Caparo Industries v Dickman

My noble and learned friend, Lord Steyn has set out the essential matters of fact in this case which statement I gratefully adopt. In my view, as in his, the majority in the Court of Appeal were entitled to find and right in finding on the evidence that the brokers had undertaken a duty not merely to obtain reinsurance cover in the sum of $11m but also to advise on the availability of reinsurance cover in the market, without which the transaction would not have gone ahead. To give that advice involved an investigation as to the market's assessment of the risks involved, as my noble and learned friend, Lord Lloyd of Berwick has shown.

It is now accepted that no such reinsurance was available and that the appellants did not advise to that effect. In failing to make the necessary inquiry as to the market's assessment of the risk and to advise that such reinsurance was not available, the appellants were in breach of their duty. I agree with Lord Lloyd and Lord Steyn that the scope of the duty in this case is not the same as in South Australia Asset Management Corporation v York Montague Ltd [1997] AC 191 . It is to use Lord Steyn's phrase on the other "side of the line drawn in SAAMCO ".

I agree with Lord Lloyd and Lord Steyn that there is no justification for limiting the damage to the loss flowing from the failure to obtain the $11m reinsurance: it is the whole loss resulting from their entering into the insurance cover when no reinsurance was available.

The Court of Appeal on this came to the right conclusion and I would dismiss the Appeal.

I have had the advantage of reading in draft the speeches of my noble and leaned friends, Lord Lloyd of Berwick and Lord Steyn. I agree with them and for the reasons which they have given, I, too, would dismiss the appeal.

The question in this case is whether the brokers, Johnson & Higgins Ltd, are liable in negligence for the whole of the foreseeable loss suffered by Aneco Reinsurance Underwriting Ltd as a consequence of entering into a treaty of reinsurance with an underwriter at Lloyds, or whether the recoverable loss is limited by the principle stated in Banque Bruxelles Lambert SA v Eagle Star Insurance Co Ltd [1997] AC 191 , also known as South Australia Asset Management Corporation v York Montague Ltd ("SAAMCO") and subsequently applied in Nykredit Mortgage Bank plc v Edward Erdman Group Ltd (No 2) [1997] 1 WLR 1627 and Platform Home Loans Ltd v Oyston Shipways Ltd [2000] 2 AC 190 .

There is no dispute as to the law. Mr Hunter QC for Aneco accepts that the brokers cannot be held liable for losses which fall outside the scope of their duty of care. Nor is there any longer any dispute as to the primary facts. Thus the sole question is whether in the particular circumstances of this case the brokers' duty of care was limited to the obtaining of satisfactory excess of loss protection on behalf of Aneco, in which case, as the brokers concede, they are liable for about $11 million, but no more; or whether as Aneco contend the brokers assumed a much wider duty of care, in which case they are liable for the full extent of Aneco's loss, amounting to about $35m.

It is convenient to start with Youell v Bland Welch & Co Ltd (No 2) (The "Superhulls Cover" case) [1990] 2 Lloyd's Rep 431, the facts of which were very similar. Brokers were instructed to obtain reinsurance on the London market on behalf of insurers, in respect of construction risks on three newbuilding vessels. The brokers informed the insurers that they had obtained reinsurance as "original". But they were wrong. The reinsurance was subject to a cut-off clause whereby the cover terminated 48 months after the commencement of construction. The brokers failed to inform the insurers. Had the insurers been given that information, they would not have accepted the reinsurance, and would have written greatly reduced lines on the original insurance. Phillips J, as he then was, held that the brokers were in breach of their duty of care both in contract and tort, and that the measure of damages was equal to the difference between the amount for which the insurers became liable on the original insurance, and the amount for which they would have been liable if they had written reduced lines. There was an alternative claim for damages based on the amount the insurers would have recovered on the reinsurance if it had not contained the 48 month cut-off clause. But counsel for the brokers conceded that the primary way in which the insurers put their case was the correct approach.

Evans LJ [2000] 1 All ER (Comm) 129 held that the decision of Phillips J in the Superhulls Cover case was of direct assistance in the present case, and was correctly decided. Mr Sumption QC submitted that the decision cannot stand in the light of SAAMCO . My noble and learned friend, Lord Millett is of the view that the decision ought to be overruled; but he would forgive the learned judge for failing to anticipate the decision in SAAMCO since the point was conceded by counsel.

My own view is that the point was correctly conceded, and that the subsequent decision in SAAMCO has not changed the relevant law, or undermined the authority of Phillips J's decision. It would be odd if it had, since the Superhulls Cover case was not included among the 53 cases cited by counsel in SAAMCO ; and its existence cannot simply have been overlooked, since Mr Sumption was leading counsel in both cases.

Why, then, does the SAAMCO "principle" not touch on the decision in the Superhulls Cover case? What indeed is the SAAMCO principle? It is surely the principle which has been common ground throughout the argument before us that a defendant is not liable in damages in respect of losses of a kind which fall outside the scope of his duty of care. There was nothing new in that principle. It has been the rule in contract since the decision in Czarnikow v Koufos [1969] 1 AC 350 , if not before. It has been the rule in tort since In Re Polemis and Furness Withy & Co Ltd [1921] 3 KB 560 was disapproved in Overseas Tankships (UK) Ltd v Morts Dock and Engineering Co Ltd (The Wagon Mound (No 1) [1961] AC 388 .

What was new and important in SAAMCO was the application of the principle to valuers, so as to exclude their liability for loss due to a fall in the market: see Platform Home Loans Ltd v Oyston Shipways Ltd [2000] 2 A C 190 , 209 per Lord Hobhouse. Thus in the case of valuers, and their like, that is to say, those who undertake to provide specific information, the SAAMCO principle gave rise to a sub-rule, that valuers are not generally liable (the word is that of Lord Hoffmann, at p 214) for all the foreseeable consequences of their negligence, but only for the consequences of the valuation being wrong. It follows that the damages will usually, though not always, be limited to the difference between their valuation and the correct value; see p 222, per Lord Hoffmann, and the Platform Home Loans case p 210, per Lord Hobhouse.

In paragraph 12 of their printed case the brokers state this sub rule as if it were a rule of general application in the law of contract.

Next I should say a word about Banque Keyser Ullmann SA v Skandia (UK) Insurance Co. Ltd. [1991] 2 AC 249. The facts were complicated, but are set out with conspicuous clarity in the speech of Lord Millett. I do not repeat them. Prior to SAAMCO , the decision of the House was regarded as depending in the end on a short question of fact, albeit one which eluded the Court of Appeal. The fraud of Mr Lee, and the failure of the insurers to disclose that fraud to the plaintiff banks was at most (in the old forbidden language) a causa sine qua non of the financial loss suffered by the banks. The causa causans was the fraud of Ballasteros. By reason of the latter fraud the insurance which the brokers failed to obtain was in truth valueless.

In SAAMCO Lord Hoffmann observed that implicit in Lord Templeman's reasoning was an assumption as to the scope of the brokers' duty of care. Unfortunately the facts surrounding the employment of the brokers in Skandia were never investigated, since the banks' claim against the brokers was settled prior to the trial on payment by the brokers of £10,500,000, being the full amount of their professional indemnity cover. So the assumption as to the brokers' duty of care in Skandia throws little, if any, light on the scope of the brokers' duty of care on the facts of the present case.

It is to those facts that I now turn. Was the duty of the brokers confined to the obtaining of excess of loss protection for Aneco, and informing Aneco that they had done so? If so, I would sympathise with Aldous LJ's conclusion. The damages would be limited to the value of the reinsurance which they failed to obtain, namely, $11m. But I am quite unable to accept that the duty of the brokers was so narrowly confined. At the very least they owed a duty to inform Aneco whether or not reinsurance was available. If they had performed that duty carefully, they would have told the insurers that reinsurance was not available, in which case "the whole thing would have collapsed", as the brokers well knew. For it would have been obvious to Aneco that the unavailability of reinsurance was due to the current market assessment of the risks. It is really fanciful to suppose that there might have been some other reason for reinsurance being unavailable. Why then should the brokers not be liable for the full extent of the losses attributable to their breach of duty? Why should it be assumed in favour of the brokers that reinsurance was available on the market, thus limiting their liability to $11m, when if they had done their job properly they would have known that it was not?

But the matter does not stop there. After reviewing the letters and telexes passing between the parties, and the oral evidence of Mr Forster on behalf of the brokers and Mr Crawley on behalf of the insurers, Evans LJ concluded that the duty went wider then a simple duty to inform; it included a duty to advise. As early as 18 November 1988 Mr Forster was writing to Mr Crawley: "Would suggest you buy XL protection for up to 10 times income and would envisage comprehensive protection being available at between 30%-40% of NPI…." Again on 7 December Mr Forster wrote:

It is said that the brokers were acting in two different capacities, as brokers for the underwriter of the underlying business in negotiating the Bullen Treaty, and as brokers for Aneco in negotiating the XL cover. This is true. But I doubt whether it is all that uncommon. In any event if the brokers had found themselves in a difficult or impossible position as was suggested, they would doubtless have dropped out. I should be surprised if, as Mr Sumption hinted, a decision in favour of Aneco will upset the whole course of business at Lloyds.

I would hold that the Superhulls Cover case was correctly decided, and that the present case is indistinguishable. In agreement with my noble and learned friend Lord Steyn I would dismiss the appeal.

The Shape of the Appeal

The central issue in this case is not one of high legal principle but an evaluative one involving matters of fact and degree. This would not have been fully apparent when the Appeal Committee granted leave to appeal. The broad question is whether London reinsurance brokers, who were in breach of duty to a Bermudian reinsurance company, are liable only for the reinsurance cover which the company lost ($11m), or for the total losses which the company suffered on the transaction ($35m). This in turn depends on an assessment whether on the facts of the case it is governed by the "scope of the duty" principle applied by the House in Banque Bruxelles Lambert SA v Eagle Star Insurnce Co Ltd [1997] AC 191 , known as South Australia Asset Management Corporation v York Montague Ltd (" SAAMCO ") or whether the brokers had undertaken or assumed a duty to advise the company as to what course of action they should take.

Mr Bullen was the underwriter of four syndicates at Lloyd's which wrote marine excess of loss business, i.e. they reinsured losses of other marine insurers so far as those losses exceeded a particular level. In the autumn of 1998 Mr Bullen discussed with Mr Forster, an experienced broker with Johnson and Higgins Ltd, the idea of a proportional reinsurance of his excess of loss account. Mr Forster drafted a treaty ("the Bullen treaty") on a basis which he believed would be attractive to the Bullen Syndicates. He had identified Aneco Reinsurance Underwriting Ltd (now in liquidation) as potential reinsurers of the Bullen treaty. In truth what was contemplated was a retrocession of this treaty but I will adopt the description of it as reinsurance used during the proceedings.

On 15 November 1988 Mr Forster (acting as broker on behalf of the Bullen Syndicates) wrote to Mr Crawley of Aneco offering Aneco a share in the Bullen treaty. From the perspective of Aneco this was inwards business. On 18 November 1988 Mr Forster wrote again to Mr Crawley indicating that Mr Bullen had asked Aneco to consider four units of the Bullen treaty. At the same time Mr Forster suggested that Aneco should purchase excess of loss protection, which he envisaged would be available at between 30-40% of the net premium income.

On 30 November 1988 a meeting took place at Lloyd's to discuss the proposal. Mr Crawley, Mr Bullen and Mr Forster attended. Mr Crawley, as Aneco's underwriter, was willing, subject to Johnson and Higgins being able to obtain satisfactory excess of loss protection, to subscribe to 3-3 ½ units of the Bullen treaty on the basis of an estimated premium income of US$400,000-US$450,000 per unit. On behalf of the brokers Mr Forster indicated that he would obtain quotations for excess of loss protection for Aneco. Mr Crawley made it clear to Mr Bullen and to Mr Forster that Aneco's willingness to participate on the lines discussed was subject to the brokers being able to obtain satisfactory excess of loss protection for Aneco. Mr Crawley confirmed this by facsimile dated 5 December 1988 as follows:

On 5 December 1988 Mr Forster, acting as a broker on behalf of Aneco, approached Mr King, a prominent underwriter at Lloyds in the marine excess loss market. Mr Forster divided the reinsurance into six layers, and proceeded to obtain quotes for all six layers from Mr King. Mr Forster passed these quotes on to Mr Crawley on 5 December 1988 (as to the first four layers) and on the next day (as to the top two layers). On Mr Crawley's instructions Mr Forster went back to Mr King and obtained some modification to and improvement of the quotes. On 7 December 1988 Mr Crawley accepted the latest and most favourable of Mr King's quotes, and instructed Mr Forster to go ahead and obtain the reinsurance for which he had obtained the quotes. That involved going back to Mr King with a firm order, and then going round the market looking for other reinsurers to follow Mr King's lead. On 30 December 1988 Mr Forster told Mr Crawley that he had got 100% subscriptions on three of the six layers and was very close to getting 100% on the remaining three layers. On that basis, Mr Crawley wrote the Bullen treaty. Mr Forster completed the slips for 100% subscription in the next few days.

The Bullen treaty was a "fac/oblig" treaty i.e. Mr Bullen had the right to choose which if any risks he would cede to his reinsurers, and the reinsurers were obliged to accept the risks so ceded. It was faculative as far as Bullen was concerned, but obligatory as far as his reinsurers (including Aneco) were concerned. It was a proportional treaty, i.e. for each risk ceded Bullen ceded a proportion of his premium equivalent to a proportion of the risk. The outward reinsurance protection of Aneco as ultimately placed was in respect of Aneco's marine excess of loss account. It was on an excess of loss basis and was for $7.8m excess of $200,000 in six layers.

Unfortunately Mr Forster negligently failed properly to present the risk to Aneco's reinsurers, some of whom subsequently avoided the policies as they were entitled to do. Euphemistically Mr Forster represented to Aneco's reinsurers that the Bullen treaty was a quota share treaty when it was in fact a fac/oblig treaty. The difference is that a quota share treaty is not facultative as far as the reassured (a person in the position of Bullen) is concerned: he must cede a set proportion of every risk which falls within the limits of the contract, so that everything which meets those criteria is automatically ceded. By contrast fac/oblig treaties are plainly open to abuse. The reassured is able to put onto his reinsurer the least attractive pieces of qualifying business in his book, while keeping what he considers to be the best business for himself. A reinsurer will tend only to reinsure another underwriter on fac/oblig terms if he has considerable trust in the way that his reassured will use it. It is common ground, now, that Mr King would not have agreed to lead the reinsurance of a fac/oblig treaty, and that on a proper presentation of the risk, it would have been impossible to get enough underwriters to subscribe the reinsurance slip, so that the reinsurance that Mr Crawley desired was never available in the market. If Mr Forster had made the enquiries, presentation and disclosure that he should have made, he would have discovered that the outwards reinsurance cover on which Mr Crawley to his knowledge relied from the start was never available. In the event, Aneco suffered a loss on the Bullen treaty of more than $35m, of which they would have recovered $11m from their reinsurers if the reinsurance which Aneco had asked for and which Johnson and Higgins claimed to have obtained had been effective.

The brokers received the usual 3% brokerage under the Bullen treaty and 10% in respect of the six excess of loss contracts.

Aneco sued the brokers in negligence. Aneco formulated its claim for damages on two alternative bases. Its primary case was a claim for all losses which it had in fact suffered by entering into the reinsurance of the Bullen treaty. Aneco put forward this claim on the basis that the brokers had wrongly advised them that the reinsurance was available in the market, and that this advice led them to enter into the Bullen treaty. An indispensable part of this way of putting the claim was that in truth alternative security was never available. The secondary case of Aneco was a claim for all the sums which would have been payable under the outwards reinsurance if it had been in place.

After a trial lasting 20 days in the Commercial Court Creswell J found that there had been non disclosure of material facts and that the brokers had been negligent and had been in breach of their duty to exercise reasonable skill and care in placing the reinsurance on Aneco's behalf: Aneco Reinsurance Underwriting Ltd v Johnson & Higgins Ltd [1998] 1 Lloyds Rep 565. After dealing meticulously with the issues on the merits, the judge turned to the question of fact whether alternative security would have been available at a broadly similar price if there had been a fair presentation of the risk. The judge found that if the risk had been fairly presented, it would have been possible to find reinsurance on terms not very different from those obtained by Mr Forster. On this view of the facts Mr Forster's advice that reinsurance was available in the market was not wrong. The foundation for Aneco's larger claim of $35m was therefore not established. The judge accordingly found that the correct measure of damages was US$10,897,752, being the value of the reinsurance cover which had been avoided by the underwriters, on the basis that this sum was the amount for which the brokers had negligently advised that Aneco had been reinsured.

Aneco appealed against the dismissal of their primary claim. Aneco challenged the judge's finding of fact on the availability of the alternative security. The Court of Appeal took into account expert evidence adduced before the judge, the materiality of the underlying business, and contemporaneous marketing sheets which effectively covered the whole of the London market where this sort of reinsurance might be placed. On this issue the Court of Appeal unanimously concluded that such reinsurance was either not available at all or not available on acceptable terms: Aneco Reinsurance Underwriting Ltd v Johnson & Higgins Ltd [2000] 1 All ER (Comm) 129. In the leading judgment Evans LJ dealt with this point paras 55-72, at pp. 148G-154D. Ward LJ expressed agreement with this part of the judgment of Evans LJ at p 164J. Aldous LJ accepted that alternative security was not available; at p 161B. This was a conclusion of major importance. It became the factual foundation of the argument of Aneco in the Court of Appeal that its primary claim ought to succeed. The issue of the correct measure was, however, still vigorously contested.

The Court of Appeal was divided on the issue. The majority (Evans and Ward LJJ) found that the correct measure was Aneco's loss flowing from the Bullen treaty, i.e. $35m. Aldous LJ concluded that the correct measure was the loss flowing from Aneco's assumption that the appropriate reinsurance had been effectively placed, i.e. $11m. Evans and Ward LJJ found that on the facts "the scope of duty" principle stated by the House in SAAMCO was inapplicable: [1997] AC 191 . The majority concluded that Mr Forster had assumed the duty of advising Aneco what course to adopt, and that he had wrongly advised that reinsurance cover was available in the market, thereby causing Aneco to enter into the Bullen treaty. Aldous LJ came to the opposite conclusion. He concluded that the duty of the brokers to Aneco was narrow, viz merely "to provide information and to place the reinsurance." at p 162E.

Before the House the Court of Appeal's conclusions on the non-availability of alternative reinsurance cover was accepted. It follows that Mr Forster's advice to Aneco that reinsurance cover was available in the market was wrong and was negligently given.

In these circumstances the principal question is: Is the correct measure of damages all of Aneco's losses under the Bullen treaty or is the correct measure equal to the recovery which Aneco would have made under the reinsurance contracts but was unable to make to the extent that those have been avoided?

Before I proceed to consider the principal question of fact it is necessary to refer to a preliminary argument by counsel on behalf of the brokers to the effect that on the state of the pleadings and the manner in which the case was conducted at trial it was not open to the majority to decide the case as they did.

Aneco put forward its case on the alternative bases which I have described. The primary case was based on the allegation that the brokers negligently advised on the availability of reinsurance in the market. The allegation was squarely pleaded. It was explored and tested in evidence. It was the cornerstone of Aneco's case throughout before Creswell J and before the Court of Appeal. Evans and Ward LJJ certainly did not think they were engaged on a frolic of their own. Moreover, Aldous LJ did not suggest in his dissenting judgment that the majority was proceeding on a basis not open to them. The technicality of the arguments of counsel for the brokers on this point reminded me of an observation of Holmes J in Braithwaite v Hall, 168 Mass 38. He said, at p 46: "Nowadays we do not require pleadings to guard against all the distortions of perverse ingenuity". The preliminary objection must be rejected.

Given that this case can be decided by applying settled principles, I do not propose to examine any problems which do not arise. Nevertheless, I must set out, without examination, the contours of established doctrine.

In the leading judgment in SAAMCO [1997] AC 191 Lord Hoffmann illustrated "the scope of duty" concept with an example. He said, at p 213D:

There was an interesting debate during the hearing of the appeal on the validity or otherwise of Lord Hoffmann's paradox. In a closely reasoned case note Professor Jane Stapleton has argued that conceptually there is no paradox: Negligent Valuers and Falls in the Property Market (1997) 113 LQR 1. In the interests of brevity, and doing less than justice to the full rigour of the argument (at pp 3-5), I cite only one passage, at p 5:

The background, relevant exchanges, documentation and oral evidence is set in great detail in the judgment of Evans LJ. While his conclusions are in issue, there can be no valid criticism of the remainder of his judgment. In these circumstances, and in a case involving simply the factual categorisation of the case, it would serve no purpose for me to cover the same ground again.

The starting point of the enquiry is not in doubt. If the brokers had carefully performed their duty to report on the availability of reinsurance they would inevitably have reported to Aneco that reinsurance cover was not available in the market. In that event, Aneco would not have entered into the Bullen treaty. The issue is simply: Did the brokers undertake a duty to advise Aneco as to what course of action they should undertake? The argument on behalf of the brokers was that they only undertook a duty to exercise reasonable care to obtain the reinsurance ordered and to report the result of their endeavours. Evans LJ, who has vast experience of the way in which reinsurance business is transacted, gave the answer to this argument. He observed that it would be "highly artificial to derive from the evidence any suggestion that Mr Forster was not advising Mr Crawley what course to take": at para 78, p 155H. There was ample material to support this conclusion. Only one item of evidence need be cited. In his evidence Mr Forster accepted that the brokers were advising Mr Crawley as to what reinsurance was available and as to the state of the market. He said:

The contrary reasoning of Aldous LJ, and the arguments of counsel for the brokers, are in my view based on an artificial and unrealistic distinction between reporting on the availability of reinsurance in the market and reporting on the assessment of the market on the risks inherent in the Bullen treaty. These are two sides of the same thing: they are inextricably intertwined. If the brokers had advised Aneco of the non availability of reinsurance cover in the market, that would inevitably have revealed to Aneco the current market assessment of the risk. There was no other credible reason for reinsurance being unavailable. On the evidence Evans LJ was correct to conclude that the brokers breach of duty was their negligent advice "with regard to the availability of reinsurance (retrocession) and therefore on the current market assessment of the risk". In my view the conclusion of Evans LJ is supported by the commercial realities and inherent probabilities in the relationship between broker and reinsured revealed by the documentary and oral evidence.

Counsel for brokers placed great weight on the argument that the conclusion of the majority places a broker, circumstanced in a dual capacity as Mr Forster was, in an invidious position. He argued that the difficulty lies in holding that the broker, who owes a duty to the insured to place the insurance, is simultaneously under a duty of care to the insurer to provide advice to him on whether or not to write the insurance at all. The answer is clear. Any problem of the brokers arising from the performance of their dual functions in this case was entirely of their own making. It cannot divert the House from arriving at the inescapable conclusion on the facts that the brokers assumed a duty to advise Aneco as to what course to take. In the result the brokers' failure to advise that reinsurance was unavailable in the market resulted in a recoverable loss of $35m. The width of the duty assumed by the brokers is determinative of this being the correct measure of damages.

Ultimately, on matters of fact the question is on which side of the line drawn in SAAMCO the present case falls. In my view the majority of the Court of Appeal came to the correct conclusion. The brokers were fortunate in obtaining leave to appeal to the House on what turned out to be issues of fact. Nevertheless, it was necessary to give the closest attention to all the arguments deployed during a three day hearing. Having done so my view is that the arguments of the brokers must be rejected.

I would dismiss the appeal with costs

This is another case which is concerned with the extent of a defendant's liability for the consequences of a breach of duty on his part which has resulted in the plaintiffs entering into a loss-making transaction. It is established by the decisions of this House in Banque Keyser Ullmann SA v Skandia (UK) Insurance Co Ltd [1991] 2 AC 249 ("Skandia") and Banque Bruxelles Lambert SA v Eagle Star Insurance Co Ltd [1997] AC 191 , known as South Australia Asset Management Corporation v York Montague Ltd ("SAAMCO"), that the plaintiff is not entitled to recover damages for the full extent of his loss merely because the defendant knew that he would not have entered into the transaction but for his own breach of duty. There must be a sufficient causal connection between the particular feature of the transaction which occasioned the loss and the subject matter of the defendant's duty of care. The amount of damages which the plaintiff is entitled to recover is limited to the amount of the loss which is attributable to the defendant's breach of duty, and this depends upon the scope of the duty in question. None of this is in dispute; the issue in the present case turns primarily on the correct identification of the scope of the defendant's duty, which is a question of fact.

Stripped to the bare essentials, and omitting for the moment the critical features on which the respondent Aneco relies, the facts are these. In 1988 Aneco was minded to enter into a proportional treaty ("the Bullen treaty") under which it would participate in the excess of loss account of certain Lloyd's marine syndicates. This was a transaction which in the events which happened was to cause Aneco a loss (in round figures) of some $35m. Aneco was not prepared to enter into the Bullen treaty unless it obtained excess of loss reinsurance for its own marine excess of loss account, including but not limited to losses arising from the Bullen treaty. Aneco employed the appellants, the brokers who had acted for the Lloyd's syndicates in obtaining Aneco as a counterparty to the Bullen treaty, to obtain the reinsurance for Aneco. Aneco made it plain to the appellants that it would not enter into the Bullen treaty unless it obtain the excess of loss reinsurance which it required. The appellants obtained it and Aneco duly entered into the Bullen treaty. Unfortunately the appellants had failed to present the risk fairly to the reinsurers, and many of them have repudiated liability. Had they not done so, Aneco's losses under the Bullen treaty would have amounted (in round figures) to $24m. As it is, they are some $35m.

Thus Aneco entered into two separate but interdependent transactions: (i) the Bullen treaty by which it reinsured the syndicates excess of loss business; and (ii) contracts of outward reinsurance (strictly retrocession) by which it obtained excess of loss reinsurance for its own marine excess of loss account, including but not limited to losses under the Bullen treaty. The appellants were Mr Bullen's brokers in the first transaction and Aneco's brokers in the second.

If these were the only facts, Aneco concedes that it could not recover damages in excess of $11m. The concession is plainly correct. The loss of $11m was due to the avoidance of most of the contracts of reinsurance. The loss of the further $24m was caused by Mr Crawley's decision to enter into the Bullen treaty, albeit with the benefit of excess of loss reinsurance, and not by the appellants' failure to obtain valid reinsurance. It had nothing to do with the absence of effective excess of loss protection. Aneco would have suffered this loss even if it had all the reinsurance protection it asked for. Aneco's claim has never been put on the basis that the appellants bore any responsibility for advising Aneco as to the amount of reinsurance it needed. That would involve an exercise of underwriting judgment; the appellants were brokers, not underwriters.

The additional fact on which Aneco relies in order to establish its claim to recover in respect of the retained loss of $24 million is that, contrary to the information which the appellants gave Aneco, the reinsurance which Aneco required was not available in the market; it would have been declined by any underwriter to whom the risk was fairly presented. If the appellants had presented the risk fairly, as they ought to have done, so the argument runs, they would have discovered that the market viewed the Bullen treaty with marked disfavour. If this had been reported to Aneco, it would then not have entered into the treaty.

In the Court of Appeal [2000] 1 All ER (Comm) 129 Evans LJ (with whom Ward LJ agreed) concluded that the appellants, as Aneco's brokers to obtain reinsurance, had accepted responsibility for advising Aneco, not merely with regard to the availability of reinsurance, but with regard to the market's assessment of the risks inherent in the Bullen treaty. If this conclusion is correct, then the appellants would not only have reported the market's unfavourable assessment of the Bullen treaty to Aneco as a matter of fact, but would have been bound to do so as a matter of legal obligation.

My Lords, the fact that the reinsurance which Aneco required was not available in the market is no longer in dispute, and if the conclusion of the majority of the Court of Appeal as to the scope of the duty undertaken by the appellants were correct I would affirm its decision. The additional loss of $24m would still have been occasioned by Aneco's decision to enter into the Bullen treaty with only partial reinsurance protection, and not by the appellants' failure to obtain effective reinsurance. But Aneco's decision would be attributable at least in part to the appellants' breach of duty in failing to ascertain and advise Aneco of the market's adverse assessment of the risks inherent in its decision to participate in the Bullen treaty.

There are, therefore, two questions for decision. The first is whether the conclusion is correct: did the appellants, as Aneco's brokers, undertake a duty to report, not only on the availability of the reinsurance which they were instructed to obtain, but on the market's assessment of the risks inherent in the Bullen treaty? That is a question of fact, not law. If they did not undertake such a duty, a second question arises: what is the measure of damages for a broker's advice that reinsurance is available when it is not? In particular, if he advises that reinsurance has been obtained (and therefore implicitly that it is available) is his liability greater if the reinsurance is in fact not available in the market than if it is? This is a question of law.

Before setting out the facts in more detail, I shall briefly set out the current state of the law.

The law has never imposed liability for all the consequences of a defendant's negligence. It has formulated general rules to restrict the scope of liability within acceptable limits by reference to concepts such as foreseeability and remoteness of damage. In traditional cases of negligent conduct which causes physical injury, it has seldom been found necessary to place limits on the scope of the duty of care or the extent of the defendant's liability for the foreseeable consequences of his acts. Claims for damages for economic loss which is the result of negligent statements or advice, however, are very different. There is a potential for foreseeable but indeterminate and possibly ruinous loss by a large and indeterminate class of plaintiffs. Foreseeability of reliance alone is not a sufficient limiting factor.

One response has been to limit the scope of the duty of care by reference to a test of "proximity". Another has been to decline to admit new categories of liability unless it is "fair, just and reasonable" to do so. Given the existence of a duty of care owed by the defendant to the plaintiff, however, it has generally been assumed that the defendant is liable for all the foreseeable consequences of his breach of duty.

This assumption led the defendants to concede the quantum of damages in Youell v Bland Welch & Co Ltd (No 2) (The " Superhulls Cover " case) [1990] 2 Lloyd's Rep 431. The facts were not materially distinguishable from those of the present case. The plaintiffs wrote insurance contracts in reliance on the advice of their brokers that reinsurance was available on appropriate terms when it was not. Since the plaintiffs would not have entered into the insurance contracts but for the advice, it was assumed that, if the advice was negligent, the brokers were liable for the full amount of the loss on the insurance contracts even though their advice was referable exclusively to the reinsurance.

In Caparo Industries plc v Dickman [1990] 2 AC 605 , however, this assumption was shown to be false. The House held that it is not sufficient for the plaintiff to prove that the defendant was in breach of a duty of care owed to him. He must also show that the particular loss fell within the scope of the duty. Lord Bridge of Harwich said, at p. 627:

At p 629 Lord Roskill dealt with liability for the provision of information in terms which are particularly apposite in the present case:

The fallacy which underlay the concession in the Superhulls Cover case was exposed in Bank Keyser Ullmann SA v Skandia (UK) Insurance Co Ltd [1991] 2 AC 249, though not until the case reached the House. A syndicate of banks which was proposing to make a secured loan to a group of companies instructed a broker to arrange credit insurance. The policies excluded liability for loss by reason of fraud on the part of the borrowers. It was a condition of the loan that the banks would advance the moneys only when all the policies were in place. The broker fabricated cover notes and falsely certified that all the policies were in place when only some of them were. The banks duly advanced the moneys. The borrowing companies defaulted, the man behind them turned out to be a swindler who had embezzled the companies' assets, the security proved to be worthless, and the insurers were able to rely on the fraud exemption to any claim under the policies. The banks sued the broker. They proved that they would not have lent the money if they had known that he had deceived them. The "but for" test of causation was satisfied. The consequence (that the banks would lend the money) was foreseeable; indeed it was intended. But the House unanimously dismissed the banks' claim. The loss was caused by the borrowers' fraud coupled with the fact that this was not an insured risk. The brokers' certificate that the insurance was in place did not cause the loss, which would have been exactly the same if all the insurance had been in place as the broker had certified.

The Superhulls Cover case and Skandia were both concerned with separate but interdependent transactions each leading to a different but easily quantifiable loss. But the principle was taken a stage further in SAAMCO . There lenders made mortgage advances on the security of land in reliance on valuations which were subsequently found to be negligent. In each case there was a single transaction and the loss, being the difference between the amount advanced and the amount recovered, was indivisible. But a large part of the loss was due to a fall in the property market between the date of the advance and the date the security was realised. The valuers were not responsible for this save in the sense that, if they had given a correct valuation, the transaction would not have gone ahead at all and the mortgagee would have suffered no loss. The "but for" test was satisfied. Eliminating the loss for which the valuers were not responsible made it necessary to undertake a mathematical exercise, and this required the proper measure of damages to be identified. The courts below simply applied the "but for" test and asked what would have happened if the valuers had performed their duty and valued the properties correctly. In that event none of the transactions would have taken place. The mortgagees were awarded the whole of their loss. The House allowed the valuers' appeals: [1997]AC 191.

Lord Hoffmann gave the only reasoned speech. He applied the principle which had emerged in Caparo v Dickman and Skandia . He observed that a person who is under a duty to take reasonable care to provide information on which someone else will rely in deciding whether to take a course of action is, if negligent, not generally responsible for all the consequences of that course of action. He is responsible only for the consequences of the information being wrong. He is not responsible for losses which would still have occurred even if the information had been correct. It was necessary to exclude from the computation of loss that part of the overall loss which would still have been sustained if the facts had been as represented. This explained why no part of the loss was irrecoverable in Skandia ; it was uninsured loss and would have been irrecoverable even if the cover had been fully in place as the broker had certified. In the Superhulls Cover case the same principle would have excluded that part of the loss which would have been retained by the plaintiff even if the desired reinsurance cover had been obtained. In SAAMCO itself it excluded that part of the loss which would have been sustained even if the property had been worth the amount at which it had been valued.

Lord Hoffmann drew a distinction between "a duty to provide information for the purpose of enabling someone else to decide upon a course of action and a duty to advise someone as to what course of action he should take." [1997] AC 191 , 214 This has been widely misunderstood. Lord Hoffmann was not distinguishing between a duty to provide information and a duty to give advice. That is a distinction without a difference, for the terms are interchangeable. He was distinguishing between a duty to provide particular information or advice on request and a duty to advise generally when it is left to the adviser to decide what matters he should consider. Even where the defendant assumes responsibility for advising generally "whether or not a course of action should be taken" it is still necessary to identify the particular course of action in question. Where the question is whether to enter into a particular transaction, it is necessary to identify the relevant transaction, for the defendant is not responsible for loss arising from any other transaction. The warning given by Lord Roskill in Caparo v Dickman to which I have already referred is particularly apposite when there are separate but interdependent transactions as in Skandia and the Superhulls Cover cases. In the former, it would have made no difference to the result if the broker had advised generally in relation to the credit insurance, so long as he had left the decision whether to accept the fraud exclusion to his principals, since the loss arose on the transaction of loan. In the latter, the concession would be correct only if the broker had assumed responsibility for advising what action should be taken in relation to the underlying insurance. It is never enough to say: the defendant was responsible for advising the plaintiff what action he should take. It is necessary to ask: in relation to what? His liability is limited to losses arising from the particular transaction in relation to which the advice was given. Where the defendant gave professional advice, his profession will usually supply the answer.

In Nyekredit Mortgage Bank Plc v Edward Erdman Group Ltd [1997] AC 191 ("Nyekredit"), which was heard at the same time as SAAMCO , the valuers were specifically asked to advise not only on the value of the property, which was a development site, but also on the projected rental value of the completed development and its lettabilty in the open market. The advice which the valuers gave under each of the specified heads was wrong. The lenders claimed that if the valuers had given them the correct figures they would have appreciated, not only that the security was insufficient, but that the development project was not viable and default was inevitable.

The House rejected this as a distinguishing factor which imposed a different liability from that in the other cases. There were two reasons for this. First, the argument was based on what would have happened if the valuers had provided accurate information. As Lord Hoffmann observed, at p 223, that is not the basis of their liability. They are not liable for the loss which is due to the fact that they failed to give correct valuations, but for the loss which is due to the fact that the true values were not as they represented them to be. Secondly, the detailed figures had been requested merely as component elements in the valuation of the security. In other words, the valuers had been asked to advise on the value of the security and not on the viability of the project. As always, in Lord Roskill's words in Caparo , [1990] 2 AC 605 , 628 it is necessary to determine "for what purposes and in what circumstances the information in question is to be given".

The SAAMCO principle is not confined to valuers. Lord Hoffmann gave an imaginary example of a doctor. In Bristol and West Building Society v Fancy & Jackson [1997] 4 All ER 582 it was applied to a number of cases against solicitors who acted for proposed mortgage lenders. In some cases they were held responsible for the whole of the loss on the transaction; in others they were not. In Fancy & Jackson itself they were held not to be responsible for any part of the loss. They ought to have reported that they did not have an official search certificate. If they had done so, the society would not have authorised the advance. But the loss which the society suffered was not due to the absence of a search certificate, for the title was clear. The society obtained what it intended to obtain when it decide to enter into the transaction. The loss which occurred would have occurred even if the solicitor was in possession of an official search certificate, as he said he was.

The law can be summarised as follows:

I shall return to points (2) and (3) when considering the second of the two questions to which I have referred: whether the measure of damages for the appellants' advice that the desired reinsurance had been obtained (and therefore implicitly that it was available) is greater if the reinsurance was in fact not available in the market than if it was.

The appellants Johnson & Higgins Ltd are insurance and reinsurance brokers. Mr Forster was a broker in their employ. Mr Bullen was the underwriter of four syndicates at Lloyd's which wrote marine excess of loss business. This is a form of reinsurance which is intended to cover particularly catastrophic losses due to hurricanes, oil rig disasters and the like.

On 15 November 1988, acting as broker to Mr Bullen, Mr Forster wrote to Mr Crawley, Aneco's underwriter, offering Aneco a share of the syndicates' marine excess of loss account. On 18 November Mr Forster followed this up by a letter giving more information about the account, and adding

What was envisaged was a proportional treaty under which the Bullen syndicates as insurer/reinsured and Aneco as reinsurer agreed to share risks within the limits of the treaty. Such treaties are of two kinds: quota share and facultative/obligatory (known as "fac/oblig"). Under a quota share treaty, the insurer is obliged to cede to the treaty a fixed proportion of every risk which falls within the limits of the treaty. Under a fac/oblig treaty the insurer has a choice whether to cede any given risk to the treaty. He cannot cede it unless it falls within the limits of the treaty, but he is not obliged to cede it if it does. The reinsurer has no choice; he cannot insist on a risk being ceded, and cannot refuse to accept his share of a ceded risk.

Fac/oblig treaties are naturally less attractive to reinsurers than quota share treaties. They are subject to the obvious risk that the insurer will retain good business for his own account and cede poor business to the treaty. There is, or at least is assumed to be, no obligation of good faith on the part of the ceding party when exercising his discretion whether to cede or retain a risk. The only constraint upon him is that he must exercise some restraint if he wishes to maintain a good reputation in the market and any hope of doing future business with existing and prospective reinsurers.

Mr Crawley was aware that the Bullen treaty was a fac/oblig treaty. As Evans LJ accepted [2000] 1 All ER (Comm) 129, 156, para 79, it was necessary for Mr Crawley to form his own view of the prospects for business of the kind which the Bullen treaty represented. The availability of sufficient excess of loss protection was a major factor, but there were others and Mr Forster was not concerned with them. Mr Crawley had decided to participate in the London market excess of loss business because he believed (correctly) that premium rates were rising as a result of the Piper Alpha disaster and (incorrectly) that the likelihood of another catastrophe coming so soon afterwards was remote. The Bullen treaty marked Mr Crawley's first foray into this market, but it was to be expected that he would write further business from time to time. He made his own enquiries about Mr Bullen and his reputation in the market and came to London in late November 1988 in order to meet Mr Bullen and size him up. He made his own assessment of Mr Bullen's reasons for wanting the treaty and what use he was likely to make of it. He negotiated the finer points of the treaty, such as overriding commission, directly with Mr Bullen. He was not deterred by the fact that the Bullen treaty was a fac/oblig treaty. Aneco's potential exposure under the Bullen treaty was not predictable with any degree of certainty. There were several reasons for this; it is sufficient to mention two of them. First, although there was a limit on the size of any one contract which Mr Bullen could cede to the treaty, there was no limit on the number of contracts which he could cede. Secondly, Aneco's exposure depended on the average rate on line of the business which was ceded to the treaty. (The rate on line is the proportion which the premium payable bears to the cover. The lower the rate on line for a given premium the greater the exposure.) Under the treaty Mr Bullen could cede business which had an original rate on line of between 2% and 20%. Mr Crawley estimated that the average rate on line of business ceded to the treaty would be 10%, but Mr Bullen was not bound to maintain any particular average rate on line. In the event his rate on line was very much lower than 10%. The combination of these two factors meant that Aneco's exposure under the Bullen treaty was far higher than Mr Crawley had envisaged.

Mr Crawley was willing to enter into the Bullen treaty without any limitation on the number of contracts which Mr Bullen could cede, without any obligation on Mr Bullen's part to maintain a minimum average rate on line, and without any limit on Aneco's aggregate exposure. He was, however, not willing to enter into the treaty without reinsurance protection, even though, for the reasons already stated, he could not predict how much cover was needed.

Mr Crawley applied a rule of thumb of his own that he should buy reinsurance for 60-70% of his total aggregate exposure. This involved his estimating the amount of the aggregate exposure. He did this by taking an estimated average rate on line of 10%, even though Mr Bullen had not bound himself to such an average and Mr Crawley had not stipulated for it as he could have done. He then multiplied the exposure produced by that rate on line by the estimated premium income, even though there was no contractual ceiling on this either. Mr Crawley did not seek advice on the adequacy of the reinsurance protection which his calculations indicated.

On 30 November, at a meeting in the Captain's Room at Lloyd's attended by Mr Crawley, Mr Bullen and Mr Forster, Mr Crawley agreed to participate in the Bullen treaty on the basis of an estimated (but not guaranteed) premium income. At the meeting Mr Crawley made it clear that Aneco's willingness to participate on the lines discussed was subject to Mr Forster being able to obtain the excess of loss protection for Aneco which Mr Crawley wanted. Mr Crawley confirmed this by fax dated 5 December to Mr Forster which contained the statement

Mr Forster then approached Mr King, a leading underwriter at Lloyd's and asked him to quote for six layers of excess of loss protection for Aneco's whole marine excess of loss account. On 5 December he sent a fax to Mr Crawley. The fax was in two parts, corresponding to the two roles in which Mr Forster was acting. The first, headed "N.T.Bullen - Special Priority Treaty", dealt with revisions to the proposed Bullen treaty. The second, headed "Your Excess of Loss Protections", contained the quotations he had obtained from Mr King for the four bottom layers of cover. He added that he would expect to obtain the further cover required and that

On 7 December Mr Forster telexed Mr Crawley again. Ominously, perhaps, he reported that Mr Bullen was unable to provide accurate information regarding his aggregates. Mr Forster added:

Mr Crawley accepted Mr Forster's advice. By telex of the same day he agreed that Aneco would participate in the Bullen treaty to the extent proposed and confirmed a firm order to place the reinsurance to the extent and on the terms indicated.

Mr Foster returned to Mr King who subscribed lines on all six layers of reinsurance. Mr Foster then broked the six layers to the following market, both at Lloyd's and the London and overseas companies markets. By 30 December the placing of the six layers was all but complete and Mr Forster so informed Aneco by telex of the same date. On receiving this information Aneco signed the slip for its participation in the Bullen treaty. By 19 January 1989 all six layers were fully subscribed. Aneco relies on the telex of 30 December as a further representation that reinsurance, if properly broked, was available. The actual representation was that such reinsurance had been or would shortly be obtained; but it is implicit in such a representation that the reinsurance is available.

Again, however, the context is critical. Mr Crawley had not asked Mr Forster to report on the market's assessment of the Bullen treaty. He did not ask to see Mr Forster's marketing sheets which contained the notes of the responses of the underwriters who declined to subscribe for any of the six layers ("no way"; "not now"; "next year"; "thanks but"; "no new XL"; "not this layer"; "needs like a hole in the head"); and Mr Forster did not show them to him. Had Mr Crawley asked for them he would have discovered that market sentiment generally was distinctly unfavourable. Many underwriters did not want to write any more marine excess of loss business, irrespective of the quality of the business or the identity of the syndicate's underwriter. Some underwriters did not like the quality of Mr Bullen's business; others did not relish having Aneco, a foreign company and unknown in the marine market, as a counterparty. Significantly, those underwriters who realised or suspected that the Bullen treaty was a fac/oblig treaty declined participation on that account.

Aneco subsequently became insolvent and is now in liquidation. It suffered a loss on the Bullen treaty of more than $35m. It would have suffered the greater part of this loss even if the reinsurance which Aneco had ordered and the appellants had obtained had been effective. The reason for the loss is not controversial. It had nothing to do with the broking of the reinsurance for which the appellants were responsible. It arose because Mr Bullen ceded much more business and at lower rates on line than Mr Crawley expected, and because it was a heavily loss-making business.

The Lloyd's underwriters repudiated liability under the reinsurance contracts for non-disclosure, and their right to do so was upheld by a Lloyd's tribunal chaired by Sir Michael Kerr. Some but not all of the companies which subscribed to the reinsurance contracts have also avoided them on the same grounds. Nine of the reinsurers have affirmed the contracts and have paid claims amounting (in round figures) to a total of $4.6m. Aneco has given credit for this sum in formulating its claim.

Aneco sued the appellants for negligence. It pleaded that the appellants had acted as Aneco's brokers

The allegation that the appellants negligently advised on the availability of reinsurance was clearly pleaded. Aneco did not allege that the appellants had any duty to advise it (i) on what course of action to take in relation to the Bullen treaty or the wisdom of its decision to participate in the Bullen treaty (whether as a fac/oblig treaty or otherwise); (ii) as to the amount of the reinsurance protection which it should obtain (both of which would have called for an exercise of underwriting judgment); or (iii) as to the market assessment of the risks inherent in the Bullen treaty. It did not allege any duty on the part of the appellants to report the comments made by underwriters who declined the business or their reasons for doing so.

In his evidence Mr Crawley did say that he sought Mr Forster's advice on the level of reinsurance that Aneco should obtain and that Mr Forster recommended buying up to 10 times premium income. But Aneco's claim to the $24m was neither pleaded nor put on the basis that Mr Forster had given negligent advice that the level of reinsurance protection which Aneco asked for was sufficient. In cross-examination Mr Crawley said that, with hindsight, they had got the amount of vertical cover wrong, but he agreed that at the time he thought that it was adequate. Aneco's claim to the $24m was put fairly and squarely on the simple ground that Aneco would not have entered into the Bullen treaty in the absence of effective reinsurance in the amount which it required, and that Mr Forster knew this.

In his evidence Mr Crawley did not suggest that he was intending to rely on Mr Forster for the market's assessment of the merits of the Bullen teaty. There is only one passage in Mr Crawley's evidence which bears on this. It occurs in his cross-examination:

In cross-examination Mr Forster accepted that he had a responsibility (in counsel's words) "for protecting Aneco's interests with regard to the outwards reinsurance protection " (emphasis added.) In this context he agreed that he was advising Aneco of the reinsurance which was available and the state of the (reinsurance) market at the time. He agreed that he had advised Mr Crawley that reinsurance protection for the layers and at the rates on line indicated could probably be obtained, that he knew that Aneco's willingness to enter into the Bullen treaty was entirely dependent on the availability of reinsurance, and that if he had been unable to obtain the required reinsurance protection "the whole thing would have collapsed."

It was not put to Mr Forster that he had any responsibility to advise Aneco what course of action he should take in relation to the Bullen treaty, or to advise not only of the availability of the reinsurance protection which Aneco required, but of the market's assessment of the risks inherent in the Bullen treaty. Nor was it put to him that he had a duty to report the comments of those underwriters who had declined to subscribe or their reasons for doing so. The only breaches of duty put to him were his failure to present the risk fairly to the reinsurers (with the result that the reinsurance which he was asked to obtain was ineffective) and his advice that such reinsurance was available and had been obtained when it was neither (with the result that Aneco entered into the Bullen treaty and the reinsurance contracts).

Cresswell J [1998] 1 Lloyd's Rep 565 found that, in breach of their duty of care as Aneco's brokers, the appellants were negligent in presenting the risk to the reinsurers by failing to disclose material facts, principally (though not exclusively) that the Bullen treaty was a fac/oblig treaty. On the basis of expert evidence, the judge found that, if the risk had been fairly presented, it would have been possible to find reinsurance on terms not greatly differing from those obtained by Mr Forster. On this view of the facts Mr Forster's advice that reinsurance was available in the market was not incorrect (let alone negligent). The judge awarded damages of $11m as representing the value of the reinsurance cover which was lost by Mr Forster's negligence.

The Court of Appeal [2000] 1 All ER (Comm) 129 unanimously reversed the judge's finding that it would have been possible to obtain reinsurance on comparable terms if the risk had been properly broked. The court found that on the balance of probabilities it would not have been possible for the appellants to find alternative leaders to Mr King or sufficient following underwriters to subscribe to the risk if it had been fairly presented. They would not have been willing to subscribe to an excess of loss protection of Aneco as a fac/oblig reinsurer of the Bullen Syndicates or would not have been willing to do so on terms which would have been commercially acceptable to Aneco. In arriving at this conclusion the Court of Appeal took into account the (differing) expert evidence dealing with alternative security, the materiality of the underlying nature of the business, and the response of potential excess of loss underwriters recorded in Mr Forster's marketing sheets which effectively covered the whole of the London market where reinsurance of the kind in question might be placed. There is no appeal from this part of the Court of Appeal's judgment. It follows that Mr Forster's advice that such reinsurance was available must be taken to be incorrect. It does not, of course, follow that it must be taken to have been negligent, but that is another story.

As I have already explained, the Court of Appeal by a majority (Evans and Ward LJJ, Aldous LJ dissenting) held that Mr Forster had undertaken responsibility for advising Aneco, not only with regard to the availability of reinsurance, but with regard to the market's assessment of the risks inherent in the Bullen treaty. Evans LJ held, at p 156, para 82 of his judgment, that the fact that no reinsurance cover was available in the market introduced an additional breach of duty by the appellants. They were liable not only for failing to obtain effective cover on the terms which they reported to Aneco, but also for failing to report that no cover could be obtained. He said that this meant that the SAAMCO principle (compensating the plaintiff only for the consequences of the defendant's advice being wrong) failed to provide proper compensation in the present case. He held, at p 156, para 83 that:

It is sufficient for the moment to observe that there is no "therefore". The conclusion in para 84 follows if, but only if, the premise in paras 82 and 83 is correct. If it is then the conclusion of the Court of Appeal is an application of the SAAMCO principle and not (as Evans LJ thought) an exception to it. Aneco is entitled to recover damages for the full amount of the loss which is attributable to the appellants' breach of duty. The greater the scope of their duty, the greater the scope for the attribution of loss to the breach.

Ward LJ agreed with Evans LJ but added some further reasons of his own. He, too, relied on the fact that reinsurance was not available at all, or at least on terms acceptable to Aneco, and that, in order to discharge their duty, the appellants should have told Aneco that the market would not cover the risk. This, of course, is not disputed. But more problematically he continued, at pp 164-165:

In the course of argument before your Lordships counsel for Aneco made two significant but inevitable concessions. First, he accepted that Mr Crawley did not seek to use the market's reaction to the reinsurance layers as "a barometer" of the risks inherent in the Bullen treaty. Secondly, he accepted that even if Mr Forster had discovered that underwriters were almost universally hostile to the Bullen treaty - they "wouldn't touch it with a barge pole" - but had nevertheless been able, albeit with great difficulty, to place the required layers of reinsurance with good and sufficient counterparties and on a fair presentation of the risk, Aneco would have had no complaint. Counsel frankly acknowledged that Mr Forster was under no obligation to report anything beyond the bare fact that the desired reinsurance was available and had been obtained (or not as the case might be). He did not suggest that Mr Forster undertook any responsibility for advising Aneco what course of action to take in relation to the Bullen treaty.

In my opinion these concessions are entirely inconsistent with paragraphs 83 and 84 of Evans LJ's judgment. But they were correctly made. Mr Forster was not bound as a matter of legal obligation to report the unfavourable comments of those underwriters who had declined the risk or their reasons for doing so or to warn Mr Crawley that market sentiment was adverse and that he should reconsider his decision to enter into the Bullen treaty. No such duty was pleaded or made out on the evidence. The appellants did not complain of the breach of any such duty and no such complaint was put to Mr Forster. If it were the ordinary duty of a broker instructed to obtain reinsurance cover in the market, one would have expected Mr Crawley to have asked to see Mr Forster's marketing sheets and Mr Forster to have provided them unasked.

Mr Forster was instructed to obtain reinsurance for Aneco. He had to test the market to find out if it was available and at what rates it could be obtained. He undertook these duties as Aneco's broker, but he undertook them in relation to the reinsurance, not in relation to the Bullen treaty. He had no responsibility for advising or reporting with regard to Mr Crawley's conditional decision to enter into the Bullen treaty. He was a broker, not an underwriter.

By reporting that he had obtained reinsurance Mr Foster implicitly represented that reinsurance was available. He could not avoid doing so. But this did not involve any representation that market sentiment was favourable. No such representation was pleaded or put to Mr Forster. Had Mr Forster reported that reinsurance was not available Mr Crawley would not have entered into the Bullen treaty. But as his own evidence demonstrates, this would not have been because of market sentiment: he would have been disappointed at his inability to do what he conceived to be good business. It would not have mattered to him why reinsurance was unavailable: it would have been enough that reinsurance was not available for whatever reason. That is why the issue cannot be tested by asking whether Mr Crawley would have entered into the Bullen treaty if reinsurance had not been available. It is necessary to consider whether Mr Foster was under a legal obligation to report adverse market sentiment even if he had been able to obtain the desired reinsurance; and what (if any) loss is attributable to the breach of that particular duty. Unless Mr Crawley was intent on using the market as a barometer of the merits of the Bullen treaty, which has been expressly disclaimed, it would seem that no loss is attributable to this particular breach, for Mr Crawley would still have gone ahead despite his knowledge of the market's attitude provided only that effective reinsurance had been obtained.

As I have already mentioned, Evans LJ held, at p 156, para 82 of his judgment that the appellants are liable not only for failing to obtain effective reinsurance, but also for failing to report that no cover could be obtained. Thus far I agree with him. It is at the next step that I respectfully part company from him. He held, para. 83, that the appellants are liable for their failure to report correctly the current market assessment of "the reinsurance risks which Aneco was proposing to undertake", ie. the risks inherent in the Bullen treaty. As Evans LJ rightly observed, those risks were central to Aneco's decision to participate in the Bullen treaty; but it does not follow that (as he found) "Mr Forster took it upon himself to advise Mr Crawley with regard to them". There is simply no basis, in the pleadings or the evidence, for finding that the appellants undertook any such duty, and it is inconsistent with the concessions made by Aneco's counsel in argument before your Lordships. No such duty would have arisen merely from the fact that, as the appellants knew, Aneco's decision to participate in the Bullen treaty was conditional on obtaining reinsurance (and therefore on the reinsurance being available). It makes no difference to the scope of the appellants' duty whether reinsurance was available or not. They were bound to report correctly whether reinsurance was available and had been obtained or not. Beyond this they were not bound to report the market's response to the risk. It was enough for Mr Crawley that the reinsurance had been obtained (or not). He did not ask to be informed, and was not concerned to know, how easy or difficult it was to obtain it. As para 84 of his judgment demonstrates, Evans LJ thought that a duty to advise on the current market assessment of the Bullen treaty followed automatically from a duty to advise in regard to the availability of reinsurance. But it does not. The two duties are quite distinct and are capable of generating different mechanisms of loss. An express representation that reinsurance has been obtained necessarily involves an implicit representation that it is available. But the only reason Mr Crawley wanted to know in advance that reinsurance was available was that it would be a waste of time to proceed any further if it was not; and his only interest later was to know that it had been obtained, and therefore implicitly that it was available. Breach of the duty to report in advance that reinsurance was available was not in itself causative of any additional loss beyond, perhaps, some trivial expenses. Breach of the representation that reinsurance was available which was implicit in the representation that it had been obtained was incapable in itself of generating any additional loss.

I have already described Ward LJ's reasoning, at pp 164-165, as inconsistent with the SAAMCO principle. The appellants did accept the responsibility of "there being no contract without cover" as he stated. The "but for" test is satisfied, as it was in SAAMC O. It simply does not follow that the appellants assumed responsibility for Aneco entering into "the transaction as such." Nor does it follow that their duty was "to protect Aneco from the very risk they knew Aneco would not accept if no satisfactory reinsurance was not available." What follows in my opinion is that they assumed a duty to protect Aneco from loss on the transaction to the extent that Aneco wished to reinsure it and relied on them to do so.

In my opinion, the appellants did not assume responsibility for advising on the market's assessment of the risks inherent in the Bullen treaty. They were not asked to give such advice and did not do so. No such duty was pleaded or proved. The appellants' only duty was to take reasonable care to ensure that the information which they did supply, viz. that the desired reinsurance was available and that they had obtained it, was correct; and they are responsible for all the foreseeable consequences of that information being wrong.

The proper measure of damages must reflect the fact that the appellants are not liable for the foreseeable consequences of Aneco's entering into the Bullen treaty, but only the consequences of Aneco's having done so on a false basis, that is to say, on the basis that it had effective reinsurance cover. It is not liable for the consequences which would have occurred even if the appellants' advice (that reinsurance was available and had been obtained) had been correct. They are not liable for these because they are the consequences of risks Aneco would have taken if the appellants' advice had been sound.

As I have already observed, by reporting that they had obtained the desired reinsurance as instructed, the appellants implicitly represented that such reinsurance was available. Is the proper measure of damages aggravated by the fact that such reinsurance was not in fact available?

In examining this question it is convenient to take first the common case where the principal has already written the underlying risks before he instructs the broker to obtain reinsurance. If such reinsurance is available in the market and the broker reports that he has obtained it when he has not, his liability is limited to the value of the lost cover. So much is clear. Where such reinsurance is not available, however, the broker is not liable for the loss of cover which no amount of care and skill could have obtained. His negligence will have caused no loss beyond some trivial expenses. Failure to obtain the unobtainable is not in itself causative of loss.

But where, as in the present case, the principal has not yet written the underlying risk and is not prepared to write it without the reinsurance, and the broker knows this, the position is different. Once the principal has acted to his potential detriment by writing the underlying risk in reliance on the broker's statement that cover has been obtained (and implicitly is available), the broker cannot reduce or avoid his liability by leading evidence that, contrary to what he told his principal, cover was not in fact available. So his liability is not less than it would have been if reinsurance had been available. But why should it be more?

Since a representation that cover has been obtained necessarily implies that it is available, it would be surprising if the measure of damages were greater where cover is not available than where it is. The absence of available cover in itself generates no additional loss beyond that due to the fact that it has not been obtained. What is the mechanism which caused Aneco to suffer an additional $24m of loss which it would not have suffered if the reinsurance had been available in the market?

Aneco argues as follows. If the appellants had performed their duty, they would have discovered that the reinsurance was not available and would have been bound to report the fact to Aneco. If Mr Crawley had asked why not, the appellants would have reported the market's assessment of the risks inherent in the Bullen treaty. So, it is said, the distinction between the availability of reinsurance and the market's assessment of the Bullen treaty is artificial; they go hand in hand; and should generate the same measure of damages.

I cannot accept this for several reasons, not least because it is inconsistent with Lord Hoffmann's rejection of Nyekredit's argument in SAAMCO [1997] AC 191 , 223 to which I have already referred.

In the first place, the reasoning is speculative and implausible. Why should Mr Crawley ask why cover was unobtainable? It was enough for him that it was unobtainable; he was not concerned to know why. Without it he would not go ahead with the Bullen treaty anyway. But this is a minor quibble. The essential point is that Aneco entered into the Bullen treaty because it was led to believe that the reinsurance had been obtained. It would have done the same whatever the true state of affairs and irrespective of the availability of reinsurance in the market. If the reinsurance was not obtained in fact, Aneco would have suffered the same loss whether or not it was available.

In the second place, it does not matter what the appellants would have done as a matter of fact; what matters is what they were bound to do as a matter of legal obligation. This is because their liability is measured by the scope of their duty . It is not measured by what would have happened if they had acted in a way in which they were not bound to act. The distinction between liability for the consequences of information they were bound to provide and liability for the consequences of information they would probably have volunteered is not an artificial one. It lies at the heart of an inquiry into the scope of a defendant's liability for the consequences of his actions. Aneco must show, not that Mr Foster would probably have reported the market's reactions in fact, but that he was legally bound to do so, and for the reasons I have given this it cannot do.

In the third place, this is not the purpose for which Aneco wanted to know that cover was available. Nyekredit argued that, if it had been provided with the information it had asked for, it would have realised that the project was not viable. That was a stronger case than the present, for the valuers were asked to provide the information in question. Nyekredit's claim failed because the information was provided for a different purpose (assessing the sufficiency of the security and not the viability of the project.) In the present case, the appellants were not asked for the market's assessment of risk; and their advice as to the availability of reinsurance was provided for a different purpose, viz. to satisfy Mr Crawley that the condition on which alone Aneco would participate in the Bullen treaty could be satisfied, and not to enable him to judge the merits of the Bullen treaty.

Finally, Aneco's argument, like that of Nyekredit, employs an illegitimate mechanism of loss and yields the wrong measure of damages. The court does not engage in speculating on what would have happened if the defendant had performed his duty. That is not the basis of his liability. The court asks only what loss is attributable to his failure to perform his duty. He is liable for all the consequences of his advice being wrong; he is not liable for the consequences which would have occurred even if the facts had been as he represented them to be. The appellants are not liable for the $24m loss which would still have occurred if the reinsurance had been available as they said it was and had been obtained as they said it had.

I consider that the concession in the Superhulls Cover case [1990] 2 Lloyd's Rep 431 was wrongly made. I would allow the appeal and restore the judgment of Cresswell J.