended that the principle underlying the English decisions extends to the point that a plaintiff sustains loss on entry into an agreement notwithstanding that the loss to which the plaintiff is subjected by the agreement is a loss upon a contingency. For our part, we doubt that the decisions travel so far. Rather, it seems to us, the decisions in cases which involve contingent loss were decisions which turned on the plaintiff sustaining measurable loss at an earlier time, quite apart from the contingent loss which threatened at a later date … If … the English decisions properly understood support the proposition that where, as a result of the defendant's negligent misrepresentation, the plaintiff enters into a contract which exposes him or her to a contingent loss or liability, the plaintiff first suffers loss or damage on entry into the contract, we do not agree with them. In our opinion, in such a case, the plaintiff sustains no actual damage until the contingency is fulfilled and the loss becomes actual; until that happens the loss is prospective and may never be incurred.”

The High Court thus held that time did not begin to run until the indemnity was called. The reason for the distinction between Wardley and Forster was that in the latter case there was “measurable loss at an earlier time, quite apart from the contingent loss which threatened at a later date”. The speeches in Sephton use the phrase “measurable loss”: in my judgment, that expression in general bears the same meaning as in the passage just cited from Wardley, namely, it means, in relation to a contingent liability, a loss which is capable of quantification and which is additional to the mere incurring of that liability. An unsecured liability is generally capable of quantification but, under the rule laid down in Wardley, for time to start to run (or the clock to tick) there must be loss in addition to that caused by the contingent liability. Although the majority of the High Court in Wardley referred to Forster, the passage from its judgment, specifically approved by Lord Hoffmann and set out above, does not contain any further requirement that the additional loss is property related.

Lord Hoffmann examined some other decisions of this court but held that they did not concern “purely contingent liabilities”. In the course of dealing with the authorities in this court at [21] and [22] of his speech, Lord Hoffmann recognises that time could begin to run from the time of entry into a transaction where contingent liabilities were incurred under that transaction. He held:

“21.

Next, there are a number of cases in the Court of Appeal which involve transactions, with both benefits and burdens, into which the plaintiff entered as a result of the negligence or breach of contract of the defendant. None of these cases concerned purely contingent obligations. It is only necessary to observe that in such bilateral transactions the answer to the question of whether damage has been suffered may be different according to whether the liability is for the consequences of the defendant not performing his duty or (as is usual in claims for misrepresentation) the consequences, or some of the consequences, of the plaintiff entering into the transaction. If the liability is for the difference between what the plaintiff got and what he would have got if the defendant had done what he was supposed to have done, it may be relatively easy, as Bingham LJ pointed out in D W Moore & Co Ltd v Ferrier [1988] 1 WLR 267, to infer that the plaintiff has suffered some immediate damage, simply because he did not get what he should have got. Thus in Knapp v Ecclesiastical Insurance Group plc [1998] PNLR 172, where the plaintiff paid a premium for a voidable fire insurance policy because his insurance broker had failed to disclose material facts, the Court of Appeal held that he had suffered immediate damage because he "did not get what he should have got", namely a policy binding on the insurers. On the other hand, if the damage is (as it was in the Nykredit (No 2) case [1997] 1 WLR 1627 and First National Commercial Bank plc v Humberts [ 1995] 2 All ER 673) the difference between the defendant's position after entering into the transaction and what it would have been if he had not entered into the transaction, the answer may be more difficult. Despite the breach of duty, the transaction may on balance have originally been advantageous to the plaintiff and some evidence may be necessary to show when he was actually in a worse position. The judgment of Mason CJ and his colleagues in the Wardley case drew attention to this distinction 175 CLR 514, 530-531:

“Another element in some of the English decisions … is the conclusion that, because the subject matter of the agreement lacked the qualities which it had been represented as having, that subject matter was therefore less valuable than it would have been if the representations had been true. That conclusion is acceptable in cases in which the contract measure of damages is appropriate but it is not acceptable here where the contract measure of damages does not apply. The application of that measure of damages [sc the difference between the value of what the plaintiff got and what he would have got if the defendant had performed his duty] may, in some situations, enable a court to conclude more readily that the plaintiff first suffers loss or damage on entry into an agreement."

Thus, if a claimant gave a personal guarantee, unsecured on any property of his, time would not begin to run for the purposes of any claim against his solicitor for negligent advice at the time he entered into the transaction until a call was made on him as guarantor.

Lord Hoffmann summarised his conclusion in Sephton thus:

“29.

It also seems to me irrelevant that a prudent accountant, drawing up the accounts of the compensation fund to give a true and fair view of its assets and liabilities, would have included provision for contingent liabilities. As Lord Radcliffe pointed out in Southern Railway of Peru Ltd v Owen [1957] AC 334, 357, the principles upon which such provisions are made does not depend upon "any exact analysis of the legal form of the relevant obligation" but upon estimates of what in practice is likely to happen. A cause of action, however, connotes a legal obligation and its existence must be determined by rules of law.

30.

In my opinion, therefore, the question must be decided on principle. A contingent liability is not as such damage until the contingency occurs. The existence of a contingent liability may depress the value of other property, as in Forster v Outred & Co [1982] 1 WLR 86, or it may mean that a party to a bilateral transaction has received less than he should have done, or is worse off than if he had not entered into the transaction (according to which is the appropriate measure of damages in the circumstances). But, standing alone as in this case, the contingency is not damage.

31.

The majority of the Court of Appeal appear to have decided the case on the basis that the Law Society did not enter into any transaction giving rise to the contingent liability. It did nothing and the contingent liability was created by the misappropriations and the previous existence of the compensation fund and the rules which governed its administration. No doubt in most cases in which a party incurs a contingent liability as a result of entering into a transaction, that liability will result in damage for the reasons already discussed in relation to bilateral transactions. But I would prefer to put my decision on the simple basis that the possibility of an obligation to pay money in the future is not in itself damage.”

At [44] of his speech, Lord Walker reviewed a number of authorities and cited an important passage from the judgment of Saville LJ in First National Commercial Bank plc v Humberts [1995] 2 All ER 673, 679:

“At the hearing and in the judgment much reliance was placed on the cases where the claimant entered into a transaction which through a breach of duty owed to the claimant provided the claimant with less rights than should have been secured, or imposed liabilities or obligations on the claimant which should not have been imposed. Examples of these cases are: Forster v Outred & Co [1982] 1 WLR 86, Iron Trade Mutual Insurance Co Ltd v J K Buckenham Ltd [1990] 1 All ER 808, and Bell v Peter Browne & Co. [1990] 2 QB 495. In all those cases, however, the court was able to conclude that the transaction then and there caused the claimant loss, on the basis that if the injured party had been put in the position he would have occupied but for the breach of duty, the transaction in question would have provided greater rights, or imposed lesser liabilities or obligations than was the case; and that the difference between these two states of affairs could be quantified in money terms at the date of the transaction.”

As the next paragraph in Lord Walker’s speech makes clear, Lord Walker accepted that, in an action against a professional adviser for negligence, time will begin to run from the time of entry into a transaction where the advice results in the client receiving a less valuable bundle of rights:

“45.

The three cases cited by Saville LJ in this passage were all cases where the client had through the negligence of his professional adviser ended up with a package of rights less valuable than he was entitled to expect - damaged or defective goods, to pursue the metaphor, rather than the undamaged and serviceable goods which he should have got. In Forster v Outred & Co [1982] 1 WLR 86 it was a mortgage (securing the existing and future liabilities of the claimant's son, who later went bankrupt) burdening the claimant's previously unencumbered freehold property…”

Lord Walker expressed his overall conclusion in paragraph 48 of his speech:

“48.

In all these cases the claimant has as a result of professional negligence suffered a diminution (sometimes immediately quantifiable, often not yet quantifiable) in the value of an existing asset of his, or has been disappointed (as against what he was entitled to expect) in an asset which he acquires, whether it is a house, a business arrangement, an insurance policy, or a claim for damages. Your Lordships have not, I think, been shown any case in which the imposition on a claimant of a purely personal and wholly contingent liability, unsecured by a charge on any of the claimant's assets, has been treated as actual loss. That would have been the position if the claimant in the Forster case [1982] 1 WLR 86 had given a personal covenant guaranteeing her son's debts (which she seems not to have done - she paid them simply to prevent enforcement of the security on her farm) and if she had not given any security over any of her own assets.”

Thus, in common with Lord Hoffmann, Lord Walker considered that, where a person gave an unsecured guarantee, time would not run against him for the purposes of any claim against a professional adviser for advice in connection with the guarantee until a demand was made under the guarantee.

At [70] of his speech, Lord Mance deals with the case where a party enters into a transaction which he would not otherwise have entered into and, as a result, his property suffers a diminution in value:

“70. In all these cases except Forster v Outred & Co [1982] 1 WLR 86 the defendant failed to preserve or procure for the claimant an asset (including a particular chose in action) which could and should have been preserved or protected by proper performance of the defendant's duty in relation to the transaction affecting the claimant's legal position. In Forster v Outred & Co the claimant's case was that, but for the defendant's negligence, she would never have entered into the transaction at all. But in that case, by doing so, she clearly depreciated the value of her house in a measurable way. However, while a defendant's failure to preserve or protect a particular asset by proper performance of his duty in relation to a particular transaction may readily be seen to have caused measurable loss, negligence causing a claimant to enter into a transaction which he would not otherwise have entered may not immediately, or indeed ever, cause measurable loss to any particular asset.”

Lord Mance’s conclusion on the application of the authorities to the particular case adopts the approach that the critical point is that the legal position of the Law Society was not changed until at least a claim was received. That supports the view that Sephton was a most unusual case since normally the legal position of a person, whether he is a natural person or a corporation, is changed when he incurs a contingent liability. Lord Mance held:

“76.

Whether or not that is, however, accepted, no English authority indicates and I do not consider that the society's present cause of action should be regarded as accruing before any change in its legal position occurred and it received any claim on the fund. First and foremost, the society's legal position remained unchanged, even in public law, at least until after it received a claim. Second, it was not possible until after a claim was received for anyone to know which client(s) of Payne & Co might suffer what loss, whether any of them might be able, and choose, to assert that they had as a result suffered hardship justifying a grant out of the fund and what the circumstances were in which the society would have to exercise its discretion to make or refuse a grant. Third, in this situation, it is not appropriate to talk of the fund or any other specific asset of the society as having suffered any loss at least until after a hardship claim was made on the society.”

No specific mention is made here of the person who executes an unsecured guarantee, but in the following passage Lord Mance goes on to give two cumulative conditions for time starting to run in the case of a purely contingent liability which would mean that he too would hold that time does not start to run with respect to a claim brought against a professional adviser in connection with advice given to the claimant regarding entry into an unsecured guarantee:

“77.

It may be that, if the facts had been known contemporaneously, some statistical or experience-based assessment could have been made of the likelihood of a claim or claims emerging, and of the fund having eventually to make payments, as a result of Mr Payne being able to continue his scheme of fraud. A similar assessment might be made of the risk of future loss of a physical asset (deeds or valuables) of which a solicitor was failing to take reasonable care, but which had not yet been lost or stolen. But I do not consider that the law should treat purely contingent loss assessed on so remote a basis as sufficiently measurable, in the absence of any change in the claimant's legal position and of any diminution in value of any particular asset. Even where negligence brings about a specific transaction and thus a change in the claimant's legal position, Lord Nicholls observed in the Nykredit (No 2) case [1997] 1 WLR 1627, 1631c-d in the passage cited in para 73 above, that the mere entry into the transaction under which "Financial loss is possible, but not certain" is not sufficient detriment.

78.

Looking at the matter more generally, I also see no particular reason to accelerate the accrual of a cause of action where there has been no transaction changing the claimant's legal position and no diminution in value of any particular asset. Where such factors are present the English authorities considered in paras 67-70 above take a clear-cut, though perhaps strict, view. The House has not been asked to review such authorities, nor would I think it appropriate to do so in the light of the way that they and the English legislation have developed. But where such factors are not present, I see attraction in the approach taken by the Australian High Court in the Wardley case 175 CLR 514, the effect of which is that unless and until a remote contingency eventuates the claimant is not expected to issue proceedings which he would not normally issue or wish to issue unless and until that point arrives.”

In summary, the Law Society came under a liability to provide compensation out of the compensation fund when section 36 of the Solicitors Act 1974 came into force, and the Law Society no doubt raised a provision in its own financial statements to reflect claims incurred but not reported. However, so far as the liability to the clients of the solicitor in that case was concerned, that did not crystallise until the conditions of liability, including the delivery of a claim in the prescribed form, were fulfilled. Until that point in time, the loss was not “measurable” for the purpose of establishing whether damage had occurred and whether the commencement of the limitation period had begun.

The speeches in Sephton draw a distinction between contingent liabilities which affect particular assets of the person subject to the liability as well as his legal position, and contingent liabilities which have little or no effect on his legal position or assets. Sephton was an example of the latter. The contingent liabilities were imposed by statute subject to certain conditions being satisfied by the person claiming to enforce the liability, and those conditions had not yet been satisfied. This distinction does not deal with at least one group of cases, namely the group of cases where the assumption of a contingent liability alters a person’s legal position but does not affect particular assets. An example of such a case is the mere execution of an unsecured guarantee by a third party who is not a party to the transaction under which the liabilities guaranteed arise. Both Lord Hoffmann and Lord Walker expressly held that in such a case, time did not start to run until the guarantee was called. They each refer to a change in the position (rather than “the legal position”) of the claimant only in the context of a claim arising as a result of entry into a bilateral transaction under which the claimant did not get what he should have got.

In my judgment, it is difficult to see why an unsecured guarantor should be in a better position than a guarantor who grants security over his property. Different persons may each give a guarantee in the same transaction and one may be unsecured; and the other secured: Mrs Forster could have given a secured guarantee for the liabilities of her son’s company and her son could at the same time have given an unsecured guarantee. Her claim against her adviser would have been time-barred six years after the execution of the guarantee but that of her son would not. Her legal position was clearly altered, and her property was diminished in value. The limitation period for her son’s claim, however, would not start to run until a demand under the guarantee was intimated or made. Yet, even though no particular assets of the son were appropriated to liabilities under the guarantee, his legal position was altered by the execution of the guarantee as he had a liability to meet amounts due under the guarantee if and when a demand was made. Sephton might also conceivably lead to a distinction being drawn between the guarantor who can point to a diminution in value of his rights of contribution and subordination as a result of the alleged wrong and one who cannot do so. Moreover it appears that a diminution in the value of a particular asset or assets is different from a diminution in the value of one’s total net worth. These distinctions are difficult to rationalise. In my judgment, this aspect of Sephton makes it desirable that at the appropriate time it should be revisited by the Supreme Court. The cases in this area which have arisen since Sephton demonstrate the importance and difficulty of the relevant law. Meanwhile, in this case, Axa seeks to capitalise on that distinction, which it is entitled to do. But I would treat any invitation to extend that distinction with considerable caution.

Lord Mance states in [77] or his speech: “But I do not consider that the law should treat purely contingent loss assessed on so remote a basis as sufficiently measurable, in the absence of any change in the claimant's legal position and of any diminution in value of any particular asset.” It was submitted in argument that the change in the claimant’s legal position, should be an alternative to, rather than additional to, a diminution in value of a particular asset. For my own part, given that the House drew a distinction between secured and unsecured contingent liabilities, and each of their Lordships agreed with each other, I do not consider it is necessary or appropriate to alter the apparent meaning of this sentence. The real difference between the speeches is that Lord Mance bases his decision on the notion of a change in legal position which the other members of the House do not. But since a change in the Law Society’s legal position did not occur until the conditions for liability were fulfilled, and there was no diminution in value of a particular asset, there is no net difference between the formulations in the Sephton case as regards the result in that case. The liability of the Law Society in Sephton existed before the solicitor’s misappropriations and the accountant’s tortious act, and so its legal position was not actually changed until the conditions for making a claim had been fulfilled (see [76] of his speech). At that point the Law Society became bound to consider the claim and to make an appropriate compensation payment. But there could in another case be different consequences from Lord Mance’s formulation. For example, the concept of change in legal position is not limited to a change in a liability to make a payment, and thus would appear to be wider than the notion of being called on to make a payment under a previously accepted contingent liability.

It follows from Sephton that there can be cases, like that of Mrs Forster, where a contingent liability is incurred but it does not crystallise into an actual liability until a future date but where damage occurs for the purposes of the commencement of the limitation period at the time when the transaction is entered into so that time starts running from that time. I will call this “the damaged asset rule”. As a matter of outcome, this rule benefits the wrongdoer, but that is a policy decision made in Sephton and is the result whenever a cause of action becomes time-barred. The fact that the wrongdoer benefits and the claimant has no knowledge of his claim against him may partly explain why Sephton creates a rule for purely contingent liabilities. If the consequence of Sephton is that the result in Forster is to be translated into the world of insurance, it could mean that the mere underwriting of a policy of insurance would constitute damage, and start the running of the limitation period. The insurer would then be placed in the same position as Mrs Forster. The beneficiary would again be the alleged wrongdoer. In the case of longtail business (such as employers’ liability), the insurer may not know that he has suffered a loss as a result of a wrong until the limitation period for bringing an action in respect of that wrong has expired. The exception to that would be where he is able to bring himself within the provisions of s 14A of the Limitation Act 1980 dealing with latent damage.

It is clear from the speeches in Sephton that there are other situations where loss is suffered immediately. The speeches refer to cases such as D. W. Moore v Ferrier [1988] 1 WLR 267. In that case, the plaintiff did not discover the deficiencies in a restrictive covenant in an employee’s contract of employment until the covenant was challenged several years later. It was held that damage occurred when the agreement was executed since at that point in time the plaintiff received a worthless covenant rather than a valuable chose in action. Accordingly time began to run from this point. Likewise in Bell v Peter Browne & Co [1990] 2 QB 495, where the defendant solicitors failed to register the interest of a husband in a house following the terms of a separation agreement, time began to run from a failure to take the necessary steps and not from the time when the husband discovered that the house had been sold and the proceeds dissipated. In these cases, there was a bilateral transaction under which the claimant should have received certain benefits but owing to the negligence of his professional adviser did not do so. I will refer to this situation as “the package of rights rule”. There is no reason in principle why this line of authority should not apply where what the claimant by virtue of the bilateral transaction places himself under a contingent liability. Lord Hoffmann at [22] of his speech in Sephton expressly contemplates that situation (see also the first sentence of the passage from the judgment of Saville LJ in Humberts which Lord Walker cited at [44] of his speech). The judge considered that what I have called the package of rights rule applied to this case.

In my judgment, the damaged asset rule and the package of rights rule are best regarded not as a series of independent qualifications on the basic rule in Sephton that the assumption of a “pure contingent liability” does not cause the limitation period to start to run, but as different cases in which the courts have tried to express a central idea. That idea has to be found by seeking the ultimate ratio in Sephton, that is, a ratio which expresses the reason for the decision on which, despite the differences in expression, all the members of the House in that case were agreed. As I see it, the concept on which all the members of the House agreed was that there had to be measurable loss before time began to be run, that is to say, loss which is additional to the incurring of a purely contingent liability. In my judgment, for this purpose, rights of contribution or subrogation must be ignored because those rights arise by operation of law, unless excluded by agreement or statute. If they were taken into account, they would undermine the basic rule which is clearly established in Sephton that a pure contingent liability is not damage.

In my judgment, the central idea in Sephton is that there has to be loss additional to that resulting from the incurring of a purely contingent liability. This reflects the formulation in Wardley with which Lord Hoffmann expressly agreed. That passage in particular held:

“Rather, it seems to us, the decisions in cases which involve contingent loss were decisions which turned on the plaintiff sustaining measurable loss at an earlier time, quite apart from the contingent loss which threatened at a later date … If … the English decisions properly understood support the proposition that where, as a result of the defendant's negligent misrepresentation, the plaintiff enters into a contract which exposes him or her to a contingent loss or liability, the plaintiff first suffers loss or damage on entry into the contract, we do not agree with them.”

The central idea which I have identified is also consistent with the “simple basis” on which Lord Hoffmann puts his decision at the end of paragraph [31] of his speech: “...the possibility of an obligation to pay money in the future is not in itself damage” (emphasis added). Lord Walker expressly agreed with the passage cited in the last paragraph from Wardley. In addition, Lord Walker, cited with approval a passage from the judgment of Saville LJ in Humberts (see the passage cited in paragraph [19] above). This concentrated on the requirement for incurring of loss rather than the acquisition of less valuable rights under a bilateral transaction, which is the particular situation with which Lord Walker deals in paragraph [45] of his judgment. I do not read Lord Walker as rejecting any part of what Saville LJ had held. On this basis the acquisition of less valuable rights is to be regarded as a species of loss and not as describing the whole genus of loss. Lord Mance spoke of the necessity for a change in legal position and a diminution in value of a particular asset (see paragraphs [77] and [78] of his judgment). This also appears to be an acceptance of a requirement for additional loss, though arguably limited to damage to a particular asset. However, as Lord Mance agreed with Lord Hoffmann and Lord Walker I do not consider that his speech should be read as ruling out the incurring of additional loss in other ways.

In my judgment, the true ratio of Sephton is that there has to be measurable loss as defined in Wardley for time to begin to run for limitation purposes. On this basis what has to be shown is additional loss, and that loss does not have to be of a kind previously considered in the case law. Types of damage considered in the case law thus far include damage to existing property (my damaged asset rule) or a failure to obtain the hoped-for benefits under a bilateral transaction (my package of rights rule).

It is important to see how Sephton has been applied in subsequent cases. It was first considered by this court in Watkins v Jones Maidment Wilson [2008] I EGLR 149 (Arden, Longmore, Thomas LJJ). In that case the appellants sought to establish that time only began to run against them, for the purposes of their claim against their solicitors in relation to allegedly negligent advice to write a letter to their builder, when the consequences of writing that letter became clear some time after the letter was in fact written. In the letter, they agreed to a postponement of the date of completion, which under the agreement with the builder resulted in their losing the right to have disputes referred to arbitration, which would have given them a more cost-effective way of resolving their disputes with their builder. The builder did not default in completing the property until a date subsequent to the date of the waiver, and it was argued that that was the date for the commencement of the limitation. This court rejected that argument and held that damage was suffered when the letter was executed as the rights conferred by the agreement constituted an asset which was diminished in value as soon as the advice was acted upon.

This court subsequently considered Sephton in Shore v Sedgwick Financial Services Ltd [2008] EWCA Civ 863 (Buxton, Keene and Dyson LJJ). This court held that Mr Shore suffered damage for the purposes of a claim against his financial advisers when pursuant to their advice he transferred his pension benefits to a new scheme. On Mr Shore’s behalf it was argued that he did not suffer loss at that date because he subscribed for the new scheme at its market value, and that scheme could have done better. The court rejected that argument:

“It is Mr Shore's case (assumed for present purposes to be established) that the PFW scheme was inferior to the Avesta scheme because it was riskier. It was inferior because Mr Shore wanted a secure scheme: he did not want to take risks. In other words, from Mr Shore's point of view, it was less advantageous and caused him detriment. If he had wanted a more insecure income than that provided by the Avesta scheme, then he would have got what he wanted and would have suffered no detriment. In the event, however, he made a risky investment with an uncertain income stream instead of a safe investment with a fixed and certain income stream which is what he wanted.” (per Dyson LJ)

This court held that under the transfer Mr Shore acquired rights that were from the time he entered into the transaction less advantageous to him. This was accordingly not a pure contingent liability case and time began to run from the making of the transfer. In the case considered in the next paragraph, Lewison J considered that this court had applied a subjective basis of valuation which he regarded as unusual but, as I read it, Dyson LJ was merely contemplating a valuation using the investment criterion which Mr Shore had stipulated, which was more risk averse than permitted by an investment in the new scheme.

In Pegasus Management Holdings SCA v Ernst & Young [2009] PNLR 11, the claimant had subscribed for shares in a company in the belief (based allegedly on advice he was given) that this would provide him with roll-over relief. In the event, relief could not be obtained because the company structure did not satisfy the relevant requirements. On the question of when time began to run for limitation purposes, Lewison J held at [111] to [112] that the claimant suffered loss as soon as he acquired the shares because his advisers did not ensure that he subscribed for shares in a company which would enable him to preserve his entitlement to roll-over relief. An appeal in this case was heard by a different constitution of this court on 27 October 2009 and judgment has been reserved.

In Poole v H M Treasury [2007] 1 Lloyd’s Rep IR 114, certain underwriting members of Lloyd's (names) sued the government for failure to implement a European directive which caused them loss as underwriters. Applying Sephton, Langley J held that the claimants were worse off when they became names even though the liabilities which they undertook as names were contingent only. Although it is open to argument that a member might be exposed to contingent liabilities arising from his underwriting business only when the business is written or as the case may be the reinsurance to close of the previous year is approved by his syndicate, it is clear that Langley J held that the limitation period started to run at a point in time when the liabilities were still contingent. Langley J noted, however, that the issue was largely if not wholly academic (paras 239 to 240).

Another case decided by Langley J was also cited to us, Companhia de Seguros Imperio v Heath (REBX) Ltd [1999] I Lloyd’s Rep IR 571, but, as little turned on it for the purposes of this appeal and the case was decided before Sephton, it is not necessary for me to deal with it in this judgment.

The primary submission of Mr Charles Hollander QC, for Axa, with regard to the vetting claims is that Sephton is dispositive of this case. He submits the issue of the ATE policies at most placed Axa under a contingent liability, and that accordingly Axa was in the same position as the Law Society in Sephton. He points out that Sephton could have been decided on the narrow basis that the liability of the Law Society was a pre-existing contingent liability imposed by statute rather than pursuant to some arrangement between the parties. He submits, however, that the House of Lords decided Sephton by reference to a more general principle, as is evident from their express approval of the analysis of the majority of the High Court of Australia in Wardley (see per Lord Hoffmann at [17] and [18], per Lord Scott at [33], per Lord Walker in respect of one statement of principle at [49] and see per Lord Mance at [75] and [78]). Mr Hollander submits that because of the approval of Wardley in Sephton, it cannot be contended that Sephton is limited to cases where the claimant's legal position is unchanged. Such a situation was expressly rejected by Lord Hoffmann at [31]. Mr Hollander points out that in Wardley the claimant obtained nothing in exchange for assuming the contingent liability under the indemnity.

Mr Hollander also submits that when the ATE policies were issued NI acquired rights, the principal right being to the payment of premium, but that there was no damage to this asset. That premium was an asset which NI could assign for full value and the vetting breaches therefore did not result in an immediate loss to NI. The premium was not therefore in any way devalued by the negligence of the panel solicitors. He accepts that NI also obtained other rights under the ATE policy such as the right to control claims and so on, but those rights were not adversely affected by the panel solicitors’ negligence. He submits that the effect of the panel solicitor's negligence was not on the asset side of the transaction, but on the liability side. Indeed, although whenever NI issues an ATE policy, it enters into a contract of indemnity, that contingent liability may never be called upon. The claim may succeed or a settlement may be achieved which avoids any call on the policy. Thus, in vetting breach cases, there is no devaluation or other detriment to any asset, there is no liability other than a contingent liability. Unless the contingent liability matures into an actual liability, NI will make a profit out of the transaction. Thus Mr Hollander submits that the contingent liability incurred by NI under the policies of ATE insurance should be treated as a stand-alone contingent liability of the same kind as a contingent liability in Sephton. For this purpose, it makes no difference how likely the contingency is to occur: see per Lord Hoffmann at [9]. This would mean that there had to be a claim made on the ATE policy under which NI became liable to provide an indemnity before time started to run.

The judge rejected the argument that the premium could be treated separately from the transaction under which the policy of insurance was issued and that there was no damage to an asset (see paras [40], [50] and [51] of the judge's judgement). He thought this was an artificial approach.

Miss Sue Carr QC, for the panel solicitors, submits this is not a pure contingency case. The House of Lords in Sephton also recognised that, where the transaction which gave rise to the contingent liability also resulted in loss in value of a particular asset, damage was incurred at that point. She submits the damage is also incurred where the contingent liability arises from a bilateral transaction which the claimant has been advised to enter into on the advice of the defendant and which does not secure for the claimant features or benefits which the defendant promised to him: see in particular paragraph [30] of the speech of Lord Hoffmann and paragraph [45] of the speech of Lord Walker, which I have set out above. The judge took the same view at [42] of his judgment. Miss Carr also relies on Davys Burton v Thom [2008] 1 NZLR 437 at 449, where Elias CJ expressed the view that, in relation to an unsecured indemnity, the execution of such an instrument might have:

“measurable economic detriment irrespective of whether its performance is secured or accounted for. ... The question of actual detriment to the person giving a guarantee or indemnity is one of fact turning on the terms of the obligation and the existence of the principal debt.”

Miss Carr submits that Mr Hollander's argument that the premium is unaffected ignores the fact that the insurer was exposed to a greater risk as a result of the vetting breaches.

In response to this point, Mr Hollander puts forward his alternative submission. He accepts that, where the client has through the negligence of the professional adviser ended up with a package of rights which is less valuable than he was entitled to expect, he suffers damage when he enters into the agreement and acquires the damaged package of rights: see for example, D. W. Moore & Co. v Ferrier. He submits, however, that in all these cases the client should have entered into or had the opportunity to enter into a particular transaction having certain features or on certain terms and that as a result of the negligence of the professional adviser he has entered into a transaction lacking one or more of these features or terms. Moreover, it should have been within the scope of the professional adviser's duty to bring about for the claimant a transaction which had the features or terms in question and as a result of the defendant’s negligence the package of assets or rights acquired by the claimant would thus be less than valuable should have been the case. The measure of damages in such a case is the loss caused by the fact that the claimant does not have the benefit of such a transaction, or a “loss of bargain” measure.

Mr Hollander submits that, in the present case, the contractual measure is not applicable. NI would not have issued policies if the vetting breaches had not occurred. This court should apply the decision of the House of Lords in Nykredit Mortgage Bank v Edward Erdman Group (No. 2) [1997] 1 WLR 1627. In Nykredit, the plaintiff had made a loan to a borrower in reliance upon the negligent survey of the property charged in support of the loan. The plaintiff would not have entered into the loan but for the negligent survey. The borrower immediately defaulted and the plaintiff obtained judgment for damages against the negligent surveyor. The House had to determine the date from which interest would run on the damages and this depended upon when damage occurred. Lord Nicholls, with whom the other members of the House agreed, held that the first step was to identify the relevant measure of loss, which was a comparison between what the plaintiff's position would have been had the defendant fulfilled its duty of care and the plaintiff's actual position. If the plaintiff would not have entered into the transaction but for the negligent advice, the comparison fell to be made between its position had it not entered into the transaction in question and its position under the transaction. The valuer was liable for the adverse consequences of the loan attributable to the deficiency in the valuation. Lord Nicholls held that damage occurred as soon as the lender sustained measurable relevant loss. Sometimes, this would be immediately on making the loan, such as in this case where the borrower's covenant to repay was worthless. In other cases, this might be at the later time when the borrower defaulted. In yet other cases, it might be at other points in time. It depended upon the facts. However, Lord Nicholls rejected the submissions that the loss did not materialise until the security was realised or after the lender became entitled to realise its security. As Lord Nicholls said, at p1633D, “within the bounds of sense and reasonableness the policy of the law should be to advance, rather than retard, the accrual of a cause of action”.

The only other speech in Nykredit was given by Lord Hoffmann, who held that loss was suffered when the lender could show that it was “worse off” than it would have been had this security been worth the sum advised by the valuer: see at p1639A. In paragraph [20] of his speech in Sephton, Lord Hoffmann held that Nykredit decides that:

“in a transaction in which there are benefits (covenant for repayment and security) as well as burdens (payment of the loan) and the measure of damages is the extent to which the lender is worse off than he would have been if he had not entered into the transaction, the lender suffers loss and damage only when it is possible to say that he is on balance worse off.”

In this passage, Lord Hoffmann uses the words “worse off”, although on another occasion he uses the words “financially worse off”. In Sephton, Lord Walker expressed the view that the latter expression was to be preferred, and I have proceeded on that basis.

Mr Hollander submits that in Nykredit the claimant suffered loss immediately upon entry into the transaction because at that stage it believed that it had more valuable security for its loan and was thus in a worse position. However this did not mean that the loss constituted damage for the purpose of the accrual of the cause of action in tort since the loss of a transaction with security for the full amount was not the applicable measure of damages. The applicable measure of damage was the difference between the claimant's actual position and the position it would have been in if the defendant had not breached its duty. In that event it would have provided a valuation which accorded with the actual value of the building. The claimant would not then have entered into a loan at all. Mr Hollander submits that the loss of bargain measure was therefore inapplicable and the question was whether the claimant was worse off having entered into the particular transaction than it would have been if an accurate valuation had been given in which case it would not have ended in transaction at all. That depended on the value to be placed on the borrower's covenant to repay. Although the security was insufficient to cover the full amount of the loan, there was no loss to the lender so long as the borrower's covenant to repay had value. Thus, as Lord Hoffmann explained in Sephton:

“20.

The Nykredit (No 2) case [1997] 1 WLR 1627 therefore decides that in a transaction in which there are benefits (covenant for repayment and security) as well as burdens (payment of the loan) and the measure of damages is the extent to which the lender is worse off than he would have been if he had not entered into the transaction, the lender suffers loss and damage only when it is possible to say that he is on balance worse off. It does not discuss the question of a purely contingent liability.”

Thus, submits Mr Hollander, where it is not the defendant's duty to bring about a particular transaction but only to provide an accurate valuation, and the claimant enters into a transaction in which there are benefits and burdens which he would not have entered into had the defendant done his duty, the entry into transaction results in loss and damage only if and when an adverse balance is struck. Time only begins to run from that point.

Mr Hollander submits that this was a “no transaction” case with respect to the vetting breaches. If those breaches had not occurred, NI would not have entered into contracts of ATE insurance at all. The measure of damage in such a case is the loss which NI suffered as a result of entering into the ATE insurance and not on “a loss of the bargain” basis. Mr Hollander therefore submits that Nykredit is dispositive against the contention that immediate loss was suffered on inception applying the loss of bargain measure.

Miss Carr submits that there is no need for a striking of the balance in this case. She points out that the courts will more readily conclude that damage was suffered when a transaction was entered into than at some subsequent date. She also relies on Pegasus at [107] where Lewison J concluded that Nykredit did not detract from the professional negligence authorities. The client suffers damage at the time of the transaction even if (she adds) the property or rights which he acquires are not worth less than what he paid for them.

Miss Carr submits that it is moreover often difficult to identify a solid distinction between a “no transaction” case and a “successful transaction” case. She submits that there is a grey area where the transaction that would have been entered into was vastly different from the transaction actually entered into. In the present case, she contends that NI has demonstrated in its pleading the difficulty of drawing this distinction. There is no reason why the commencement of the limitation period should start at a different date depending on whether the clause could be removed and the transaction entered into or on whether the transaction should have been avoided. She points out that Lord Hoffmann in South Australia Asset Management Corporation v York Montagu Ltd [1997] 1 AC 191 at 218 held that the distinction between “no transaction” cases and “successful transaction” cases was not based on any principle and should be abandoned.

Miss Carr further submits that the no transaction/successful transaction distinction illustrates the difficulty of categorising the cases. She submits that dating the commencement of the limitation period involves a fact-specific enquiry rather than a categorisation. She points out that in Sephton Lord Hoffmann did not suggest at [21] that there was a difference just because the case was a “no transaction” case.

Mr Hollander submits that his analysis with respect to vetting claims can be applied also to the conduct breaches consisting of a failure to notify. In each case where the panel solicitor should have notified NI that the prospects of success had fallen below 50% NI lost the opportunity to withdraw cover at that stage. This might lead to a situation in which NI incurred greater loss than it would otherwise have done so. In the type of conduct case where the panel solicitor had failed to progress the case as it should have done, for example by failing to obtain a medical report, the prospects of success of the claim were thereby reduced. But again, the contingency remains, even though the risk of the contingency occurring is greater than should be the case otherwise.

It follows from my earlier analysis that damage occurs at the time when the conduct breach takes place if and in so far as NI is thereby exposed to larger liabilities than it would have been but for the conduct breaches consisting of a failure to notify.

The panel solicitors have filed a respondents’ notice in which they advance an alternative argument put to the judge but which the judge held that it was unnecessary for him to rule (see [108] to [111] of his judgment). The panel solicitors contend in effect that NI suffered damage for limitation purposes as soon as the ATE business was written because the liabilities were such as to affect the value of its business or goodwill on the open market. In the light of the conclusions reached thus far, it is unnecessary to decide the questions arising on the respondents’ notice.

I have had the privilege of reading in draft the judgments of Longmore LJ and Lloyd LJ. I agree with the judgment of Longmore LJ. I respectfully do not agree with the conclusions drawn by Lloyd LJ from the authorities, and in particular his central conclusion, on the basis of those authorities, that NI suffered damage as a result of entry into a policy of ATE insurance solely because it had assumed a contingent liability. At the end of the day the major issue that divides us is whether the contingent liability of NI under the policies that it issued stood alone, or whether NI incurred other measurable loss at the time those policies were issued. My view is that there was then other measurable loss: if, for instance, immediately on signature NI had sought to agree to transfer to a purchaser the benefit of its rights remaining after discharge of its obligations under any one of the policies in respect of which there had been a vetting breach, it would have received less for it than it would have done if the vetting breach had not occurred. In respectful disagreement with Lloyd LJ, both Sephton and Wardley are distinguishable from this case. In Sephton, there was no transaction concluded between the clients claiming compensation and the Law Society when their rights against the Law Society (to claim compensation) arose. In Wardley, summarised in [14] above, by contrast, there was a transaction but there was no benefit moving from the beneficiary of the indemnity to the State of Western Australia in return for its execution of the indemnity. The beneficiary agreed to make, and did make, advances to a particular customer (R). Therefore, the State of Western Australia suffered no damage by entering into the indemnity on the defendant’s misrepresentation, other than that inherent in the issue of the indemnity, until the indemnity was called. The claim against W, the defendant, in misrepresentation did not form part of the package of rights obtained under the transaction of indemnity. In this case, however, NI acquired the right to receive and retain the premium and other rights. The maturing of the contingent liabilities into actual liabilities was, to borrow a phrase used by the High Court of Australia in Jobbins v Capel Court Corporation (1989) 25 FCR 226 at 231, and cited in Wardley at 528, “a reaping of the tares sown with the crop”. Tares make a crop less valuable. So, too, the vetting breaches resulted in a diminution in the value of the package of rights that NI obtained under the policy.

For the reasons given above I would dismiss the appeal and make no order on the respondents’ notice.

Law Society v Sephton has made it clear that “damage” for the purpose of the accrual of a cause of action in the tort of negligence will not be constituted by a mere contingent liability. Another way of making the same point is to say that a contingent liability does not “of itself” constitute “damage”. There must be something more.

In cases of negligent advice, the person relying on the advice will usually have entered a transaction of some kind which has turned out to be “flawed” in some way. One might expect that a “flawed” transaction of this kind would be something more than a contingent liability even if the liability, to which that transaction might give rise, depends on some future event. Sometimes the position will be that, if the claimant had been given the right advice, he would have entered a different and better transaction, in which case his damages are the difference between the value of the transaction into which he ought to have entered and the value of the transaction into which he did enter. That is akin to the contractual measure of damages (“loss of the bargain”). Sometimes if he had been given the right advice he would not have entered into a transaction at all, in which case his damages are the difference between the position in which he would have been if he had never made the transaction and the position in which he finds himself at the time of the institution of proceedings.

Mr Hollander QC for the claimant insurers submits that there is nothing more than a contingent liability on the insurers on the facts of this case because, if the solicitors had performed their duty instead of (as must be assumed) being in breach of it, no policies of ATE insurance would have been written in favour of the litigants whose claims ultimately failed and insurers would, therefore, have never had to pay out on any policies underwritten by them. In other words he submits that this case is in the latter (non-transaction) category rather than the former (better transaction) category. For that reason insurers have no more than a contingent liability and did not suffer “damage” when the policies incepted but only when the underlying claim “failed” in the sense set out in paragraph 17 of the judgment below.

It is not, however, possible to say that the entering into of a flawed transaction constitutes damage when it is in one category of case but not when it is in another. The fact that the flawed transaction has been entered into will usually be damage from the claimant’s point of view. The fact that the recipient of the advice might have hoped for a better transaction or might have hoped to avoid any transaction makes no difference to the fact that he has entered into a flawed transaction which he would not have done if he had been competently advised. If such a flawed transaction has come into existence that will, in my view, usually be the damage which the recipient of the advice has suffered and that is more than the existence of a mere contingent liability.

This is supported by Nykredit v Edward Erdman [1997] 1 WLR 1627 in which the defendants had negligently over-valued a property on which the claimant lenders in March 1990 advanced money on mortgage. If a non-negligent valuation had been given, the claimants would not have lent money on the security of the property. It was, therefore, a non-transaction category case. The personal covenant of the borrower was, on the facts of the case, a worthless covenant and there was an immediate default. But no sale took place until February 1993. The House of Lords decided, first, that the claimants’ damages were only the difference between the defendant’s valuation and the true valuation rather than the difference between the defendant’s valuation and the sum for which the property was ultimately sold. The House therefore awarded the claimants the sum of £1.4 million plus interest. The question then arose of the date from which interest should be calculated; that meant that the House had to determine when the claimants’ cause of action accrued. The defendants contended that the cause of action arose at the date of sale in 1993 when the amount of the claimants’ loss crystallised, but the House held that the claimants’ cause of action in tort (1630A) arose at the date of the loan transaction in March 1990. Their full loss of £1.4 million had been suffered by December 1990 (1635A) and interest should, therefore, be calculated from that date. But the cause of action arose when the claimants “first sustains damage” (1630C) or sustains “measurable relevant loss” (1631D) and, in a case where the borrower’s covenant had no value, that was when the transaction was entered into. Lord Nicholls of Birkenhead recognised that the position might be different if the borrower’s covenant performed satisfactorily (1633G).

In Law Society v Sephton & Co the Law Society did not enter into any transaction as a result of Mr Mascard of Sephton & Co negligently certifying Mr Payne’s accounts. Their liability arose because they were required by section 36 of the Solicitors Act 1974 to maintain and administer a compensation fund for the purpose of relieving clients’ losses caused by a solicitor’s dishonesty. The conditions under which such losses would be met by the Law Society were set out in the Solicitors Compensation Fund Rules 1995. All that could be said when Mr Payne misappropriated his clients’ funds was that it was possible that the Law Society would become liable to compensate that client

“contingent upon the misappropriation not being otherwise made good and a claim in proper form being made” (550G)

Not unnaturally Lord Hoffmann described that as a “purely contingent obligation” (551F and 552F).

Lord Hoffmann distinguished the case in which the claimant had entered into a transaction on the basis of negligent advice as irrelevant to cases of a pure contingent liability and pointed out (551A) that Nykredit was not a case of contingent liability at all. It was a case of the lender suffering a loss as a result of making a loan which he would never have made if he had been given prudent advice. The only question was whether there was an immediate loss or a contingent loss depending on how good (and for how long) the borrower’s covenant was a good covenant. Lord Hoffmann held (551E, para 20) that Nykredit decided

“that in a transaction in which there are benefits (covenant for repayment and security) as well as burdens (payment of the loan) and the measure of damages is the extent to which the lender is worse off than he would have been if he had not entered the transaction, the lender suffers loss and damage only when it is possible to say that he is on balance worse off. It does not discuss the case of a purely contingent liability.”

Lord Hoffmann goes on (in para 21) to observe that where a transaction has been entered into which has both benefits and burdens, the answer to the question whether damage has been suffered may be different according to whether the liability is for the consequences of the claimant entering into the transaction. If the claimant did not get what he should have got it will be easy to infer that he has suffered some immediate damage. If the claimant’s damage is, however, the difference between his position having entered into the transaction and his position if he had not entered into the transaction at all (the Nykredit case),

“the answer may be more difficult”

And “some evidence may be necessary to show when he was actually in a worse position”.

In the light of these two authorities the judge was, in my view, right to hold (para 37) that there was no authority for Mr Hollander’s primary submission that, where a claimant would not have entered into the relevant transaction if he had been given the correct advice, the claimant has not, as a matter of law, suffered damage. It may be obvious that he has suffered damage as it was in the Nykredit case itself; or it may be necessary (in a benefit and burdens case) to adduce evidence as to the time when the claimant is worse off.

I do not regard para 48 of Lord Walker of Gestingthorpe’s speech as being in any way inconsistent with the above analysis. In para 45 he summarises cases where, as a result of negligent professional advice, the claimant has ended up with a package of rights less valuable than he was entitled to expect and then says that in those cases

“the claimant has … suffered a diminution … in the value of an existing asset of his, or has been disappointed (as against what he was entitled to expect) in an asset which he acquires whether it is a house, a business arrangement, an insurance policy or a claim for damages.”

He then distinguishes such cases from a case in which there is

“the imposition on a claimant of a purely personal and wholly contingent liability”

where no asset has been acquired. This case falls within the first category of cases because the claimants have acquired an insurance policy which will in due course generate claims which will exceed the premium. It is therefore a flawed transaction case not an imposition of a purely personal and wholly contingent liability.

Nor do I think that Lord Mance’s analysis is inconsistent with what I have said unless it is (perhaps) to be said that the word “and” in the third sentence of para. 77 is intended to import as a requirement of measurable loss that there is both a change of position and a diminution in value of a particular asset. Like my Lady I doubt if he does mean that because he expressly says he agrees with Lord Hoffmann and Lord Walker.

Wardley (1992) 175 CLR 514 is in many ways closer to the present case because it can be said that the indemnity granted by the State of Western Australia is analogous to the insurance contracts in the present case. But the analogy cannot be pressed too far. The indemnity cover was similar to liability insurance where it is axiomatic that a cause of action against the insurer does not arise until the claim is ascertained by agreement judgment or award, see Post Office v Norwich Union [1967] 2 QB 363. In the present case the insurers agree to pay to the claimant the amount he has borrowed from the Funder in respect of the insurance premium and any disbursements, although it is fair to say that the insurer is also, of course, liable to pay the amount of the defendant's costs and disbursements for which the claimant is liable. Insurers' liability can arise if and when they withdraw their indemnity just as much as it can arise once the claim has failed (see paras 9, 11 and 17 of the Statement of Assumed Facts).

The most that can be said in the present case is that the loss suffered by the claimant insurers is contingent upon the claim, which is (ex hypothesi) likely to fail, actually failing. But that does not make the case a case of a “mere contingent liability” because the claimants have entered into a flawed transaction which they ought not to have entered into. To my mind that is the damage which the claimants have suffered and that occurred at the time of the inception of the policies. It is true that the insurers are not immediately worse off as a result of entering into the ATE policies because they receive the premiums up front and it will be a short time before they are “on balance worse off” to use Lord Hoffmann’s phrase in para. 20 of Sephton, but that will be well before the underlying claim has “failed” which is the time argued for by Mr Hollander. In this context it is important to appreciate that the alleged negligence is that the defendant solicitors were negligent in selecting claims which had less than a 51% chance of success. The loss to insurers was therefore in the natural order of things bound to occur. I therefore agree with Arden LJ (para. 61) that there was measurable relevant loss on the inception of the policies in that any valuation of the policies at that time would have to take into account the assumed fact that there had been no proper vetting. I also agree with her in relation to what have been called the “conduct breaches”.

I have read a draft of what will be Lloyd LJ’s dissenting judgment. But I do not, with respect, think that Sephton compels us to accept Mr Hollander’s argument. Lloyd LJ says more than once that NIG has been put in a seriously worse commercial position than it ought to have been; indeed that is the whole essence of the claimants’ case. In those circumstances it seems to me to be distinctly uncommercial to say that the insurers have suffered no loss and I would be troubled by such a conclusion. It is also worth remembering Lord Nicholls’ observations in Nykredit at 1633D:

“…. within the bounds of sense and reasonableness the policy of the law should be to advance, rather than retard, the accrual of a cause of action. This is especially so if the law provides parallel causes of action in contract and in tort in respect of the same conduct. The disparity between the time when these parallel causes of action should be smaller, rather than greater.”

In a case where, on any view, the natural cause of action is for breach of contract, the courts should not favour a much later date of accrual for the co-existing action in tort unless they are compelled to do so. This, in my view, is not that case. In agreement with Arden LJ I would dismiss this appeal.

The question in this appeal is what it was that set the limitation period running for a claim in tort by NIG, assignor of the claimant, against the defendant solicitors for negligence in respect of entry into, or the conduct of claims under, after-the-event (ATE) legal expenses insurance policies relating to personal injury or certain other types of claim. The question can be addressed most clearly as regards claims that the solicitors were in breach of duty in committing NIG to policies, by their negligence in vetting claims before a policy was issued. In contract the cause of action accrued on entry into the policy. In tort it accrued when NIG suffered loss as a result. In the nature of an insurance policy, the insurer’s liability was contingent. On behalf of the appellant Mr Hollander Q.C. argued that loss is not suffered in such a case unless and until an event occurs upon which a claim arises under the policy. For the various firms of solicitors who are defendants in this litigation, Miss Sue Carr Q.C. argued that the judge was right to hold that loss was suffered on entry into the policy. The aggregate claims against the solicitors amount to a very large sum of money. If the judge is right, a substantial part of this amount cannot be recovered because these proceedings were brought too late.

The documentation relating to the ATE policies and the circumstances in which they came to be entered into is quite elaborate. I will first describe the overall arrangements, and then say something about the detail of the policies.

Because the policies were ATE, each of them related to a specific claim against a given third party, unlike a before-the-event (BTE) policy which would provide cover against legal expenses and liabilities incurred in relation to any claim or a class of claims, whether as claimant or as defendant. ATE policies developed as a result of the Access to Justice Act 1999, section 29, under which a party to proceedings can, for the first time, recover by way of costs the cost to him of insuring a costs liability in the proceedings. This came into effect on 1 April 2000.

In effect, entry into ATE policies on behalf of NIG was delegated to Panel solicitors. Under the agreement between CLE and the particular firm of solicitors, the firm was an appointed representative of CLE and thus of NIG. The firm, as agent for CLE and indirectly for NIG, had authority to receive proposals or applications and to accept business. Cases where the prospects of success were assessed as being less than 51% were not acceptable, and the reasonably anticipated minimum value of the claim had to be not less than £1,000 (later increased to £1,500). The assessment was carried out by the firm. If in the course of the case the prospects fell to below 50%, CLE was entitled to withdraw from the case, so the solicitors had to have a procedure for monitoring prospects and for reporting relevant changes. In addition the firm owed a duty of care to CLE and NIG (as well as to the Scheme Claimant, the nominal client) as regards the conduct of the case.

The ATE policies were issued in a number of different forms. Under a typical policy, the basic provision as regards cover was as follows:

“Consideration and Insurance

The Insured having made a proposal and declaration and having entered into a Conditional Fee Agreement with the Appointed Solicitor and having paid or agreed to pay the Premium, the Company will, subject to the terms, conditions and exclusions of the Policy (compliance with such conditions being a condition precedent to the liability of the Company) indemnify the Insured up to the Limits of Indemnity in respect of Opponent’s Legal Costs under section 1 below and/or Deficiency of Damages under section 2 below incurred by the Insured exclusively in connection with the Proceedings, provided that the Proceedings are conducted exclusively within the Territorial Limits.”

The Insured was the relevant Scheme Claimant. This provision was amplified, as its text suggests, by sections 1 and 2 dealing with different contingencies. Under section 1, the Company was to indemnify the insured for Opponent’s Legal Costs (defined as one would expect) payable by the Insured to the Opponent pursuant to a court order or a settlement approved in writing by the Company, any costs payable the other way being set off for this purpose. Section 2 covered two different cases. If the Insured was not Successful (as defined), the Company was to indemnify the Insured for the Insured’s Disbursements, the Premium and interest payable on the loan of those amounts. If the Insured was Successful, or benefitted from a settlement approved by the Company, but the amount payable to the Insured did not cover the whole of the Insured’s Disbursements, the Premium and the loan interest, the Company was to indemnify the Insured for the balance. For this purpose “Successful” was defined as meaning that an order is made for the payment of damages by the Opponent to the Insured. Insured’s Disbursements meant reasonable and proper payments made or authorised by the Appointed Solicitor in connection with the proceedings (including those incurred before commencement).

The conditions of the policy included one relating to prospects of success, and another as regards control of the claim proceedings. Section 4.1 was as follows:

“Prospects of Success

The Company has provided cover under this Policy on the basis that it offers reasonable prospects for the recovery of damages in the Proceedings. The Company or the Coverholder on its behalf may discontinue cover if during the course of the Proceedings it considers that such prospects no longer exist. If the Company or the Coverholder discontinues cover it shall notify the Insured in writing and shall inform the Insured of its reasons which, subject to the provisions of Section 5 (Disagreement) below, shall be final. If cover is discontinued, the Company shall be liable only for Opponent’s Legal Costs up the date of notification of the discontinuance of cover and Insured’s Disbursements up to that date. Notwithstanding notification, the Company shall be entitled to retain the whole of the Premium.”

I do not need to set out the details of the condition dealing with control of the Proceedings.

Clause 6 dealt with payment. No payment was to be made by the Company under the Policy before the conclusion of the Proceedings, except if an order were made against the Insured during the course of the Proceedings requiring payment of any of the Opponent’s Legal Costs prior to the conclusion of the Proceedings. In that event, the Insured was to be indemnified for that amount as soon as the Insured was required to make the payment.

Looking at the Policy as a whole, therefore, whether NIG would become liable to make any (and if so what) payment under the policy would depend on the outcome of the Proceedings. If the claim were successful, whether by order or by agreement, and damages and costs were recovered, with no liability for any part of the Opponent’s Legal Costs to be set off, then NIG would be under no liability to indemnify the Scheme Claimant for anything. If it were not successful, or not sufficiently so to cover the Scheme Claimant’s liability for his own disbursements and the premium and interest, then NIG would be liable to make a payment by way of indemnity, but, normally, not only would this not arise until the conclusion of the proceedings (because until then the outcome would not be known) but in any event no payment would fall to be made until then. Only in the unusual case of an order to pay Opponent’s Legal Costs during the proceedings would NIG become liable to pay any money out before the conclusion of the proceedings. Plainly, on issue of the policy, NIG came under a liability, but it was a contingent liability at that stage, which might or might not become an actual liability, according to later events.

Cover could be withdrawn on the grounds that the prospects of success of the claim were no longer adequate. In that event, presumably, the Claim would have to be discontinued for lack of funding. NIG would then be liable under the policy for Opponent’s Legal Costs up to the date of notification of the withdrawal. Because the Claim would not have been Successful, NIG would also be liable to indemnify the Insured in respect of the Premium, Disbursements and interest on the Funder’s loan. The Insured was thus protected from liability for the Premium, the Disbursements and interest on the loan because they would be paid either by the Defendant, if the Claim was Successful, or by NIG if the Claim was not Successful or if the quantum on Success was insufficient to cover the liability. In addition, as already noted, NIG agreed to indemnify the Funder if NIG avoided the policy.

The present proceedings were commenced on 17 June 2008, so the cut-off date is 17 June 2002. So far as vetting breaches are concerned, the simple case is where the ATE policy had been issued before 17 June 2002. Mr Justice Flaux held that in all such cases a claim in tort based on alleged negligence in vetting was barred by the Limitation Act 1980. As regards conduct breaches, he held that the claim was barred if, as a consequence of the breach of duty, there had, before 17 June 2002, been a material diminution in the prospects of success of the Scheme Claim.

On behalf of AXA, Mr Hollander argued that damage was not suffered by NIG until the particular Scheme Claim failed, that failure giving rise for the first time to a claim under the policy (leaving aside as exceptional and possibly only hypothetical any case in which an indemnity had to be provided early because of an order for payment of Opponent’s Legal Costs during the Claim Proceedings). He submitted that the judge’s reasoning is inconsistent with the decision of the House of Lords in Law Society v. Sephton [2006] UKHL 22, [2006] 2 AC 543 and with the earlier decision of the House of Lords in Nykredit Mortgage Bank v Edward Erdman Group (No. 2) [1997] 1 W.L.R. 1627 (to which I will refer as Sephton and Nykredit). Much of the argument before us turned on the effect of these and other cases in this area. I therefore turn to these next.

In the early 1980’s the answer to the present question might have been thought to lie in the decision of this court in Forster v Outred & Co [1982] 1 W.L.R. 86. Mrs Forster had mortgaged her freehold property as security for a loan made to her son. He later went bankrupt, and the lender made a demand under the mortgage, following which she paid the sums secured. She then issued proceedings against the solicitors who had acted for her at the time of the mortgage, claiming damages for negligent advice, in contract and in tort, at and before the time the mortgage was executed. That action went into suspended animation, and was eventually struck out for want of prosecution. In the meantime, her solicitors issued a second writ claiming the same relief, but this was more than 6 years after the date of the mortgage, though within 6 years of the claim under the mortgage. The Court of Appeal held that the second writ was barred by limitation because, although there might never have been a claim, she had suffered actual damage by executing the mortgage by which her property was encumbered with a liability which might mature into a financial loss, so that her claim in tort accrued when the mortgage was executed. Stephenson LJ accepted a proposition put forward by Mr Murray Stuart-Smith Q.C., which he summarised at page 93E:

“Mr Stuart-Smith contends, on behalf of the defendants, that when she signed the mortgage deed she suffered actual damage. By entering into a burdensome bond or contract or mortgage she sustained immediate economic loss; her valuable freehold became encumbered with a charge and its value to her was diminished because she had merely the equity of redemption, varying in value at the whim of her son’s creditors; she could not sell the land without discharging the mortgage; she could not prevent her son from borrowing on the security of her mortgage to the extent of the full value of the land; she could have sued the defendants in February 1973 for an indemnity or for damages on the basis of the diminished value of the land or the amount of the outstanding debt to the mortgagor.”

In the course of his review of the issues, Stephenson LJ came to the question of actual damage, and said this at page 94C:

“What is meant by actual damage? Mr Stuart-Smith says that it is any detriment, liability or loss capable of assessment in money terms and it includes liabilities which may arise on a contingency, particularly a contingency over which the plaintiff has no control; things like loss of earning capacity, loss of a chance or bargain, loss of profit, losses incurred from onerous provisions or covenants in leases. They are all illustrations of a kind of loss which is meant by ‘actual’ damage.”

At the end of his judgment he said this at page 98D:

“Although there is no more direct authority than those cases among those which have been cited to us, I would accept Mr Stuart-Smith’s statement of the law and would conclude that, on the facts of this case, the plaintiff has suffered actual damage through the negligence of her solicitors by entering into the mortgage deed, the effect of which has been to encumber her interest in her freehold estate with this legal charge and subject her to a liability which may, according to matters completely outside her control, mature into financial loss, as indeed it did. It seems to me that the plaintiff did suffer actual damage in those ways, and subject to that liability and with that encumbrance on the mortgage property was then entitled to claim damages (not, I would think, an indemnity and probably not a declaration) for the alleged negligence of the solicitor which she alleges caused her that damage. In those circumstances her cause of action was complete on 8 February 1973 and the writ which she issued on 25 March 1980 was issued too late to come within the six years’ period of limitation.”

In turn, Dunn LJ, agreeing, said at page 100A:

“In this case, as soon as she executed the mortgage the plaintiff not only became liable under its express terms but also, and more importantly, the value of the equity of redemption of her property was reduced. Before she executed the mortgage deed she owned the property free from encumbrance; thereafter she became the owner of property subject to a mortgage. That, in my view, was a quantifiable loss and as from that date her cause of action against her solicitor was complete, because at that date she had suffered damage. The actual quantum of damages would, of course, depend on events between that date and the date when the damages had finally to be assessed, but the cause of action was complete when she executed the mortgage, without proof of special damage.”

Sir David Cairns agreed with both judgments.

The decision in Forster v Outred & Co, and the terms of the judgments in the case, were considered by the House of Lords in Nykredit. The relevant issue in that case was the date from which interest was payable, following earlier decisions of the House of Lords in a group of cases generally referred to as South Australia Asset Management Corporation v York Montague Ltd (or SAAMCO) [1997] AC 191. There the cause of action arose because the defendant valuers were required by the plaintiff lender to value property on the security of which they were considering making a loan. They overvalued the property negligently at £3.5 million, it being agreed that the true value at that date was £2.1 million. The lender advanced £2.45 million on the security of the mortgage in March 1990; the borrower defaulted at once and by December 1990, taking into account the continuing cost to the lender of providing the money and the diminishing value of the property as the market deteriorated, the lender had sustained its full allowable loss of the difference between the amount of the negligent valuation and the true value at that time, namely £1.4 million (see [1997] 1 W.L.R. 1627 at 1629H). The House of Lords held that the cause of action had arisen when a relevant and measurable loss had first been revealed and that because the borrower had defaulted at once, and the property had at all times been worth less than the amount lent, the first moment of actual loss was about the time of the loan transaction itself; it was also held that interest should be awarded on the full £1.4 million from the date in December 1990 by which the full loss had been sustained.

This was not a case of a contingent liability. However, Lord Nicholls considered Forster v Outred & Co and other cases as to when a cause of action first accrues. At page 1630D he quoted the passage from Stephenson LJ’s judgment in which he set out Mr Stuart-Smith’s submission about actual damage (see paragraph [97] above), said that Stephenson LJ accepted it, and said that he agreed, subject to the loss being relevant, that is to say within the measure of damage applicable to the wrong in question.

One other case which was mentioned in Nykredit was a decision of the High Court of Australia, Wardley Australia Ltd v State of Western Australia (1992) 175 CLR 514. This arose from a claim on a statutory provision under which a person who suffers loss or damage by conduct of another person done in contravention of various provisions might recover the amount of the loss or damage by action against that other person, but subject to a time limit of three years from the date when the cause of action accrued. The headnote summarises the decision as being that where, as a result of misleading or deceptive conduct, a person grants an indemnity under which he is obliged to make payment when the loss of the party indemnified is ascertained and quantified, the indemnifier suffers no loss until the contingency is fulfilled and time does not start to run until then.

The State of Western Australia had granted an indemnity in favour of the National Australia Bank against a facility granted by that bank to Rothwells Ltd. Wardleys were alleged to have engaged in misleading or deceptive conduct in connection with the execution of the indemnity. Rothwells, a merchant bank, had suffered severe liquidity problems in the stock market collapse of 1987, following substantial withdrawals of deposits. Rescue arrangements included the State giving the indemnity in question. Wardleys were said to have made material misrepresentations to the State as to the financial and asset position of Rothwells. Rothwells drew down substantially on the facility against which the indemnity was given, and shortly thereafter it repaid the drawings, but this repayment was challenged by the liquidators when Rothwells went into liquidation as a result of a petition presented soon afterwards. The State brought its claim within the three year period, but later, more than three years after the date of the indemnity, it sought to add a new allegation of a different misrepresentation made on a separate occasion. The judge held that this was outside the time limit, the date of accrual being the execution of the indemnity. The Full Court of the Federal Court allowed the State’s appeal, and the High Court dismissed Wardley’s appeal against that. Several judgments were given. In the leading judgment, of Mason CJ, Dawson J, Gaudron J and McHugh J, this point was made (at pages 524-5):

“The liability was, therefore, in conformity with the opinion of the Full Court, contingent and executory. The likelihood, perhaps the virtual certainty, that there would be a loss, in the light of Rothwell’s actual financial position as it stood when the indemnity was executed, did not transform the liability into an actual or present liability at that time.”

At page 527 they went on to recognise that when a plaintiff is induced by misrepresentation to enter into an agreement which is or proves to be disadvantageous, he suffers detriment in a general sense on entry into the agreement, but that sort of detriment has not been equated with the legal concept of loss or damage. At page 529 they explained Forster v Outred on the basis that the execution of the mortgage had an immediate effect on the value of the plaintiff’s property. Commenting on the English cases at page 531 they said:

“It has been contended that the principle underlying the English decisions extends to the point that a plaintiff suffers loss on entry into an agreement notwithstanding that the loss to which the plaintiff is subjected by the agreement is a loss upon a contingency. For our part we doubt that the decisions travel so far. Rather, it seems to us, the decisions in cases which involve contingent loss were decisions which turned on the plaintiff sustaining measurable loss at an earlier time, quite apart from the contingent loss which threatened at a later date.”

Consistently with that, at page 532, they said:

“If, contrary to the view which we have just expressed, the English decisions properly understood support the proposition that where, as a result of the defendant’s negligent misrepresentation, the plaintiff enters into a contract which exposes him to a contingent loss or liability, the plaintiff first suffers loss or damage on entry into the contract, we do not agree with them. In our opinion, in such a case, the plaintiff sustains no actual damage until the contingency is fulfilled and the loss becomes actual; until that happens the loss is prospective and may never be incurred.”

Mr Hollander relied strongly on that reasoning, and on the policy grounds given at the end of the judgment to support this conclusion, at page 533:

“The conclusion which we have reached is reinforced by the general considerations to which we referred earlier. It is unjust and unreasonable to expect the plaintiff to commence proceedings before the contingency is fulfilled. If an action is commenced before that date, it will fail if the events so transpire that it becomes clear that no loss is, or will be incurred. Moreover the plaintiff will run the risk that damages will be assessed on a contingency basis, in which event the compensation awarded may not fully compensate the plaintiff for the loss ultimately suffered. These practical consequences which would follow from an adoption of the view for which the appellants contend outweigh the strength of the argument that the principle applicable to the cases in which the plaintiff acquired property (or a chose in action) should be extended to cases where an agreement subjects the plaintiff to contingent loss. In such cases it is fair and sensible to say that the plaintiff does not incur loss until the contingency is fulfilled.”

Brennan J, Deane J and Toohey J each gave judgments agreeing in the result.

I turn at last to Law Society v Sephton in which the relevant liabilities were contingent and, unusually, did not arise from a transaction such as the indemnity in Wardley or the mortgage in Forster v Outred & Co. The Law Society sued in respect of its liability under the Solicitors’ Compensation Fund, to make payments by way of compensation to clients who had lost money by reason of the dishonesty of a solicitor, claiming that the accountants who had reported on the accounts of the dishonest solicitor should compensate the fund by reason of their negligence in reporting on the accounts. Such liability was held to exist in Law Society v KPMG Peat Marwick [2000] 1 W.L.R. 1921. In the Sephton litigation the question was when the cause of action arose. The Society intervened in the practice of the relevant solicitor in 1996, and made its first compensation payment out of the fund in October 1996. The accountant’s report related to each of the years from 1988 to 1995. The claim form was issued in 2002, just within 6 years from discovery of the deficiency by the Law Society’s investigating accountant. The House of Lords held that the cause of action accrued when a claim was made for compensation from the fund, so that the proceedings were not barred by limitation. It is necessary to examine in some detail what was said by Lord Hoffmann, Lord Walker of Gestingthorpe and Lord Mance.

Lord Hoffmann said that he found the answer to the case in Wardley, and he approved Forster v Outred but subject to an important qualification. At paragraph 14 he said this (the second quotation to which he refers is of the passage quoted at paragraph [97] above):

“Stephenson LJ said (at p 98) that he accepted Mr Stuart-Smith’s statement of the law. The ambiguity in these passages (in an unreserved judgment in an interlocutory appeal) arises from the inclusion of the words “it includes liabilities which may arise on a contingency” in the second quotation. As appears from the first passage, the thrust of Mr Stuart-Smith’s argument was that the mortgage, although the liability which it secured was contingent, had the immediate effect of depressing the value of Mrs Forster’s farm. But the reference to contingent liabilities in the second passage could give the impression that merely incurring a possible future liability (for example, by giving a guarantee or indemnity unsecured upon any property) counted as immediate damage.”

Having referred to the approval of Stephenson LJ’s comment in Nykredit, but observing that this did not concern contingent liabilities, Lord Hoffmann went on at paragraph 21 to say this (Arden LJ sets out the whole paragraph, and paragraph 22, in her paragraph [16] above):

“Next, there are a number of cases in the Court of Appeal which involve transactions, with both benefits and burdens, into which the plaintiff entered as a result of the negligence or breach of contract of the defendant. None of these cases concerned purely contingent obligations. It is only necessary to observe that in such bilateral transactions the answer to the question of whether damage has been suffered may be different according to whether the liability is for the consequences of the defendant not performing his duty or (as is usual in claims for misrepresentation) the consequences, or some of the consequences, of the plaintiff entering into the transaction.”

He reviewed other cases and observed at paragraph 22 that they had no relevance to a case in which a purely contingent obligation has been incurred. At paragraph 25 he said that, apart from one case which he thought wrongly decided, there seemed to be no authority “inconsistent with the opinion of the High Court of Australia in Wardley that incurring a contingent liability is not, as such, actionable damage”.

He concluded his speech by referring to considerations of policy, but saying that the issue had to be decided on principle. I will set out all of his paragraphs 28 to 31.

“28.

I respectfully think that the reasons of policy advanced by the High Court in Wardley 175 CLR 514 are somewhat overstated. It is often the case that a plaintiff, who has plainly suffered some damage, is obliged to commence proceedings before the full effects of his injury can be known. This frequently happens in actions for personal injury. On the other hand, I do not agree with Hobhouse LJ that Wardley and section 14A of the 1980 Act are different solutions to the same problem. They are solutions to different problems. Allowing a plaintiff three years from the date on which he knows that a cause of action has arisen does not help if a mere contingent liability is treated as damage and the plaintiff, at the end of the three years, still does not know whether the contingency will eventuate or not. On the other hand, Wardley is no answer to a case like Cartledge v E Jopling & Sons Ltd [1963] AC 758, in which the plaintiff had on any view suffered damage but did not know it.

29.

It also seems to me irrelevant that a prudent accountant, drawing up the accounts of the Compensation Fund to give a true and fair view of its assets and liabilities, would have included provision for contingent liabilities. As Lord Radcliffe pointed out in Southern Railway of Peru Ltd v Owen [1957] AC 334, 357, the principles upon which such provisions are made does not depend upon “any exact analysis of the legal form of the relevant obligation” but upon estimates of what in practice is likely to happen. A cause of action, however, connotes a legal obligation and its existence must be determined by rules of law.

30.

In my opinion, therefore, the question must be decided on principle. A contingent liability is not as such damage until the contingency occurs. The existence of a contingent liability may depress the value of other property, as in Forster v Outred & Co [1982] 1 WLR 86, or it may mean that a party to a bilateral transaction has received less than he should have done, or is worse off than if he had not entered into the transaction (according to which is the appropriate measure of damages in the circumstances). But, standing alone as in this case, the contingency is not damage.

31.

The majority of the Court of Appeal appear to have decided the case on the basis that the Law Society did not enter into any transaction giving rise to the contingent liability. It did nothing and the contingent liability was created by the misappropriations and the previous existence of the Compensation Fund and the rules which governed its administration. No doubt in most cases in which a party incurs a contingent liability as a result of entering into a transaction, that liability will result in damage for the reasons already discussed in relation to bilateral transactions. But I would prefer to put my decision on the simple basis that the possibility of an obligation to pay money in the future is not in itself damage.”

Lord Scott agreed with the other speeches, and agreed in terms with what Lord Hoffmann said about Wardley, and that this analysis provided the answer to the appeal. Lord Rodger agreed with the other speeches without adding any particular comment.

Lord Walker reviewed the cases and identified a category of transaction cases, where the negligence of the professional has resulted in a diminution in the value of an asset held or acquired, that diminution being sometimes immediately quantifiable but in other common cases not yet quantifiable. At paragraph 48, having referred to this category, he said:

“Your Lordships have not, I think, been shown any case in which the imposition on a claimant of a purely personal and wholly contingent liability, unsecured by a charge on any of the claimant’s assets, has been treated as actual loss. That would have been the position if the claimant in Forster [1982] 1 WLR 86 had given a personal covenant guaranteeing her son’s debts (which she seems not to have done — she paid them simply to prevent enforcement of the security on her farm) and if she had not given any security over any of her own assets.”

In his paragraph 49, he agreed with the passage in the leading judgment in Wardley which I have quoted at paragraph [105] above. On that basis he disagreed with the arguments for the accountant, saying at paragraph 52:

“I would therefore reject Sephton’s submission that it was enough that the Law Society was, every time that misappropriations were made after the issue of defective accountant’s certificates, at risk of having to meet claims from clients from whom Mr Payne misappropriated funds — even if that risk (of a future eventuality) is beguilingly expressed as “exposure to claims” (suggesting a present or current condition). It was in a sense a detriment, but it was not a detriment of the sort described in Forster v Outred & Co [1982] 1 WLR 86, 94 as understood and developed in the later authorities.”

He also agreed with the reasons given by Lord Hoffmann and Lord Mance.

The most important passage in the speech of Lord Mance is at paragraphs 75 to 79:

“75.

Wardley was considered in Knapp [1998] PNLR 172. Hobhouse LJ viewed it at p.178A as adopting a different approach to the English approach. Buxton LJ at p.192A-D endorsed the rejection in Wardley of any proposition that

“the plaintiff necessarily suffers loss on entry to an agreement notwithstanding that the loss to which [he] is subjected by the agreement is loss upon a contingency: what is required is actual loss on entry, quite apart from the contingent loss threatened at a later date.”

The rejection in Wardley and by Buxton LJ of any such proposition is consistent with the English authorities discussed in paragraphs 71 to 73 above. Since the State’s liability through entering into the indemnity was purely contingent on whatever might happen to Rothwells and the bank facility and no particular State asset was depreciated in value on entry into of the indemnity, I think that the same result as the High Court reached in Wardley should also be reached in this jurisdiction.

76.

Whether or not that is, however, accepted, no English authority indicates and I do not consider that the Society’s present cause of action should be regarded as accruing before any change in its legal position occurred and it received any claim on the Fund. First and foremost, the Society’s legal position remained unchanged, even in public law, at least until after it received a claim. Second, it was not possible until after a claim was received for anyone to know which client(s) of Payne & Co might suffer what loss, whether any of them might be able, and choose, to assert that they had as a result suffered hardship justifying a grant out of the Fund and what the circumstances were in which the Society would have to exercise its discretion to make or refuse a grant. Third, in this situation, it is not appropriate to talk of the Fund or any other specific asset of the Society as having suffered any loss at least until after a hardship claim was made on the Society.

77.

It may be that, if the facts had been known contemporaneously, some statistical or experience-based assessment could have been made of the likelihood of a claim or claims emerging, and of the Fund having eventually to make payments, as a result of Mr Payne being able to continue his scheme of fraud. A similar assessment might be made of the risk of future loss of a physical asset (deeds or valuables) of which a solicitor was failing to take reasonable care, but which had not yet been lost or stolen. But I do not consider that the law should treat purely contingent loss assessed on so remote a basis as sufficiently measurable, in the absence of any change in the claimant’s legal position and of any diminution in value of any particular asset. Even where negligence brings about a specific transaction and thus a change in the claimant’s legal position, Lord Nicholls observed in Nykredit [1997] 1 WLR 1627 in the passage at p.1631C-D cited in paragraph 73 above, that the mere entry into the transaction under which “Financial loss is possible, but not certain” is not sufficient detriment.

78.

Looking at the matter more generally, I also see no particular reason to accelerate the accrual of a cause of action where there has been no transaction changing the claimant’s legal position and no diminution in value of any particular asset. Where such factors are present the English authorities considered in paragraphs 67-70 above take a clear-cut, though perhaps strict, view. The House has not been asked to review such authorities, nor would I think it appropriate to do so in the light of the way that they and the English legislation have developed. But where such factors are not present, I see attraction in the approach taken by the Australian High Court in Wardley 175 CLR 514, the effect of which is that unless and until a remote contingency eventuates the claimant is not expected to issue proceedings which he would not normally issue or wish to issue unless and until that point arrives.

79.

Sephtons point out that in Nykredit [1997] 1 WLR 1627, 1630 Lord Nicholls endorsed the Court of Appeal’s approval in Forster v. Outred & Co. of counsel’s submission that actual damage meant “any detriment, liability or loss capable of assessment in money terms”. There can, as I have said, be situations of measurable monetary detriment or loss, even though there has been no change in a person’s legal position. But the passage which I have quoted from p.1631C-D in Nykredit shows that Lord Nicholls was far from accepting that everything that can loosely be described as “detriment” will constitute damage for the purposes of a tort claim. Lord Nicholls’ words are not in any event to be read as a statute.”

Mr Hollander argued that it can now be seen that, if Forster v Outred & Co had involved Mrs Forster entering into an unsecured guarantee of her son’s liabilities to the lender, then her claim would not have been statute-barred, because it would have been a case of a pure contingent liability, not affecting any asset of hers, either already owned or acquired, and a cause of action in tort would not have arisen until a claim was made on the guarantee. By parity of reasoning, he submitted, NIG’s liabilities under the policies were purely contingent, and the same result should follow as regards the date on which a cause of action in tort arose or would arise.

He contended that, even if Sephton could have been decided on the basis that there was in that case (unusually) no transaction from which the contingent liability arose, that was expressly not the basis on which Lord Hoffmann decided the case: see his paragraph 31 quoted at paragraph [113] above. Lord Hoffmann recognised that in many cases involving entry into a transaction which gives rise to a contingent liability, damage may be suffered at once, but he held in terms that the possibility of an obligation to pay money in the future is not damage in itself.

The decision of the House of Lords therefore endorsed Wardley and, as regards the English cases, affirmed the correctness of Forster v Outred & Co as a decision, but limited the width of the observations made in that case, so that if Mrs Forster had entered into a personal unsecured guarantee of her son’s debts, no loss would have accrued unless and until a call was made on the guarantee. That being so, it is necessary to see where the distinction lies between a case such as Wardley, on the one hand, and other cases where the loss has been held to arise on entry into a transaction on the other.

In particular, because of what Lord Hoffmann said at paragraph 21 of Sephton, it is necessary to consider the class of cases involving transactions into which a party enters as a result of negligence by the defendant. In Knapp v Ecclesiastical Insurance Group [1998] PNLR 172, the plaintiff paid a premium for a fire insurance policy which was void because his broker had failed to disclose material facts to the insurer. Through the broker’s negligence, he failed to get what he wanted, namely a valid policy. Similarly in DW Moore v Ferrier [1988] 1 W.L.R. 267, the plaintiff entered into contracts of employment with employees containing restrictive clauses which were later found to be void as being in unreasonable restraint of trade, the provision having been drafted for them by the defendant solicitors. Through the solicitors’ negligence, they failed to get what they wanted, namely a contract of employment with valid restraints. In each of those cases the damage was held to have been suffered as soon as the plaintiff entered into the relevant transaction. The same is true of McCarroll v Statham Gill Davies [2003] PNLR 25 (cited in Watkins – see below – though not directly to us) where the claimant had entered into a partnership agreement on the advice of the defendant, which he said was less favourable than it should have been, as a result of the defendant’s negligence. The cause of action arose on his entry into the agreement.

In Watkins v Jones Maidment Wilson [2008] EWCA Civ 134, the claimants entered into an agreement to buy a plot of land from a builder, who was also to build a house on it for them. The contract included a clause allowing the claimants to terminate the agreement if the building work had not been completed by a given date (August 31 1998), and leaving the claimants in that case liable only to pay the reasonable value of the work actually done, with a provision for expert determination of that amount. After delays in the work, at a time when they believed (wrongly) that the only defects in the building were minor, and with advice from the defendant solicitors, on 6 August 1998 the claimants waived the right to terminate if the work was not complete by the end of that month. The defects turned out to be more serious, and the claimants incurred substantial expense in the resulting dispute with the builder. On 26 August 2004 they sued the solicitors asserting two distinct claims. One was of negligence in drafting the original agreement; the other arose from their advice as a result of which the letter of 6 August 1998 was written. The solicitors relied on a limitation defence. Arden LJ (with whom Longmore LJ and Thomas LJ agreed) held that the causes of action had accrued before 26 August 1998, and the claim was therefore out of time. The claimants’ rights had been affected immediately, both on entry into the original agreement and, thereafter, on writing the 6 August letter. For present purposes the second claim, concerning the 6 August letter, is the more relevant, because the claimants (arguing the case in person) relied on Sephton in support of their appeal on this point. It was argued that the termination provision could not have been used before 31 August, and might not have needed to have been used, so that a claim resulting from the consequences of its loss did not arise until after 31 August. The judge held that the provision had a value before that date, and therefore when the claimants waived the right in advance on 6 August 1998, their action affected the value of their existing rights immediately. Arden LJ said at paragraph 32:

“In my judgment, the judge was correct for the reasons he gave. When the Watkins entered into the building agreement they acquired a bundle of rights. That bundle of rights was of lesser value than they were on their case led to believe that it would be. Those rights were an asset capable of valuation. Thus, the Watkins suffered measurable loss when they acted on the allegedly negligent advice to enter into the later transaction. Accordingly, that claim is statute-barred. ”

Thus, the transaction entered into as a result of the alleged negligent advice altered the claimants’ position immediately, in terms of the value of their rights, even though the extent of that effect was not immediately quantifiable, and at least in theory it might not have arisen at all, if the builders had been more competent or co-operative or both. Mr Hollander submitted that an important factor in the case was the assumed fact that as a result of the defendant’s breaches of duty the claimants entered into the original agreement under which they did not secure the contractual rights which they had instructed the defendants they wanted, and lost a real chance of securing more favourable contractual terms from the builder. It seems to me that this is irrelevant to the argument about the waiver of rights, though it is relevant (expressly so) to the argument about the entry into the original agreement.

Miss Carr relied strongly on Shore v Sedgwick Financial Services [2008] EWCA Civ 863. In that case the claim was for damages arising from the claimant having transferred his pension benefits from an occupational scheme of his employer to a personal pension scheme and making various particular arrangements in respect of it. It was not suggested that the financial value of the assets transferred was less than that of his rights as they had been in the occupational pension scheme. On one basis, therefore, the transfer did not cause him any loss. He argued that the loss arose only later, when he withdrew the maximum possible amount of income from the new scheme, rather than purchasing an annuity. Dyson LJ held that the cause of action accrued when the value of his rights was transferred out of the occupational pension scheme into the personal pension scheme. He said at paragraph 48:

“The PFW scheme was a different kind of transaction from the advance of a loan on the security of a mortgage on property. It was a transaction under which Mr Shore obtained a bundle of rights which, from the outset, were less advantageous to him than the benefits that he enjoyed under the Avesta scheme. On the facts of this case, it was not necessary to wait to see what happened to determine whether Mr Shore was financially worse off in the PFW scheme than he would have been in the Avesta scheme. For these reasons, I would hold in relation to the primary claim that Mr Shore first suffered loss on 28 April 1997.”

The judge’s reference to the benefits as being less advantageous “to him” has been criticised or questioned. But even if those two words are ignored, it is not difficult to understand why in a case of that kind the loss was suffered at the time of the transfer. The secondary claim was based on negligent failure to advise the claimant to take an annuity as from July 1997 (a couple of months after the transfer). It was said that the steps which he did take gave him a higher income than an annuity would have done until May 2000, so that he did not suffer loss until that date. Dyson LJ rejected this argument as well, at paragraph 70:

“I would reject this argument for the same reasons as those for which I have rejected the corresponding argument in relation to the primary claim. The essence of the complaint is that SFS failed to take steps to protect him from the risks inherent in the PFW scheme by purchasing an annuity as from July 1997. For the reasons already given, Mr Shore suffered loss in July 1997 when he did not acquire an annuity which would yield a secure income instead of the uncertain income that could be withdrawn under the PFW scheme.”

I do not find in this decision any compelling analogy with the present case, such as Miss Carr suggests. The case has nothing to do with contingent liabilities. Like many of the other cases it involved entry into a transaction which presented risks for the party in question, which might or might not materialise, but not risks in the nature of contingent liabilities.

As Lord Hoffmann said in Sephton at paragraph 22, the claimant may pay money, transfer property, incur liabilities or suffer diminution in the value of an asset, and in return obtain less than he should have done. In such a case entry into the transaction can be seen to give rise to quantifiable damage, even though further damage might arise later. But he distinguished these cases from those in which purely contingent obligations are incurred. At paragraph 30 he recognised that a contingent liability may depress the value of property, as it did in Forster v Outred & Co, or it may mean that a party to a bilateral transaction has received less than he should have done, or is worse off than if he had not entered into the transaction. In paragraph 31 he also recognised that, for reasons such as these, a party to a bilateral transaction who incurs contingent liabilities under the transaction will often suffer loss on entry into the transaction. But Mr Hollander argued that a case in which the whole purpose of the transaction is the contingent liability cannot be put in the same category. True, Sephton was not concerned with a bilateral transaction, but Wardley was, and the House of Lords not only approved Wardley, but did so on the basis of the statement of principle, that a contingent liability is not as such damage until the contingency occurs, and that the possibility of an obligation to pay money in the future is not damage in itself.

What Lord Mance said in the latter part of paragraph 75 (cited above at paragraph [118]) is also inconsistent with a conclusion that coming under a contingent monetary liability, not affecting any particular asset, is of itself damage. At paragraph 77 he referred to the need for a change in the claimant’s legal position and of the diminution of value of a particular asset if contingent loss is to be regarded as sufficiently measurable as to give rise to a cause of action. The first sentence of paragraph 78 is to the same effect. As I read it, those requirements are cumulative. To incur a contingent liability is itself (in almost all cases - the Law Society’s position was exceptional) a change in the legal position of the person under the liability, but that is not enough of itself, according to Sephton.

On that basis, Mr Hollander argued that the judge was wrong to hold that, in the present case, as regards vetting breaches, loss was suffered by NIG as soon as the policy was issued. The relevant duty of the solicitors was to assess the legal prospects of the particular claim and to accept it under the scheme, and so to issue the ATE policy to the relevant Scheme Claimant, only if the claim satisfied the stated conditions as to prospects of success and quantum of claim. If the solicitors had done their job properly, the policies complained of would not have been issued at all and NIG would not have been at risk. The risk, however, is not sufficient to constitute damage, because it was only a contingent liability, and no particular asset of NIG was affected, as Mrs Forster’s freehold property was on the facts of her case. NIG’s legal position was affected by the issue of the policy, but so was that of the State of Western Australia in the Wardley case, by the execution of the indemnity, and so would Mrs Forster’s position have been if even she had only entered into an unsecured personal guarantee. That was not sufficient to cause immediate damage, according to the High Court of Australia, and the House of Lords agreed with that, as a matter of English law, in Sephton. The fact that the contingent liability arises under a bilateral transaction therefore makes no difference by itself. Mr Hollander therefore asks rhetorically what it is that distinguishes the present case from Wardley, or from Sephton. Not the fact that there was a transaction, because there was a transaction in Wardley. Could it be the fact that NIG obtained monetary consideration, unlike the State of Western Australia? It is difficult to see why that should be sufficient.

We were also referred to the decision of the Supreme Court of New Zealand in Davys Burton v Thom [2009] 1 NZLR 437, in which Sephton was considered. Members of the court expressed differing views, unnecessarily to the decision in the case, as to when actual damage would be suffered in the case of an unsecured personal guarantee. The observations are interesting, but were made by a court which is not bound, as we are, to follow Sephton.

The judge recorded Mr Hollander’s submissions to him as proceeding on the basis that there were three relevant categories of case. The first is where as a result of the defendant’s negligence, the claimant has incurred a potential liability which may or may not accrue, but has not suffered any separate damage on entry into the transaction. In such a case he argued that the potential liability is a contingent liability, and is not to be considered as amounting to loss to the claimant unless and until the contingency occurs, however likely it is to occur. The second category is where it was the defendant’s duty to bring about a transaction having particular characteristics or features, and it was possible for the defendant to fulfil that duty, but due to his negligence, the claimant entered a different transaction lacking one or more of those features. In those cases actual damage may be suffered on inception of the transaction where the assets (tangible or intangible) obtained under the transaction may properly be viewed as ‘damaged or defective goods’. The third type of case is where but for the defendant’s negligence, the claimant would not have entered into the particular transaction at all. In these cases the claimant will suffer loss only when the balance between the value of the liabilities acquired under the particular transaction and the value of any assets obtained under it becomes negative.

The judge rejected this approach. He said that such rigid categorisation was not supported by the decided cases, and did not work in any event because some cases fell into more than one of the second and third categories. He also rejected the proposition that the present case was to be equated with Sephton, because he said that actual damage had been suffered when each ATE policy incepted. He said that, if the categories were to be applied, the present case would fall into the second category, because the solicitors’ duty in question was ‘to bring about a transaction with particular characteristics or features’. At paragraph 30 he said that because the defendant’s duty in tort was to use reasonable care in vetting claims to ensure that the transaction in question had a particular feature, that is to say prospects of success of 51% or more, the damage was suffered in each case when the flawed transaction (i.e. the policy with less than a 50% prospect of success) was entered into, on the basis that it was then that NIG did not get what it ought to have got. In turn at paragraph 31 he rejected an argument of Mr Hollander as follows:

“Mr Hollander submits that the relevant asset, the premium, was not ‘damaged’ when the transaction was entered into and that the asset remained worth the same, unless and until the claim failed for whatever reason and it was only then that NIG suffered actual damage. I agree with Miss Carr that the fallacy in that argument is that it ignores, in a wholly artificial manner, the fact that, because the prospects of success were 50% or less, the insurer is exposed to a greater risk for the relevant premium than was intended and thus the insurer is not getting what it was the panel solicitors’ duty to ensure that it got.”

I agree with the judge that it is not helpful to apply an over-rigid classification of cases in this area. He said at the end of paragraph 59 that “whether or not actual damage has been suffered in such a case depends not upon the division of cases into rigid categories, but upon the facts of the particular case”. I agree. If too much attention is given to classification, one may be led into an unproductive focus on the precise terms of the classification, rather than the substance of the issue. Thus, it seems to me that, in answer to the proposition that the solicitors’ duty in the present cases was a positive one, that is, to exercise reasonable care to bring about transactions with particular characteristics or features, the duty could at least equally well be expressed in the negative, namely that the duty was to take reasonable care not to bring about transactions with particular characteristics or features, that is to say, not to cause the insurer to enter into policies where the prospects of success were too small or the likely damages too low. The consequence of the negligence, which the judge in the passage cited described as “the insurer is not getting what it was the panel solicitors’ duty to ensure that it got” could as well, or as it seems to me better, be expressed as “the insurer is getting something that it was the panel solicitors’ duty to ensure that it did not get”.

This is not like a case where the professional’s task arises in relation to a transaction with a third party which could be entered into either on terms consistent with the client’s instructions or on less favourable terms not so consistent, such as Knapp, DW Moore v Ferrier and McCarroll, cited above. This was a case in which the professional’s duty was to cause the principal to enter into transactions which had particular characteristics, with third parties in relation to whom those characteristics were satisfied, and not into transactions with others where they were not satisfied. The negligence arose in the fact of entry into the transaction at all, not as regards the terms on which it was entered into. For my part I do not find the analogy with “damaged or defective goods” helpful, if it is intended to convey that, but for the negligence, the client might have acquired “goods”, that is to say here the policy, which were not “damaged or defective”. The claimant’s case here is that, but for the negligence, it would not have entered into any policy with the particular Scheme Claimant. It is not a case (such as could arise on the part of an insurer) where the complaint is that, because of the negligence of a professional, such as a broker, the level of the risk was underestimated and as a result the premium charged for the policy was inadequate. These were not policies which NIG would have entered into on any terms, if it had known the true assessment of the prospects of the relevant Scheme Claim.

The judge’s conclusion on this is set out in his paragraph 96. He held that the result of the negligence was that NIG was committed to a series of flawed transactions, in the form of policies where the prospects of success were less than 51%, and that NIG was therefore exposed to a greater degree of risk than it was entitled to expect if the panel solicitors had complied with their duty. NIG suffered actual loss when committed to the flawed policy and thus exposed to the risk.

Mr Hollander submitted that the judge was wrong to reject his arguments, and that the judge’s reasoning to which I have referred did not meet the points arising from the decision in Sephton. He argued that that decision is precisely in point because what the solicitors’ negligence exposed NIG to was a contingent liability to which they should not have been subject. Being a contingent liability, it might not materialise, and would not do so for some time. In the meantime NIG had not suffered actual loss because it had received the premium. Although it was exposed to a liability in return for the premium which was greater than should have been the case, that was a contingent liability which, from its nature and in accordance with Sephton, was only to be treated as constituting damage, for this purpose, when it became an actual liability.

From a commercial and economic point of view, it seems an obvious proposition that if an insurer enters into a disadvantageous policy, it suffers actual loss at the time of the policy. Miss Carr submitted that this is consistent with the judge’s decision, and with earlier authority including Companhia de Seguros Imperio v Heath (RBEX) Ltd [1999] Lloyd’s Rep IR 571, and is not inconsistent with the decision in Sephton.

In Imperio Langley J had to consider a claim in negligence against insurance brokers. One of the issues was limitation. Rejecting one aspect of the argument (on the part of Mr Flaux Q.C.) that the claim in tort was not barred, the judge said, at 587:

“The questions whether and when damage is suffered are questions of fact … But the receipt of something less valuable or the transfer of something of value without an agreed protection is itself damage. There must be actual damage within the measure of damage applicable to the wrong in question for the cause of action to accrue but an increase in a plaintiff's obligations plainly may constitute such actual damage. Nor is it relevant that the damage may later become more serious or capable of more precise quantification.

I have already recorded that the claims in respect of fronting are alleged in paragraph 16 of the Points of Claim to have caused loss to Imperio by exposing the company to a greater liability or potential liability than the net liability subscribed and for which the premium was calculated and paid. In the course of the hearing Mr Flaux was not inclined to dispute that such exposure for no compensation was damage and thus that the cause of action in tort accrued, as the cause of action in contract, when Imperio was committed by Heaths to front. In this, in my judgment, Mr Flaux was plainly right. Insurance is the business of undertaking risk for reward. An insurer who is committed to a greater risk than agreed for no reward has thereby suffered a real loss.”

Those comments are pertinent, but it seems to me that (as Flaux J candidly accepted at paragraph 62 of his judgment) they are less cogent than they would otherwise have been, given that Counsel did not dispute the proposition that “such exposure for no compensation” was actual damage. The point was not pursued on appeal and, unsurprisingly, the case was not cited to the House of Lords in Sephton.

Poole v HM Treasury [2006] EWHC 2371 (Comm), another decision of Langley J, was also relied on by Miss Carr in this context. The claimants relied on the then recent decision in Sephton in response to a limitation plea. Having referred to that case and stated that it concerned a contingent liability, the judge said this at paragraph 239:

“But, in this case, I do not think the court is addressing a loss which is only prospective and might never be incurred. By becoming a Name (and joining a Syndicate or Syndicates) a Name was committed to the liabilities of that Syndicate under RITC it had written and on all business (including RITC) it did write in the course of his membership. Quite apart from joining fees and the provision of charges and guarantees (ASF 2.2.27) there were therefore actual liabilities undertaken albeit unquantified. This was not a contingency standing alone in the sense with which, as I understand their speeches, their Lordships were concerned in Sephton: see in particular, Lord Hoffmann at paragraph 30. The Names were worse off than if they had not become or continued to be Names.”

His reasoning, therefore, was not that actual damage was caused even though the liabilities were contingent, but rather that the liabilities were not merely contingent. It is also worth noting that, at paragraph 240, he commented that the issue was academic, at least as regards any direct significance.

Neither of these decisions is binding on this court, and it seems to me that we have to come to our own decision as a matter of principle in the light of Sephton, on the basis of a proper view of the nature of the liabilities to which NIG became exposed by the solicitors’ negligence, and the consequences of that negligence more generally, and, so far as relevant, a comparison with the position that NIG would have been in but for that negligence.

Miss Carr placed considerable reliance on Shore and on Watkins. Neither case was concerned with contingent liabilities. In the case of Watkins, to which I have referred at paragraph [123] above, the comparison (on the second claim) had to be between the position in which the claimants found themselves, as a result of the (alleged) negligence of the defendant, namely having given up a right under the existing contract which, as it turned out later, would have been valuable, and the position in which they would have been if they had not given it up. That comparison showed clearly that they were at once in a less good position because they no longer had a contractual right which could have been valuable (as the judge held) even before the latest date for its exercise, not least because of its possible usefulness in negotiation with the builder. Accordingly, that case does not seem to me to assist on the resolution of the present issue.

Shore is even further away from the facts of the present case, and from any issue concerning contingent liabilities, and I do not feel it necessary to lengthen this judgment any further by discussing it.

On the basis that the proper analysis of the question before us should be guided by the facts of the case, rather than by how those facts might fit into one or another system of classification, it seems to me that the essential facts, on the vetting claims, are that, by reason of the solicitors’ negligence, NIG was caused to enter into a policy in favour of a Scheme Claimant which did not comply with its preconditions as regards such policies, and to which it ought therefore not to have been bound. By comparison, if the solicitors had not acted negligently, NIG would not have entered into that policy. The consequences were, on the one hand, that NIG received the premium under the policy, which it would not otherwise have received, and, on the other hand, that it was exposed to the contingent liabilities under the policy. Because of the solicitors’ negligence as regards vetting, those liabilities were more likely than not to materialise.

As I have said, that situation, and its repetition over many instances, put NIG into a seriously worse commercial and economic position than that in which it wished to be, ought to have been, and would otherwise have been. If the true facts had been known at the time, NIG’s accounts could and would have included a provision for the contingent liabilities arising under the unsatisfactory policies, in addition to whatever provision was appropriate for the compliant policies under which, of course, there were also contingent liabilities. Any potential investor, for example, would have regarded investment in NIG as less attractive as a result. The solicitors had mounted an alternative argument along these lines, which the judge rejected at paragraphs 108 to 111. The solicitors sought to rely on the same point before us in a Respondent’s Notice.

It seems to me, however, that the adverse effect on NIG’s commercial and economic position does not provide the answer to the question whether, in the eyes of the law, NIG had already suffered actual damage. Lord Hoffmann rejected that sort of factor in terms at paragraph 29 of his speech in Sephton, and Lord Mance made a similar point in his paragraph 77, both passages being quoted above, at paragraphs [113] and [118]. The same point was relevant on the facts and was made by the High Court in Wardley. That was one of the judge’s reasons for rejecting the argument now pursued in the Respondent’s Notice: see paragraph 111.

Miss Carr also submitted that it was relevant that NIG was in the business of a commercial insurer, issuing policies and undertaking risks in return for premiums, unlike the position of either Mrs Forster or the State of Western Australia, each of which entered into a one-off transaction. I do not see that this is a relevant distinction. How often a particular person is exposed to the negligence of professional advisers does not seem to me to be relevant to the analysis of the nature of the loss suffered, though it may be to the nature and level of the duty and possibly to issues such as contributory negligence. It is true that Deane J in Wardley referred to the State undertaking an isolated or one off contingent liability, and to the State not being in the business of incurring such risks: see (1992) 175 CLR at pages 541 and 544-5. However, I find no other support for this being a relevant factor, and it is not mentioned by the other judges in the High Court.

If, then, the correct comparison is that which I have described at paragraph [146] above, how does that stand in relation to the decision in Sephton? In terms of the judgment in Wardley, approved expressly by Lord Hoffmann, Lord Scott and Lord Mance, NIG entered into a contract as a result of the solicitors’ negligence which exposed it to a contingent liability, namely the liability under the policy (see the passage quoted at paragraph [106] above). As regards Lord Hoffmann’s speech, it is not a case in which the solicitors should have caused NIG to enter into a different transaction with the particular Scheme Claimant, and where they caused loss because of the disadvantageous terms of the contract: it should not have been brought into contract with the particular Scheme Claimant at all (see his paragraph 22). The existence of the contingent liability did not depress the value of any other property of NIG (his paragraph 30). It did not mean that NIG, as a party to a bilateral transaction (i.e. the policy) received less than it should have done, and that is because it ought not to have been bound to a transaction with that policyholder at all (paragraph 30 again). In that sense, NIG was worse off than if it had not entered into the transaction, but that is so only because of the existence and burden of the contingent liabilities. Taking then what Lord Walker said, quoted at paragraph [115] above, the case does seem to me to be analogous with the hypothetical variant on Forster v Outred in which Mrs Forster had entered into an unsecured personal guarantee of her son’s liabilities. In terms of what Lord Walker said at paragraph 52, NIG did suffer, in a sense, a detriment, but not a detriment of the sort described in the relevant cases. As regards Lord Mance’s speech, this is not a case where there has been any transaction changing NIG’s legal position, other than entry into the policy, thereby incurring the contingent liabilities, nor has there been any diminution of the value of any particular asset: see his paragraph 78. In that paragraph Lord Mance speaks of “a remote contingency”, whereas it may fairly be said that in the present case the contingency was far from remote – the fact that it was all too proximate is the basis of the allegation of negligence. However, it does not seem to me that there is any other suggestion in the case that the likeliness or otherwise of the contingent liability becoming an actual liability has anything to do with the principle. Among other things it would be inconsistent with Lord Mance’s rejection at paragraph 77 of a hypothetical assessment, at the time, of the likelihood of claims arising. The point is also made in Wardley, for example in the passage quoted above at paragraph [104] (“perhaps the virtual certainty”) and also by Toohey J: see (1992) 175 CLR at 556.

It seems to me that Mr Hollander’s submission, based on the passages to which I have referred, that the detriment suffered by NIG as a result of the solicitors’ negligence is the incurring of contingent liabilities, and that Sephton decided that this does not count as actual damage, is cogent and formidable. Miss Carr relied, to the contrary, on Lord Hoffmann’s words in his paragraph 31: “No doubt in most cases in which a party incurs a contingent liability as a result of entering into a transaction, that liability will result in damage for the reasons already discussed in relation to bilateral transactions.” In Wardley the State entered into a bilateral transaction, under which it undertook a contingent liability. If the true facts had been known, immediately after the indemnity had been given, the likelihood of the contingent liability becoming actual would have been reckoned to be high. Both the leading judgment and that of Brennan J proceeded on that assumption, but nevertheless it was held that no loss was suffered as a result of the giving of the indemnity itself, nor until an absolute liability to pay had arisen: see the passage quoted at paragraph [104] above and Brennan J (1992) 175 CLR at 538. Likewise, in the hypothetical variant on Forster v Outred, if Mrs Forster had entered into an unsecured personal guarantee, that would have been a bilateral transaction. It is clear from the decision in Sephton that in neither of these cases did the actual damage occur until the contingent liability became actual.

Miss Carr submitted that the language of the House of Lords in Sephton, especially references to “a purely contingent liability” or to a contingent liability standing alone, could not be considered to apply to facts such as the present. She drew a distinction between such words, on the one hand, and cases of bilateral transactions with benefits and burdens (such as Knapp, or DW Moore v Ferrier), cases of packages of rights less valuable than the client was entitled to expect (the same examples are in point), and cases where the claimant incurs a risk secured against or impacting on an asset (such as Forster v Outred). She submitted that the present is not a case of a pure contingency for several reasons. First, NIG was a commercial insurer whose business involved accepting risks for reward, and which entered into the CLE scheme as part of that business with a view to profit. Secondly, in entering into the scheme and the particular policies under the scheme NIG committed itself to the funding and administration of the claims, and to indemnifying the Funder and the Scheme Claimant if and insofar as necessary. By doing so, in cases where the panel solicitor had been negligent in vetting, it exposed its assets to greater obligations than should have been the case, with a weaker prospect than it should have had of the costs being recovered from the third party. That increased risk was no isolated or standalone contingency but a direct result of NIG’s entry into the policy, a bilateral transaction under which the premium was inadequate to cover the enhanced risk.

For all that, the fact remains that until a claim arose under each policy, NIG’s liability was contingent, and nothing but contingent. NIG’s position was, of course, worse as soon as the policy was entered into than it should have been. But it was worse because, and only because, of the contingent liability which it incurred, being one which it ought not to have incurred, and which it only incurred as a result of the solicitors’ negligence. Mrs Forster’s position would have been worse if she had given an unsecured guarantee of her son’s indebtedness, and it would have arisen directly from a bilateral transaction. However, it is clear, from what the House of Lords said about that hypothetical case, that they regarded the incurring of actual damage in such a case as occurring only when the liability under the guarantee was called. The State of Western Australia’s position was worse as soon as it entered into the bilateral transaction under which it gave the indemnity to Rothwells’ bankers. Again actual damage was only suffered when the indemnity was called.

For those reasons I cannot accept Miss Carr’s attempt to differentiate the present case from those which the House of Lords had in mind in Sephton.

The perplexing nature of the conundrum which is posed by the application of Sephton to this case is illustrated by the fact that, in their judgments, Arden LJ and Longmore LJ have come to the opposite conclusion to that which I favour. There is much common ground between us, and I do not wish to dwell at any length on the points where we disagree. Arden LJ puts the points in favour of the judge’s conclusion on the basis of a deep analysis of the speeches in Sephton. Longmore LJ’s reasoning relies particularly on an analogy with Nykredit and on analysing the present case as one of flawed transactions. I will deal briefly with some of the points on which our respective analyses diverge.

Arden LJ says at her paragraph [60] that NIG suffered loss in law on inception of each policy because the liability under each ATE policy was greater than it should have been. She expresses the view at her paragraph [59] that the premium and the risk under the policy cannot be separated the one from the other, so as to treat the premium as if it was not harmed. She identifies actual damage to NIG from the issue of each policy in the fact that the ability to NIG to use the sums representing premiums to meet claims, or to create or contribute to a reserve for meeting claims, was adversely affected because the matching liabilities were greater than they should have been. In turn at her paragraph [62] she observes that, in addition to the incurring of liabilities under the policies of insurance, NIG suffered additional and immediate loss in the fact that those liabilities were more burdensome than they should have been if the breaches had not occurred.

My difficulty with that approach is that, in terms of the analysis expressed in Sephton, I do not see that the additional loss which she identifies in paragraph 62 is additional to, or in any way different from, the loss resulting from incurring the contingent liabilities. I recognise that, in economic and commercial terms, NIG suffered by reason of the issue of the policies, even before liability arose to make any payment under them. But that is not a kind of loss which, as it seems to me, can be recognised as actual damage because of what the House of Lords said in Sephton and in particular their approval of Wardley (see paragraphs [147] and [148] above). I do not see how, in legal terms, the incurring of the contingent liabilities under the non-compliant policies can be detached from the economic or commercial consequences of those liabilities having been incurred.

Moreover, while I would agree that if security for a contingent liability were created over an asset such as a debt, or a portfolio of debts, as distinct from over real property, actual damage would be suffered at that moment, by analogy with Forster v Outred, it seems to me that the position is essentially different where no security is given, so that the effect of the contingent liability having been incurred is on the general assets of the person under that liability, rather than on any given asset by way of security. I do not understand it to be any part of the present case that the premiums received under the ATE policies were to be earmarked in any specific way to meet liabilities under the policies. Nothing in the statement of assumed facts suggests that this was the case. Presumably NIG had other premium income under other underwriting activities which it could, and no doubt did, call on to meet liabilities under these policies. Thus, the adverse effect of NIG’s participation in the Scheme and the Panel solicitors’ negligence was its impact on NIG’s general assets, not on any particular asset or assets. The granting by the State of Western Australia of its indemnity had an adverse effect on its general assets and, if Mrs Forster had given an unsecured personal guarantee, that would have had an adverse impact on her overall financial position. Accordingly I do not see that it is consistent with Sephton to identify such an effect as constituting actual damage, where the reason for the general adverse financial or commercial effect is the incurring of one or more contingent liabilities, and nothing else. I do not see that the premium under a non-compliant policy can properly be said to have been harmed any more or any less than any other of NIG’s general assets.

In the present case, the policy into which the solicitors caused NIG to enter with a Scheme Claimant, in breach of the terms of the instruction and authority, gave NIG a premium which it would not otherwise have received. That was therefore an advantage for NIG. However, it also exposed NIG to a liability under the policy which was clearly a contingent liability. It would not become actual until the Claim had failed (except, as mentioned above, in the possible case, but itself uncertain and future, of an early order to pay Opponent’s Legal Costs). Liability under the policy would arise if cover was withdrawn but, as discussed at paragraph [93] above, this would be the result of the Claim not being Successful, in terms of the policy definitions. Given that the prospects of success were not reasonably assessable as being as high as 51%, it was more than likely that the liability would become actual, but it does not follow that there would be a claim under the policy in every case of negligence. To the extent that, notwithstanding negligence by the solicitors, a claim has not in fact arisen, NIG was better off, and no claim will have been brought against the relevant solicitors. On the judge’s decision, however, a claim in tort could have been brought in every such case, whether or not it turned out to cause NIG real financial loss because a claim was made under the policy.

Miss Carr submitted that there was no logical distinction between the present case and Nykredit, in which the surveyors provided a negligent valuation, as a result of which the lender entered into a transaction which it would not have undertaken, and it was held that the cause of action in tort did not arise until it was possible to say that the lender was on balance worse off, having regard to the balance of benefits and burdens. I see some force in that comment. However, Lord Hoffmann distinguished Nykredit in Sephton at paragraphs 19-20 on the basis that it was not concerned with contingent liabilities. It seems to me, therefore, that we are bound to treat a contingent liability case as being different. Of course, in any given case there may also be other aspects of the transaction which constitute damage. Here, however, NIG sustained no loss, of any kind, on entry into the policy other than that resulting from incurring the contingent liability to pay sums of money under the policy. It seems to me that Sephton, having expressly affirmed Wardley in this respect, binds this court to hold that incurring such a liability does not constitute actual damage. I respectfully disagree, for this reason, with Arden LJ’s conclusion that NIG suffered loss on issue of the policy which was additional to that which was caused by it incurring the contingent liabilities.

Likewise, my difficulty with the flawed transaction approach is to see how a principled distinction can be drawn under which the ATE policies in the present case were flawed transactions, causing immediate loss, but the indemnity in Wardley, and the hypothetical unsecured guarantee in an adapted version of Forster v Outred, were not, and did not give rise to immediate loss.

Hitherto I have considered only the vetting breach claims. On the basis of the analysis set out above, it seems to me that the same result follows as regards the conduct breaches. In those cases there had been no negligence in committing NIG to the policy, but the negligence lay in one of two kinds of conduct. One is where the solicitor failed to advise NIG (through CLE) of circumstances which showed that either or both of the prospects of success and the quantum of the claim was now too low, so that NIG was unable to exercise its right to withdraw from the policy and exclude its exposure for future costs. The other is where the solicitors conducted the claim negligently so that the chances of success in the claim became materially worse.

It seems to me that the first type of case must be governed by the same principle as applies where the negligence is in causing NIG to enter into the policy in the first place. In each case NIG comes under, or is under, a liability which is still only contingent. In the conduct cases it could have escaped from that liability, or at least have substantially limited its exposure under the liability. The contingent nature of the liability seems to me to show that, in terms of the tort claim, actual damage would be suffered only when and if the liability becomes an actual liability to pay under the policy.

The same must be true of the other kind of conduct breach, where the negligence is such as to increase the chances of the liability becoming an actual liability to pay, because the claim has been badly conducted.

I have found this case far from easy to decide, not least because the application of the decision in Sephton seems to me to produce a result which is at odds with that which one would anticipate, in terms of the commercial and economic reality of NIG’s position.

In the end, despite Miss Carr’s submissions and the respective reasoning of Arden and Longmore LJJ, I am persuaded by Mr Hollander’s arguments as to the effect of the decision of Sephton, and I would allow the appeal. I would reject the arguments in the Respondent’s Notice which do not seem to me to be essentially different from the points arising on the appeal. On the Respondent’s Notice points, it seems to me that the judge was right in his paragraph 111. My respectful disagreement with the judge is, in essence, on the basis that he should have accepted the same arguments as fatal to the solicitors’ primary argument as well. I would hold that the preliminary issue should be answered by saying that, in relation to each type of claim, time started running for a claim in tort only when NIG first came under an actual liability to make a payment under the relevant policy.