I have had the advantage of reading in draft the speech to be delivered by my noble and learned friend, Lord Steyn. As to the issue of liability raised by the cross-appeal, I agree with his reasoning and conclusions: I would restore the judge's finding that Smith New Court Securities Ltd. ("Smith") had established that Citibank N.A. (through Mr. Roberts) fraudulently induced Smith to purchase the Ferranti shares by making the second and third representations, but not the first representation.
The damages issue which is the subject-matter of the appeal raises for decision for the first time in your Lordships' House the question of the correct measure of damages where a plaintiff has acquired property in reliance on a fraudulent misrepresentation made by the defendant. The position in the present case is complicated by the fact that there are two frauds involved. The first, "Roberts fraud," is the fraudulent representation made by Mr. Roberts on behalf of Citibank on 21 July 1989 which induced Smith to buy 28,141,424 Ferranti shares for £23,141,424. The second, "Guerin fraud," is the fraud practised by Mr. Guerin on Ferranti. Although the Guerin fraud was committed before 21 July 1989, its existence was unknown to Citibank, Smith, Ferranti and the market until after that date. Shortly stated the question is who should bear the risk of the Guerin fraud: Smith (which still held the Ferranti shares when the Guerin fraud was discovered) or Citibank which by its servant had fraudulently induced Smith to buy the Ferranti shares.
The relevant facts lie within a comparatively narrow compass. The judge held that Smith bought the Ferranti shares at 82¼p as a market making risk, i.e. with a view to holding them on its books over a comparatively long period to be sold on at a later date. He further held that Smith would not have bought at that price at all apart from the Roberts fraud. Such a purchase is to be contrasted with a "bought deal" where the market maker buys the shares with a view to its agency branch selling them on in smaller parcels to its clients, such sales usually taking place within a matter of hours. The judge held that, if Smith had been considering a bought deal, it could not have bid more than 78p, at which price Citibank would not have sold to Smith. From these facts, two points emerge: first, as a result of the Roberts fraud Smith bought the Ferranti shares with a view to holding them for a comparatively long period; second, if Smith had bid on the basis of a bought deal, it would not have acquired the shares.
The history of the Ferranti shares subsequent to 21 July 1989 was as follows. On 11 August 1989 Ferranti published its Annual Reports and Accounts for the year ending 31 March 1989 which confirmed the results stated in a preliminary announcement made on 14 July 1989. On 11 September 1989 the directors of Ferranti announced that information had come to their attention which required a restatement of the 1989 accounts. At their request dealing in Ferranti shares was suspended. On 29 September the Chairman of Ferranti wrote to the shareholders telling them that Ferranti had been the victim of a major fraud by Mr. Guerin. Trading in Ferranti shares resumed on 3 October. On 17 November Ferranti published its revised audited accounts which showed that the effect of the Guerin fraud was even worse than had been thought.
Smith had retained all the Ferranti shares which it had bought on 21 July. But from 20 November 1989 onwards Smith began to trickle these shares onto the market and obtained prices ranging from 49p down to 30p. By 30 April 1990 it had sold them all for a total of £11,788,204, i.e. at a loss of £11,353,220. It was not suggested at the trial that Smith's retention of the shares until November 1989 or their subsequent sales were in any way unreasonable.
The judge found that Smith first learned of Mr. Roberts fraud on 5 December 1989 although there was evidence to suggest that by mid-November 1989 Smith was suspicious of the truth of Mr. Roberts' representations. On 2 January 1990 solicitors for Smith wrote to Citibank purporting to rescind the contract for the purchase of the 28m. Ferranti shares. At the trial, the claim to rescind was persisted in until the closing speech for Smith when it was expressly abandoned for reasons not examined before your Lordships. Thereafter the only claim by Smith has been for damages for deceit.
Both before the trial judge, Chadwick J.  B.C.L.C. 1104, and the Court of Appeal, Nourse, Rose and Hoffmann L.JJ.  1 W.L.R. 1271, the argument proceeded on the basis that, where a fraudulent misrepresentation has induced the plaintiff to enter into a contract of purchase, the measure of damages is, in general, the difference between the contract price and the true open market value of the property purchased, valued as at the date of the contract of purchase. This was the law as laid down in a series of cases decided at the end of the 19th century, usually in relation to shares purchased in reliance on a fraudulent prospectus: see Twycross v Grant (1877) 2 C.P.D. 469; Waddell v Blockey (1879) 4 Q.B.D. 678; Peek v Derry (1887) 37 Ch.D. 541 and subsequently treated as settled law by the Court of Appeal in McConnel v Wright  1 Ch. 546. It was common ground that there was one exception to this general rule: where the open market at the transaction date was a false market, in the sense that the price was inflated because of a misrepresentation made to the market generally by the defendant, the market value is not decisive: in such circumstances the "true" value as at the transaction date has to be ascertained but with the benefit of hindsight: McConnel v Wright.
Now in the present case the market value of the Ferranti shares at the transaction date (21 July 1989) was inflated, since the existence of the Guerin fraud was then unknown: there was, in one sense, a false market. But that false market was not attributable to the fraud of the defendant, Citibank: the Roberts fraud had no impact on the open market price. The difference between Chadwick J. and the Court of Appeal lay in the fact that Chadwick J. held that there was a latent defect (i.e. the Guerin fraud) in the Ferranti shares and that, even though the false market was not due to the fraud of Citibank, he had to find the "true" value of the Ferranti shares, using hindsight. He accordingly valued the Ferranti shares at what would have been their open market value had the market known of the Guerin fraud on 21 July 1989. The judge fixed this value at 44p per share giving a total true value of the shares on the transaction date of £12,382,226. He accordingly awarded as damages the difference between the contract price and that figure, i.e. £10,764,005.
The Court of Appeal on the other hand took the view that it was only legitimate, in the case of quoted shares, to depart from the market price as at the transaction date where that price was falsified by the defendant's misrepresentation. In all other cases the market value has to be taken to be the quoted value. The Court of Appeal therefore reduced the damages to £1,196,010, being the difference between the contract price and the value of the shares at 78p a share, being the value at which on 21 July Smith itself would have been prepared to buy. The Court of Appeal was conscious that, in so holding, they were throwing the whole risk of catastrophic events onto the innocent purchaser rather than the fraudulent vendor. But they held that such injustice stemmed from the rigidity of two rules (both of which had been common ground before them): first, the denial of rescission where a plaintiff cannot return in specie the very shares which were the subject matter of the fraudulent sale; second, the rule which requires the damages to be calculated as at the date of sale.
As to the first rule referred to by the Court of Appeal, the reasons why Smith abandoned their claim to rescind were not explored before your Lordships. I will therefore say nothing about the point save that if the current law in fact provides (as the Court of Appeal thought) that there is no right to rescind the contract for the sale of quoted shares once the specific shares purchased have been sold, the law will need to be closely looked at hereafter. Since in such a case other, identical, shares can be purchased on the market, the defrauded purchaser can offer substantial restitutio in integrum which is normally sufficient.
As to the second rule referred to by the Court of Appeal - the rule requiring damages to be assessed as at the date of the transaction - Mr. Grabiner, for Smith, submitted that the basis on which the 19th century cases were decided was erroneous and that later decisions show the right approach to the assessment of damages. I agree with those submissions and rather than consider the sterilities of the argument surrounding the 19th century cases proceed at once to consider the more modern law.
As ever in considering damages in tort, the starting point must be to repeat, yet again, the well-known statement of Lord Blackburn in Livingstone v Rawyards Coal Co. (1880) 5 App.Cas. 25, 39:
I do not think there is any difference of opinion as to its being a general rule that, where any injury is to be compensated by damages, in settling the sum of money to be given for reparation of damages you should as nearly as possible get that sum of money which will put the party who has been injured, or who has suffered, in the same position as he would have been in if he had not sustained the wrong for which he is now getting his compensation or reparation.
To that statement he added these important words which are less frequently quoted:
That must be qualified by a great many things which may arise - such, for instance, as by the consideration whether the damage has been maliciously done, or whether it has been done with full knowledge that the person doing it was doing wrong. There could be no doubt that there you would say that everything would be taken into view that would go most against the wilful wrongdoer - many things which you would properly allow in favour of an innocent mistaken trespasser would be disallowed as against a wilful and intentional trespasser on the ground that he must not qualify his own wrong, and various things of that sort.
In Clark v Urquhart  A.C. 28, 67-68 Lord Atkin cast doubt on whether the measure of damages laid down in the 19th century cases as summarised in McConnel v Wright (supra) was correct. He said:
I find it difficult to suppose that there is any difference in the measure of damages in an action of deceit depending upon the nature of the transaction into which the plaintiff is fraudulently induced to enter. Whether he buys shares or buys sugar, whether he subscribes for shares, or agrees to enter into a partnership, or in any other way alters his position to his detriment, in principle, the measure of damages should be the same, and whether estimated by a jury or a judge. I should have thought it would be based on the actual damage directly flowing from the fraudulent inducement. The formula in McConnel v Wright may be correct or it may be expressed in too rigid terms. I reserve the right to consider it if it should ever be in issue in this House.
In the High Court of Australia, Dixon J. in Potts v Miller (1940) 64 C.L.R. 282, 299-300, whilst loyally applying the old inflexible rule was plainly unhappy with it: see also Toteff v Antonas (1952) 87 C.L.R. 647.
The decision which restated the law correctly is Doyle v Olby (Ironmongers) Ltd.  2 Q.B. 158. In that case the plaintiff had been induced by the fraudulent misrepresentation of the defendant to buy an ironmonger's business for £4,500 plus stock at a valuation of £5,000. Shortly after the purchase, he discovered the fraud and started the action. But despite this he had to remain in occupation: "he had burned his boats and had to carry on with the business as best he could." After three years, he managed to sell the business for £3,700, but in the meantime he had incurred business debts. Lord Denning M.R., after referring to Lord Atkin's dictum, stated the principle as follows, at p. 167:
On principle the distinction seems to be this: in contract, the defendant has made a promise and broken it. The object of damages is to put the plaintiff in as good a position, as far as money can do it, as if the promise had been performed. In fraud, the defendant has been guilty of a deliberate wrong by inducing the plaintiff to act to his detriment. The object of damages is to compensate the plaintiff for all the loss he has suffered, so far, again, as money can do it. In contract, the damages are limited to what may reasonably be supposed to have been in the contemplation of the parties. In fraud, they are not so limited. The defendant is bound to make reparation for all the actual damages directly flowing from the fraudulent inducement. The person who has been defrauded is entitled to say:
All such damages can be recovered: and it does not lie in the mouth of the fraudulent person to say that they could not reasonably have been foreseen. For instance, in this very case Mr. Doyle has not only lost the money which he paid for the business, which he would never have done if there had been no fraud: he put all that money in and lost it; but also he has been put to expense and loss in trying to run a business which has turned out to be a disaster for him. He is entitled to damages for all his loss, subject, of course to giving credit for any benefit that he has received. There is nothing to be taken off in mitigation: for there is nothing more that he could have done to reduce his loss. He did all that he could reasonably be expected to do.
In the same case Winn L.J., said, at p. 168:
It appears to me that in a case where there has been a breach of warranty of authority, and still more clearly where there has been a tortious wrong consisting of a fraudulent inducement, the proper starting point for any court called upon to consider what damages are recoverable by the defrauded person is to compare his position before the representation was made to him with his position after it, brought about by that representation, always bearing in mind that no element in the consequential position can be regarded as attributable loss and damage if it be too remote a consequence: it would be too remote not necessarily because it was not contemplated by the representor, but in any case where the person deceived has not himself behaved with reasonable prudence, reasonable common sense, or can in any true sense be said to have been the author of his own misfortune. The damage that he seeks to recover must have flowed directly from the fraud perpetrated upon him.
The damages awarded by the Court of Appeal in that case were calculated (admittedly on a rough and ready basis as to the figures) as follows, at pp. 169-170. The plaintiff was treated as having lost £9,500, the price paid for the business and stock. Against this, he had to give credit for £3,500, i.e. not for the value of the business at the transaction date but for the amount he actually received on the resale of the business three years later. To this £3,500 there were added other benefits which he had received so as to give a total of £7,000 benefits received to be set against the sum lost of £9,500, i.e. a balance of loss of £2,500. In addition, the plaintiff was awarded by way of consequential damages the sum of £3,000 in respect of liabilities incurred by him in running the business. Thus the total award for direct and consequential damages was £5,500.
Doyle v Olby (Ironmongers) Ltd. establishes four points. First, that the measure of damages where a contract has been induced by fraudulent misrepresentation is reparation for all the actual damage directly flowing from (i.e. caused by) entering into the transaction. Second, that in assessing such damages it is not an inflexible rule that the plaintiff must bring into account the value as at the transaction date of the asset acquired: although the point is not adverted to in the judgments, the basis on which the damages were computed shows that there can be circumstances in which it is proper to require a defendant only to bring into account the actual proceeds of the asset provided that he has acted reasonably in retaining it. Third, damages for deceit are not limited to those which were reasonably foreseeable. Fourth, the damages recoverable can include consequential loss suffered by reason of having acquired the asset.
In my judgment Doyle v Olby (Ironmongers) Ltd. was rightly decided on all these points. It is true, as to the second point, that there were not apparently cited to the Court of Appeal the 19th century cases which established the "inflexible rule" that the asset acquired has to be valued as at the transaction date: the successful appellant was not legally represented. But in my judgment the decision on this second point is correct. The old "inflexible rule" is both wrong in principle and capable of producing manifest injustice. The defendant's fraud may have an effect continuing after the transaction is completed, e.g. if a sale of gold shares was induced by a misrepresentation that a new find had been made which was to be announced later it would plainly be wrong to assume that the plaintiff should have sold the shares before the announcement should have been made. Again, the acquisition of the asset may, as in Doyle v Olby (Ironmongers) Ltd. itself, lock the purchaser into continuing to hold the asset until he can effect a resale. To say that in such a case the plaintiff has obtained the value of the asset as at the transaction date and must therefore bring it into account flies in the face of common sense: how can he be said to have received such a value if, despite his efforts, he has been unable to sell.
Doyle v Olby (Ironmongers) Ltd. has subsequently been approved and followed by the Court of Appeal in East v Maurer  1 W.L.R. 461 and Downs v Chappell  3 All E.R. 344. In both cases the plaintiffs had purchased a business as a going concern in reliance on the defendant's fraudulent misrepresentation. In each case after discovery of the fraud they sold the business at a loss and recovered by way of damages the difference between the original purchase price and the price eventually realised on a resale, i.e. the old date of transaction rule was not applied. In Banque Bruxelles Lambert S.A. v Eagle Star Insurance Co. Ltd.  3 W.L.R. 87 your Lordships treated the measure of damages for fraud as being in a special category regulated by the principles of Doyle v Olby (Ironmongers) Ltd.
Turning for a moment away from damages for deceit, the general rule in other areas of the law has been that damages are to be assessed as at the date the wrong was committed. But recent decisions have emphasised that this is only a general rule: where it is necessary in order adequately to compensate the plaintiff for the damage suffered by reason of the defendant's wrong a different date of assessment can be selected. Thus in the law of contract, the date of breach rule "is not an absolute rule: if to follow it would give rise to injustice, the court has power to fix such other date as may be appropriate in the circumstances:" per Lord Wilberforce in Johnson v Agnew  A.C. 367, at 401A. Similar flexibility applies in assessing damages for conversion (I.B.L. Ltd. v Coussens  2 All E.R. 133) or for negligence (Dodd Properties Ltd. v Canterbury City Council  1 W.L.R. 433). As Bingham L.J. (as he then was) said in County Personnel (Employment Agency) Ltd. v Alan R. Pulver & Co.  1 W.L.R. 916, 925-926:
While the general rule undoubtedly is that damages for tort or breach of contract are assessed at the date of the breach .... this rule also should not be mechanistically applied in circumstances where assessment at another date may more accurately reflect the overriding compensatory rule.
In the light of these authorities the old 19th century cases can no longer be treated as laying down a strict and inflexible rule. In many cases, even in deceit, it will be appropriate to value the asset acquired as at the transaction date if that truly reflects the value of what the plaintiff has obtained. Thus, if the asset acquired is a readily marketable asset and there is no special feature (such as a continuing misrepresentation or the purchaser being locked into a business that he has acquired) the transaction date rule may well produce a fair result. The plaintiff has acquired the asset and what he does with it thereafter is entirely up to him, freed from any continuing adverse impact of the defendant's wrongful act. The transaction date rule has one manifest advantage, namely that it avoids any question of causation.
One of the difficulties of either valuing the asset at a later date or treating the actual receipt on realisation as being the value obtained is that difficult questions of causation are bound to arise. In the period between the transaction date and the date of valuation or resale other factors will have influenced the value or resale price of the asset. It was the desire to avoid these difficulties of causation which led to the adoption of the transaction date rule. But in cases where property has been acquired in reliance on a fraudulent misrepresentation there are likely to be many cases where the general rule has to be departed from in order to give adequate compensation for the wrong done to the plaintiff, in particular where the fraud continues to influence the conduct of the plaintiff after the transaction is complete or where the result of the transaction induced by fraud is to lock the plaintiff into continuing to hold the asset acquired.
Finally, it must be emphasised that the principle in Doyle v Olby (Ironmongers) Ltd.  2 Q.B. 158, strict though it is, still requires the plaintiff to mitigate his loss once he is aware of the fraud. So long as he is not aware of the fraud, no question of a duty to mitigate can arise. But once the fraud has been discovered, if the plaintiff is not locked into the asset and the fraud has ceased to operate on his mind, a failure to take reasonable steps to sell the property may constitute a failure to mitigate his loss requiring him to bring the value of the property into account as at the date when he discovered the fraud or shortly thereafter.
In sum, in my judgment the following principles apply in assessing the damages payable where the plaintiff has been induced by a fraudulent misrepresentation to buy property:
The defendant is bound to make reparation for all the damage directly flowing from the transaction;
Although such damage need not have been foreseeable, it must have been directly caused by the transaction;
In assessing such damage, the plaintiff is entitled to recover by way of damages the full price paid by him, but he must give credit for any benefits which he has received as a result of the transaction;
As a general rule, the benefits received by him include the market value of the property acquired as at the date of acquisition; but such general rule is not to be inflexibly applied where to do so would prevent him obtaining full compensation for the wrong suffered;
Although the circumstances in which the general rule should not apply cannot be comprehensively stated, it will normally not apply where either (a) the misrepresentation has continued to operate after the date of the acquisition of the asset so as to induce the plaintiff to retain the asset or (b) the circumstances of the case are such that the plaintiff is, by reason of the fraud, locked into the property.
In addition, the plaintiff is entitled to recover consequential losses caused by the transaction;
The plaintiff must take all reasonable steps to mitigate his loss once he has discovered the fraud.
Before seeking to apply those principles to the present case, there are two points I must make.
First, in Downs v Chappell  3 All E.R. 344, 361, Hobhouse L.J. having quantified the recoverable damage very much along the lines that I have suggested, sought to cross-check his result by looking to see what the value of the business would have been if the misrepresentations had been true and then comparing that value to the contract price. Whilst Hobhouse L.J. accepted that this was not the correct measure of damages, he was seeking to check that the plaintiff was not being compensated for a general fall in market prices (for which the defendant was not accountable) rather than for the wrong done to him by the defendant. In my view, such a cross-check is not likely to be helpful and is conducive to over-elaboration both in the evidence and in argument.
Second, in Royscot Trust Ltd. v Rogerson  2 Q.B. 297 the Doyle v Olby (Ironmongers) Ltd. measure of damages was adopted in assessing damages for innocent misrepresentation under the Misrepresentation Act, 1967. I express no view on the correctness of that decision.
How then do those principles apply in the present case? First, there is no doubt that the total loss incurred by Smith was caused by the Roberts fraud, unless it can be said that Smith's own decision to retain the shares until after the revelation of the Guerin fraud was a causative factor. The Guerin fraud had been committed before Smith acquired the shares on 21 July 1989. Unknown to everybody, on that date the shares were already pregnant with disaster. Accordingly when, pursuant to the Roberts fraud, Smith acquired the Ferranti shares they were induced to purchase a flawed asset. This is not a case of the difficult kind that can arise where the depreciation in the asset acquired between the date of acquisition and the date of realisation may be due to factors affecting the market which have occurred after the date of the defendant's fraud. In the present case the loss was incurred by reason of the purchasing of the shares which were pregnant with the loss and that purchase was caused by the Roberts fraud.
Can it then be said that the loss flowed not from Smith's acquisition but from Smith's decision to retain the shares? In my judgment it cannot. The judge found that the shares were acquired as a market-making risk and at a price which Smith would only have paid for an acquisition as a market-making risk. As such, Smith could not dispose of them on 21 July otherwise than at a loss. Smith were in a special sense locked into the shares having bought them for a purpose and at a price which precluded them from sensibly disposing of them. It was not alleged or found that Smith acted unreasonably in retaining the shares for as long as they did or in realising them in the manner in which they did.
In the circumstances, it would not in my judgment compensate Smith for the actual loss they have suffered (i.e. the difference between the contract price and the resale price eventually realized) if Smith were required to give credit for the shares having a value of 78p on 21 July 1989. Having acquired the shares at 82¼p for stock Smith could not commercially have sold on that date at 78p. It is not realistic to treat Smith as having received shares worth 78p each when in fact, in real life, they could not commercially have sold or realised the shares at that price on that date. In my judgment, this is one of those cases where to give full reparation to Smith, the benefit which Smith ought to bring into account to be set against its loss for the total purchase price paid should be the actual resale price achieved by Smith when eventually the shares were sold.
Finally I must mention a point raised by Mr. Sumption, namely, that it is not open to Smith to argue before your Lordships that the damages should be assessed in the manner which I have proposed. On the pleadings, the damages claimed were the difference between the contract price and either the "true" value on 21 July 1989 or their value when the fraud was discovered. Mr. Sumption urged, in addition to the point on the pleadings, that it would be unfair to Citibank to entertain the new argument since, if the point had been pleaded Citibank would itself have pleaded and led evidence of a failure by Smith to mitigate its loss and as to the reasonableness of Smith's conduct.
Although the pleading point is technically correct (and could be cured by amendment) I am not satisfied that Citibank is prejudiced by allowing this point to be argued before your Lordships. In opening his case before the judge, Mr. Grabiner conceded that, in any event, he could not recover more than Smith's actual realised loss. It could therefore have been an issue at the trial, if Citibank had chosen to make it one, whether Smith should have sold earlier or at a better price. Indeed, as I understand it, those matters were investigated in that context at the trial and the judge found that Smith had acted reasonably. In the circumstances, I can see no injustice to Citibank in deciding this case on the new point raised before your Lordships.
For these reasons I would hold that the damages recoverable amount to £11,352,220 being the difference between the contract price and the amount actually realised by Smith on the resale of the shares. However, as there was no appeal by Smith against the judge's assessment of the damages at £10,764,005, Smith's claim must be limited to that latter amount. I would therefore allow the appeal and restore the judge's order.
Lord Keith of Kinkel
For the reasons given in the speeches to be delivered by my noble and learned friends, Lord Browne-Wilkinson and Lord Steyn, which I have read in draft and with which I agree, I would allow the appeal, dismiss the cross-appeal and restore the order of the trial judge as to damages.
The speech to be delivered by my noble and learned friend, Lord Steyn deals with both liability and damages. On the issue of liability I gratefully adopt the analysis of law and fact, and agree that in respect of the second and third occasions it has, but in respect of the first it has not, been established that an actionable misrepresentation was made by Citibank N.A. (acting through Mr. Roberts) to Smith New Court Securities Ltd. I also agree that the conclusion regarding the first occasion, which differs from that of the Court of Appeal, does not affect the liability of Citibank for having induced Smith to enter into the purchase of Ferranti shares.
On the question of damages I further agree that they should be assessed on the basis of the difference between the contract price and the amount actually realised by Smith on the resale, and would therefore allow the appeal and restore the award of the trial judge on damages. On this aspect of the dispute I wish, however, to add a very few words. Notwithstanding the high authority of its source I cannot regard the judgment of Lord Denning M.R. in Doyle v Olby (Ironmongers) Ltd.  2 Q.B. 158 as an invariable guide to the assessment of damages for fraudulent misrepresentation. The appeal in that case was not fully argued by counsel. The judgments were not reserved and do not sit very easily together. To my mind the propositions which on the argument of the present appeal were sought to be extracted from the decision were painted with too broad a brush to deal accurately with all the problems which may arise.
True, the assessment of damages often involves so many unquantifiable contingencies and unverifiable assumptions that in many cases realism demands a rough and ready approach to the facts. True also that in a case of fraud there are good reasons for departing in some respect from the ordinary rules: and the irrelevance of foreseeability provides an example. Nevertheless, there are instances where more is required in the way of analysis than can be found in Doyle v Olby (Ironmongers) Ltd., and this is one. For my part, I would suggest that in the future when faced with situations such as the present courts would do well to be guided by the seven propositions set out by my noble and learned friend, Lord Browne-Wilkinson in the latter part of his speech. The fourth and fifth of these are, I believe, amply sufficient to show that the damages awarded by the judge ought to be upheld.
Lord Slynn of Hadley
I have had the advantage of reading the text of the speech prepared by my noble and learned friends, Lord Browne-Wilkinson and Lord Steyn. I agree that for the reasons they give the appeal of Smith New Court Securities Ltd. should be allowed and that the order of Chadwick J. as to damages be restored and that the cross-appeal on liability should be dismissed.
There is an appeal and cross-appeal against the judgment of the Court of Appeal dated 17 February 1994 to be considered. The Court of Appeal upheld a judgment dated 25 March 1992 by Chadwick J. so far as he concluded that the plaintiffs in an action in the Chancery Division had established actionable fraudulent misrepresentations in respect of the sale of a parcel of shares. Despite the fact that Chadwick J. had come to his conclusions in a witness action the Court of Appeal held that he had misdirected himself on one aspect of the issues of fact on liability. Accordingly the Court of Appeal felt free to consider the case afresh without being rigidly bound by all the judge's findings of fact. In the result the Court of Appeal affirmed the judgment on liability of Chadwick J. on additional grounds. The correctness of the Court of Appeal's decision on liability is the subject matter of the cross-appeal. The arguments on the cross-appeal were addressed to issues of pure fact. Moreover, in material respects the arguments of counsel for the cross-appellant challenged concurrent findings of fact of the trial judge and the Court of Appeal.
By contrast the appeal challenges the decision of the Court of Appeal on a question of principle, namely the correct measure of damages in an action for deceit. The judge adopted a valuation method to assess damages and held that the buyers were entitled to damages in a sum of the order of £10.7m. In arriving at this conclusion the judge took into account a subsequent fall in the value of the shares, which was caused by a pre-existing and unconnected fraud which had been perpetrated on the company concerned. The Court of Appeal ruled that as a matter of law the judge applied the wrong measure of damages, and that the correct measure was the difference between the price paid by the buyer and the price which, absent the misrepresentations, the shares would have fetched on the open market on the acquisition date. It was common ground that on that date the fraud perpetrated by the third party was not known to the market. On this basis the Court of Appeal reduced the damages to which the buyers were entitled to a sum of the order of £1.1m.
My Lords, a detailed review of the testimonial battleground at trial has left me in no doubt that the cross-appeal ought to be dismissed. I will have to explain my reasons for this firm conclusion in some detail. The helpful judgments of Chadwick J. and the Court of Appeal are now respectively reported:  B.C.L.C. 1104 and  1 W.L.R. 1271. It is therefore possible to deal with the cross-appeal on liability somewhat more economically than would otherwise have been the case. I will then turn to the important question of law as to the correct legal measure of damages. I will explain why I think that the appeal should be allowed and that the award of damages made by Chadwick J. should be restored.
The story of the Ferranti shares
In July 1988 Citibank N.A., a company carrying on business as a bank in London, made available a loan facility of £23m. to Parent Industries Inc., a United States company. As security for the loan Parent charged 28m. shares in Ferranti International Signal Inc. to Citibank. A Mr. Guerin, a former director of Ferranti, was the beneficial owner of Parent. Mr. Peck was Mr. Guerin's man at Parent and occupied the position of President. By mid-July 1989 Parent was in default under the loan agreement. Citibank was urgently considering the realisation of the security for its loan. In the phrase often used at the trial it was a "forced sale" situation.
On 21 July 1989 Citibank, acting through the brokers Scrimgeour Vickers (Asset Management) Ltd., sold the 28m. Ferranti shares to Smith New Court Securities Ltd. ("Smith") for about £23m., the price per share being 82¼ p. Mr. Roberts, a senior employee of Citibank and director of Scrimgeour, arranged the sale. In doing so he dealt with Mr. Lewis and Mr. Abrahams, two directors of Smith and market makers by occupation.
At the trial the case of Smith was that Mr. Roberts induced Smith by fraudulent misrepresentations to buy the Ferranti shares. In order to understand the nature of Smith's case it is necessary to explain the vicissitudes of the value of shares in Ferranti. On 14 July 1989 Ferranti made a preliminary announcement of its financial results for the year ended 31 March 1989. On the basis of that information the market value of the parcel of Ferranti shares was probably of the order of 78p to 82p per share, i.e. a few pence lower than the prices quoted on the Stock Exchange. Ferranti, Citibank and Smith, as well as the individuals acting for these companies, and the market generally, did not know that a massive fraud had been perpetrated on Ferranti. In real terms the market price of the Ferranti shares on 21 July 1989 was a fictitious price.
There was a false market in Ferranti shares. The fact of the fraud and its impact on the value of Ferranti shares only became known in September 1989. By a letter of 29 September 1989 together with unaudited group accounts the Chairman of the Board of Ferranti explained to shareholders that the fraud had caused a reduction in the net worth of the Ferranti Group as at 31 March 1989 of approximately £170m. (from £370.8m. to £198.5m.) and a reduction in profit for the year of approximately £18m. (from £29.3m. to £10.9m.). In November 1989 Ferranti published the revised audited accounts for the year ended 31 March 1989. Those accounts confirmed the pessimistic predictions made in late September. In these changed circumstances, and between 20 November 1989 and 30 April 1990, Smith disposed of the Ferranti shares by selling them in the market in relatively small parcels at prices ranging from 49p per share to 30p per share. The difference between the total price paid by Smith and the total of the prices received was £11.3m.
That brings me to the events of 21 July 1989 so far as they are relevant to the alleged fraudulent misrepresentations. In order to render the shape of the case intelligible it will be necessary to give a chronological account of the sequence of events, with the rival contentions of the parties as to the principal disputes interspersed. Shortly before 9.30 a.m. on 21 July 1989 Mr. Roberts asked Mr. Lewis whether Smith would be interested in buying the Ferranti shares. At 9.43 a.m. Mr. Lewis phoned Mr. Roberts. Mr. Abrahams was also a party to the conversation. The discussion lasted 13 minutes. Mr. Lewis confirmed that Smith was interested in purchasing the Ferranti shares. Smith's case was that during the conversation Mr. Roberts said that Smith would be in competition with two other bidders interested in buying the Ferranti shares, namely a company in the Citicorp Group and another bidder not in the securities industry. Mr. Roberts identified the first company as Citicorp Scrimgeour Vickers Ltd. ("C.S.V."), a company carrying on business as stockbrokers and market makers in London. This was the first alleged representation. At trial Mr. Roberts said that he went no further than to say that there were at least two other parties interested. There was in fact no bidder for the Ferranti shares from outside the securities industry. The dispute as to what was said in the 9.43 a.m. conversation was a major issue at the trial. What was not in issue was that Mr. Lewis and Mr. Abrahams came to believe that Smith would be in competition for the Ferranti shares with a bidder from outside the securities industry.
Following the 9.43 a.m. conversation, Mr. Abrahams and Mr. Lewis attended a meeting with other employees of Smith to discuss whether Smith should bid for the Ferranti shares and, if so, at what price. Mr. Lewis and Mr. Abrahams told those present at the meeting that Smith would be bidding in competition against C.S.V. and one other bidder from outside the securities industry. The decision taken at the meeting was that Smith should bid 82p per share. The reasoning that led to this decision is of some relevance. The facts are common ground. Smith had a choice. It could have bought the shares as a "bought deal" or as a market making risk. The first would have involved Smith buying the Ferranti shares and selling them through Smith's agency arm, at a profit, within a matter of hours to institutional clients. In a transaction of the magnitude of buying 28m. shares in Ferranti it would have been normal for Smith to do a bought deal. Instead Smith chose to do a transaction of the second type. This involved buying the shares with a view to holding them as a market making risk and only selling them as and when the opportunity or opportunities to do so might arise. Smith took the view that the Ferranti shares could not be sold to institutional clients at a price above 80p per share without a recommendation to clients, which Smith's agency arm was not prepared to give. In order to do a bought deal Smith would have had to buy the Ferranti shares at below 80p. If Smith had not believed that it was in competition with a bidder from outside the securities industry, it would have bid for the shares at a price consistent with doing a bought deal at a profit. That price would have been 78p per share. It was agreed that if Smith had bid 78p per share, the bid would not have been accepted by Citibank.
In the presence of Mr. Abrahams, Mr. Lewis telephoned Mr. Roberts at 10.42 a.m. This call lasted about two minutes. It was made some 20 minutes after the conclusion of the pricing meeting. At the trial Mr. Lewis and Mr. Abrahams testified that Mr. Lewis told Mr. Roberts that Smith had decided to bid for the Ferranti shares; that Mr. Lewis asked him to attend at Smith's offices so that Smith could make the bid in person; and that Mr. Lewis asked Mr. Roberts to bring with him the other two bids in sealed envelopes to Smith's offices. Mr. Roberts agreed that Mr. Lewis asked him to come to Smith's offices to hear the bid. But Mr. Roberts denied that anything else was said about bringing the other bids in sealed envelopes.
Mr. Roberts arrived at the offices of Smith shortly after noon. A meeting then took place between Mr. Roberts and three employees of Smith, namely Mr. Lewis, Mr. Abrahams and Mr. Marks. Mr. Marks had been present at the pricing meeting. It is common ground that Mr. Lewis said that Smith would bid 82p per share. Mr. Roberts did not have authority to accept the bid but he said that he would recommend the bid. What is in dispute is the rest of the conversation. Mr. Lewis said that Mr. Roberts said at the start of the meeting that he would disclose the competing bids after Smith had made its bid. This was called the second representation. Mr. Lewis, Mr. Abrahams and Mr. Marks said that after Mr. Lewis made the bid at 82p Mr. Roberts said that Aeritalia (an Italian company) had bid 81p for the shares and C.S.V. had bid 75-77p. This was called the third representation. Mr. Roberts said that he made no mention of bids: he said that he said that Aeritalia and C.S.V. had given indications in the 81p and 75-77p regions. Neither Aeritalia nor C.S.V. had made any bid. It was conceded that if the second and third representations had been made, they would have been fraudulently made.
Mr. Roberts returned to his office and told the decision-makers at Citibank about the bid. While the bid had formally lapsed because it was not immediately accepted Smith remained a willing buyer of the Ferranti shares at 82p per share throughout the afternoon. Shortly after 5.00 p.m. on the same day the bargain was struck. It was done in a telephone conversation between Mr. Lewis and Mr. Abrahams, on the Smith side, and Mr. Fisher, a director and senior dealer at Scrimgeour. The main reason for the additional ¼p was to establish that the contract was made after trading hours on Friday 21 July, so that it would not have to be reported under Stock Exchange rules until the following business day.
The rest of the story can be taken quite briefly. After the suspension of Ferranti shares in September 1989, Smith started to investigate the circumstances in which it had purchased the Ferranti shares. Smith discovered that Aeritalia had never bid for the Ferranti shares. That discovery led to the institution of the proceedings in January 1990.
The trial and the judgment of Chadwick J. on liability:  B.C.L.C. 1104
The trial took place between 25 November 1991 and 17 January 1992. The judge gave judgment on 25 March 1992. He found that in advance of earlier criminal proceedings against Mr. Roberts Serious Fraud Office officials had asked Mr. Lewis and Mr. Abrahams to pool their recollections; that they falsely denied this at the criminal trial; and that in the civil trial they falsely pretended that they had forgotten how their statements came into existence. In the result the judge found that the first representation, which depended exclusively on the evidence of Mr. Lewis and Mr. Abrahams, had not been proved. But, in the light of the totality of the evidence before him, the judge found that Mr. Roberts had made the second and third representations on behalf of Citibank; that those representations were false; and that Smith had been induced to enter into the contract by those fraudulent misrepresentations.
The appeal and the judgment of the Court of Appeal:  1 W.L.R. 1271
Citibank appealed against the judge's findings on liability. Smith served a respondent's notice which invited the Court of Appeal to uphold the judge's conclusions on liability on additional grounds. And that is what the Court of Appeal did. The Court of Appeal held that the judge had misdirected himself in respect of the 9.43 a.m. conversation by considering the credibility and reliability of Mr. Lewis and Mr. Abrahams in isolation. After a review of all the evidence the Court of Appeal found, as a matter of fact, that all three representations were made and made fraudulently and that they induced Smith to enter into the contract. In the result the Court of Appeal dismissed Citibank's appeal on liability.
The submissions for Citibank
Counsel for Citibank put in the forefront of his submissions the undisputed proposition that while as a matter of law fraud only has to be proved to the civil standard, proof to that standard must necessarily take into account the consideration that the more serious the allegation is, the greater the proof is needed to persuade a court that it can be satisfied that the allegation is established. In other words, the very gravity of an allegation of fraud is a circumstance which has to be weighed in the scale in deciding as to the balance of probabilities: In re H. (Minors) (Sexual Abuse: Standard of Proof)  A.C. 563, 586C - 587F, per Lord Nicholls of Birkenhead. But counsel accepted that both Chadwick J. and the Court of Appeal correctly directed themselves in accordance with this standard.
Counsel for Citibank reviewed the minutiae of the evidence. He highlighted undoubted inconsistencies between the accounts of Mr. Lewis and Mr. Abrahams. He argued that there were improbabilities inherent in their accounts. But throughout his speech there was the theme that since Chadwick J. found on proper grounds that Mr. Lewis and Mr. Abrahams had lied it is impossible to sort out in their evidence truth from falsehood. That is an argument worthy of careful consideration. It has rightly been said that a cocktail of truth, falsity and evasion is a more powerful instrument of deception that undiluted falsehood. It is also difficult to detect. But counsel had to face the fact that on the third representation Mr. Marks supported the accounts of Mr. Lewis and Mr. Abrahams. And at the trial counsel never challenged the credibility of Mr. Marks. Counsel for Citibank put the matter quite simply: he said Mr. Marks' evidence was too thin a thread to bear the weight of an elaborate case of fraud. Moreover, counsel argued that the Court of Appeal was not entitled to substitute their view for that of the judge on the first representation. Once it was accepted that the first representation had not been proved he said that it was simply impossible to be satisfied that the second and third representations were made. In the broadest outline these were the principal submissions of counsel for Citibank.
The approach to an attack on concurrent findings of fact
The principle is well settled that where there has been no misdirection on an issue of fact by the trial judge the presumption is that his conclusion on issues of fact is correct. The Court of Appeal will only reverse the trial judge on an issue of fact when it is convinced that his view is wrong. In such a case, if the Court of Appeal is left in doubt as to the correctness of the conclusion, it will not disturb it. That is the first difficulty in the way of upholding the arguments of counsel for Citibank. But there is an additional obstacle. The Court of Appeal upheld the findings of fact of the trial judge on the actionability of the second and third representations. While the jurisdiction of the House is not in doubt, it is most reluctant to disturb concurrent findings of fact. There are two reasons for this approach.
First, the prime function of the House of Lords is to review questions of law of general public importance. That function it cannot properly discharge if it often has to hear appeals on pure fact. This point is underlined by the fact that, despite the economy of presentation of counsel, the hearing on liability lasted more than three days.
Secondly, in the case of concurrent findings of fact, the House is confronted with the combined views of the first instance judge and the Court of Appeal. A suggestion that the House can be expected to take a different view on concurrent findings of fact generally gives rise to an initial sense of disbelief. Nevertheless, I must examine the merits of the argument of counsel for Citibank.
The reasons why the concurrent findings are unassailable
It seems to me that there are five principal reasons why the attack on the concurrent findings of fact must fail.
First, having found that Mr. Lewis and Mr. Abrahams had lied on a collateral matter, the judge approached their evidence with great caution. Rightly, he rejected the notion that falsity in one thing involves falsity in all. Reviewing their accounts as to the second and third representations against the whole body of evidence he accepted, as he was entitled to do, their accounts.
Secondly, the judge was plainly considerably influenced by the fact that on the third representation Mr. Marks in all material respects supported Mr. Lewis and Mr. Abrahams. He accepted the evidence of Mr. Marks.
Thirdly, it is not in dispute that between 10.00 and 10.30 a.m. on the morning of 21 July 1989, at the pricing meeting, which was attended by Mr. Lewis, Mr. Abrahams and Mr. Marks, Smith fixed their bid at 82p on the footing that they would be bidding in competition with two other bidders, one of whom was from outside the securities industry. This fact strongly supported Smith's case.
Fourthly, while disputed by counsel for Citibank, it seems to me inescapable that on Citibank's theory of the case, Mr. Lewis and Mr. Abrahams fabricated the story that they had been told that there were other bidders and the price of the bids several months before the issue of misrepresentation arose. There is the undisputed evidence of Mr. Smith, another employee of Smith, that the account of the rival bids surfaced on 21 July 1989, i.e. the day of the transaction. Given this fact, and Mr. Marks' evidence, the theory of a fabrication is absurd.
Fifthly, as against these factors, the judge had to weigh the evidence of Mr. Roberts. The judge rejected Mr. Roberts' evidence. That necessarily involved a finding that Mr. Roberts gave untruthful evidence. The judge was entitled to take this course. Taking into account counsel's submissions I have reviewed the whole of the evidence of Mr. Roberts, given over more than two days. He was a most unimpressive witness. He testified that at the midday meeting he had said that Aeritalia had given an indication (shorthand for saying they were interested parties) at 81p. It is perfectly clear, however, that Aeritalia was only interested in an option to buy the Ferranti shares for two months. The sale of the shares was, however, a matter of urgency and both Citibank and Parent, acting through Mr. Peck, wanted an outright sale. Mr. Roberts said that he had been told by another Citibank employee that Mr. Peck had said that Aeritalia might make an outright bid. In Mr. Roberts' own words that was "a zero possibility" by midday on 21 July 1989.
Cumulatively, these five factors are sufficient in the particular circumstances of this case to demonstrate convincingly that the attack on the concurrent findings of Chadwick J. and the Court of Appeal must be rejected. In sustaining the second and third representations as actionable fraudulent misrepresentations Chadwick J. in my judgment came to a correct conclusion. So far as the Court of Appeal affirmed the findings of Chadwick J. I am in respectful agreement with their concurrent views.
The Court of Appeal's views on the facts
It is now necessary to consider the exceptional course taken by the Court of Appeal regarding the first representation. It will be recollected that the judge did not find the first representation proved but he did find the second and third representations proved. The clue to this conclusion is to be found in the following passage in the judgment at first instance  B.C.L.C. 1104, 1116C:
.... it would, in my view, be unsafe to make a finding of dishonesty against Mr. Roberts on the unsupported evidence of Mr. Lewis and Mr. Abrahams, I approach the examination of the events of 21 July 1989 on the basis that little, if any, weight can be given to their evidence where it is in conflict with that given by Mr. Roberts.
The Court of Appeal concluded that the judge erred by not subsequently reviewing this conclusion in the light of all the evidence. Counsel for Citibank vigorously challenged the conclusion of the Court of Appeal. It is necessary to analyse the position on a step by step basis.
In making findings of credibility and reliability it is unsafe for a trial judge to compartmentalise the case. In Attorney-General of Hong Kong v Wong Muk Ping  A.C. 501, 510, Lord Bridge of Harwich explained:
It is commonplace of judicial experience that a witness who makes a poor impression in the witness box may be found at the end of the day, when his evidence is considered in the light of all the other evidence bearing upon the issue, to have been both truthful and accurate. Conversely, the evidence of a witness who at first seemed impressive and reliable may at the end of the day have to be rejected. Such experience suggests that it is dangerous to assess the credibility of the evidence given by any witness in isolation from other evidence in the case which is capable of throwing light on its reliability; ....
In other words, an initial and provisional conclusion that a witness is not credible on a particular point may be falsified when considered against the possibilities, probabilities and certainties emerging from the whole body of evidence before the court. That is the error into which the judge fell. He ought to have reconsidered his understandable unwillingness to act on the unsupported evidence of Mr. Lewis and Mr. Abrahams in respect of the first representation in the light of the evidence about the pricing meeting, Mr. Marks' account and the inherent probabilities. There is no internal indication in his judgment that he ever did so.
It follows that the Court of Appeal was entitled to conclude that in respect of the first representation the trial judge misdirected himself. That meant that the Court of Appeal was at large to disregard the judge's findings of fact, even though based on credibility. I understood counsel for Citibank at one stage to suggest that this vitiates all the judge's findings of fact and the whole case on fraud collapses. That is quite unrealistic. The impact of a misdirection is not governed by fixed rules. The appropriate course is dictated by considerations of common sense and fairness as well as close attention to the nature of the misdirection and the circumstances of the particular case. Here the Court of Appeal was fully entitled to take the view that the misdirection only vitiated the judge's findings on the first representation. In all other respects the Court of Appeal was entitled to act on the judge's findings so far as they were unaffected by the misdirection.
On the first representation the Court of Appeal was entitled to come to its own conclusion. The principal reasons for the conclusion of the Court of Appeal were spelt out as follows  1 W.L.R. 1271, 1279:
Events at the pricing meeting and the making of the second and third representations as found by the judge are all inexplicable unless the first representation had also been made .... The judge failed to stand back and consider his finding as to the first representation in the light of his findings as to the second and third. Had he done so, we have little doubt that he would have been driven to conclude, as we do, that the first representation was also made.
In effect the Court of Appeal held that on the first representation the judge should in all the circumstances also have accepted the evidence of Mr. Lewis and Mr. Abrahams, and rejected the evidence of Mr. Roberts. To this extent I respectfully agree with the admittedly exceptional course taken by the Court of Appeal.
That is, however, not the end of the matter. On the first representation, counsel for Citibank was able to demonstrate that, even on an acceptance of the evidence of Mr. Lewis and Mr. Abrahams, there was considerable scope for misunderstanding between the participants in the 9.43 a.m. telephone conversation. In the discussions at 10.42 a.m. and at midday Mr. Lewis and Mr. Abrahams on their evidence (and the evidence of Mr. Marks) unambiguously spoke of actual bids. But Mr. Lewis and Mr. Abrahams were less clear about the discussion at 9.43 a.m.: they both said that Mr. Roberts either spoke of bids or about bids to be made. Counsel for Smith argued that Mr. Roberts impliedly represented that he had bona fide and reasonable grounds for saying that bids would be made that day and that he had no such grounds. There is force in this argument. But on any view that is a far less clear-cut position than existed in respect of the second and third representation. That brings me back to another passage in the judgment at first instance. The judge said  B.C.L.C. 1104, at 1129C:
I am not satisfied that the first representation was made in the earlier telephone conversations in the morning of 21 July 1989 in sufficiently unequivocal terms for it to form the basis for an action in deceit ....
Making due allowance for the judge's earlier misdirection, I attach weight to this passage. On balance I too am not persuaded that the evidence of Mr. Lewis and Mr. Abrahams established the first representation in clear enough terms sufficient to justify a finding of deceit. Differing from the Court of Appeal on the interpretation of the evidence of Mr. Lewis and Mr. Abrahams, I would hold that in respect of the 9.43 a.m. conversation an actionable fraudulent misrepresentation has not been established.
Conclusion on cross-appeal
In my view the conclusion that the actionability of the first representation has not been established does not affect the outcome of this appeal. The misrepresentations at the midday meeting on 21 July 1989 induced Smith to enter into the transaction shortly after 5 O'clock on that day. After all, the trial judge found on ample evidence (including that of Mr. Marks) that Smith would have withdrawn from the transaction if these misrepresentations had not been made. The Court of Appeal agreed with this conclusion. So do I. The essentials of the tort of deceit were established. I would dismiss the cross-appeal on liability.
Given the fact that the subsequent dramatic fall in the value of Ferranti shares was caused by the disclosure of an earlier fraud practised on Ferranti by a third party the question is whether Smith is entitled to recover against Citibank the entire loss arising from the fraudulently induced transaction. Smith submits that the Court of Appeal adopted the wrong measure. Smith seeks to recover damages calculated on the basis of the price paid less the aggregate of subsequent realisations. Citibank contends that the loss attributable to the subsequent disclosure of the fraud by a third party is a misfortune risk and is irrecoverable. Citibank argues that the Court of Appeal adopted the correct measure.
Horses and shares
The fraud perpetrated by Mr. Roberts on Smith related to shares quoted on the Stock Exchange. Undoubtedly, the legal measure of damages in an action in deceit when applied to transactions in shares may throw up special problems. It is not simply a matter of the perception of the market as to the value of the shares. If loss is to be determined by way of the price paid less a valuation of the shares at a given date, the determination of the real or true value of the shares, absent the deceit forming the basis of the claim, may give rise to difficult hypothetical problems. Even more difficult problems arise if it is alleged that for extrinsic reasons there has been a false market, e.g. because investors have been misled by widespread false statements about the value of the stock of the company. None of these practical considerations justify the adoption of a special rule in respect of share transactions. The same legal principle must govern sales of shares, goods, a business or land. It is therefore possible to simplify the problem. The example given by Cockburn C.J. in Twycross v Grant (1877) 2 C.P.D. 469 is instructive. He said, at pp. 544-545:
If a man buys a horse, as a racehorse, on the false representation that it has won some great race, while in reality it is a horse of very inferior speed, and he pays ten or twenty times as much as the horse is worth, and after the buyer has got the animal home it dies of some latent disease inherent in its system at the time he bought it, he may claim the entire price he gave; the horse was by reason of the latent mischief worthless when he bought; but if it catches some disease and dies, the buyer cannot claim the entire value of the horse, which he is no longer in a condition to restore, but only the difference between the price he gave and the real value at the time he bought.
Counsel for Citibank argued that Cockburn C.J. erred in saying that if the horse had some latent disease at the time of the transaction the buyer may claim the entire price he paid. He argued that in such a case there was no sufficient causal link between the latent disease and the eventual death of the horse. Counsel for Smith argued that the transaction, which was induced by deceit, directly led to the loss of the entire value of the horse. On any view it is clear that, if Cockburn C.J. is right, the law imposes liability in an action for deceit for some consequences that were unforeseen and unforeseeable when the tortfeasor committed the wrong. And if that is right it may tell us something about the correct disposal of the present case.
The justification for distinguishing between deceit and negligence
That brings me to the question of policy whether there is a justification for differentiating between the extent of liability for civil wrongs depending on where in the sliding scale from strict liability to intentional wrongdoing the particular civil wrong fits in. It may be said that logical symmetry and a policy of not punishing intentional wrongdoers by civil remedies favour a uniform rule. On the other hand, it is a rational and defensible strategy to impose wider liability on an intentional wrongdoer. As Hart and Honoré, Causation in the Law, 2nd ed. (1985), p. 304 observed, an innocent plaintiff may, not without reason, call on a morally reprehensible defendant to pay the whole of the loss he caused. The exclusion of heads of loss in the law of negligence, which reflects considerations of legal policy, does not necessarily avail the intentional wrongdoer. Such a policy of imposing more stringent remedies on an intentional wrongdoer serves two purposes. First it serves a deterrent purpose in discouraging fraud. Counsel for Citibank argued that the sole purpose of the law of tort generally, and the tort of deceit in particular, should be to compensate the victims of civil wrongs. That is far too narrow a view. Professor Glanville Williams identified four possible purposes of an action for damages in tort: appeasement, justice, deterrence and compensation: (1951) 4 Current Legal Problems 137. He concluded, at p. 172:
Where possible the law seems to like to ride two or three horses at once; but occasionally a situation occurs where one must be selected. The tendency is then to choose the deterrent purpose for tort of intention, the compensatory purpose for other torts.
And in the battle against fraud civil remedies can play a useful and beneficial role. Secondly, as between the fraudster and the innocent party, moral considerations militate in favour of requiring the fraudster to bear the risk of misfortunes directly caused by his fraud. I make no apology for referring to moral considerations. The law and morality are inextricably interwoven. To a large extent the law is simply formulated and declared morality. And, as Oliver Wendell Holmes, The Common Law (edited by M. De W. Howe), (1968), p. 106, observed, the very notion of deceit with its overtones of wickedness is drawn from the moral world.
The old cases
For more than a hundred years at least English law has adopted a policy of imposing more extensive liability on intentional wrongdoers than on merely careless defendants. This policy was trenchantly spelt out by Lord Blackburn in Livingstone v Rawyards Coal Co. (1880) 5 App.Cas. 25. He said, at p. 39:
There could be no doubt that there you would say that everything would be taken into view that would go most against the wilful wrongdoer - many things which you would properly allow in favour of an innocent mistaken trespasser would be disallowed as against a wilful and intentional trespasser on the ground that he must not qualify his own wrong, and various things of that sort.
Since Victorian times there have been great developments in our law of obligations. But there has been no retreat from the policy spelt out by Lord Blackburn. On the other hand, the way in which the law can distinguish between the intentional wrongdoer and a man who caused loss by a foolish but honest mistake was not worked out clearly in the old cases. Pasley v Freeman, decided more than 200 years ago, marks the emergence of the tort of deceit: (1789) 3 Durn. & E. 51. In cases framed in deceit the measure of damages was held to involve ascertainment of the "real" or "face" value of the shares at the time of allotment or purchase. See Davidson v Tullock (1860) 36 L.T. 97; Peek v Derry (1887) 37 Ch.D. 541; reversed on liability (1889) 14 App.Cas. 337; Arkwright v Newbold (1881) 17 Ch.D. 301, reversed on liability, 17 Ch.D. 313; Broome v Speak  1 Ch. 586, (affirmed Shepheard v Broome  A.C. 342). Except for some useful general observations on valuation as a method of measuring loss, and the explanation of the enquiry into a past hypothetical event in the sense of the valuation of shares absent the fraud, I do not think those cases help much. And, except for the obiter dictum of Cockburn C.J. in Twycross v Grant, C.P.D. 469, 544, the earlier cases do not touch on the problems in the present case. Finally, even in the last century it was realised that there must be sensible and practical limits to the heads of loss for which even an intentional wrongdoer can be held liable. Thus in Twycross v Grant Cockburn C.J. said that if the fraudulently misdescribed horse subsequently catches a disease and dies the buyer cannot claim the entire value of the horse. But it took a long time before the precise nature of those limitations were clearly understood and explained.
Doyle v Olby (Ironmongers) Ltd.  2 Q.B. 158
Eventually, the idea took root that an intentional wrongdoer is not entitled to the benefit of the reasonable foreseeability test of remoteness. He is to be held liable in respect of "the actual damage directly flowing from the fraudulent inducement:" see the obiter dictum of Lord Atkin in Clark v Urquhart  A.C. 28, 68: and compare dicta of Dixon J. in Potts v Miller (1940) 64 C.L.R. 282, 298-299, and in Toteff v Antonas (1952) 87 C.L.R. 647, 650. It was, however, not until the decision of the Court of Appeal in Doyle v Olby (Ironmongers) Ltd. that the governing principles were clearly laid down. By fraudulent misrepresentation the defendant induced the plaintiff to buy a business. The trial judge awarded damages to the plaintiff on the basis of a contractual measure of damages, i.e. the cost of making good the representations. The Court of Appeal ruled that this was an error and substituted a higher figure assessed on the basis of the tort measure, i.e. restoration of the status quo ante. Lord Denning M.R. explained, at p. 167:
In contract, the damages are limited to what may reasonably be supposed to have been in the contemplation of the parties. In fraud, they are not so limited. The defendant is bound to make reparation for all the actual damages directly flowing from the fraudulent inducement. The person who has been defrauded is entitled to say:
All such damages can be recovered: and it does not lie in the mouth of the fraudulent person to say that they could not reasonably have been foreseen.
Winn and Sachs L.JJ. expressed themselves in similar terms.
The logic of the decision in Doyle v Olby (Ironmongers) Ltd. justifies the following propositions:
The plaintiff in an action for deceit is not entitled to be compensated in accordance with the contractual measure of damage, i.e. the benefit of the bargain measure. He is not entitled to be protected in respect of his positive interest in the bargain.
The plaintiff in an action for deceit is, however, entitled to be compensated in respect of his negative interest. The aim is to put the plaintiff into the position he would have been in if no false representation had been made.
The practical difference between the two measures was lucidly explained in a contemporary case note on Doyle v Olby (Ironmongers) Ltd.: Treitel, "Damages for Deceit," (1969) 32 M.L.R. 556, 558-559. The author said:
If the plaintiff's bargain would have been a bad one, even on the assumption that the representation was true, he will do best under the tortious measure. If, on the assumption that the representation was true, his bargain would have been a good one, he will do best under the first contractual measure (under which he may recover something even if the actual value of what he has recovered is greater than the price).
Concentrating on the tort measure, the remoteness test whether the loss was reasonably foreseeable had been authoritatively laid down in The Wagon Mound in respect of the tort of negligence a few years before Doyle v Olby (Ironmongers) Ltd. was decided: Overseas Tankship (U.K.) Ltd. v Morts Dock & Engineering Co. Ltd. (The Wagon Mound)  A.C. 388. Doyle v Olby (Ironmongers) Ltd. settled that a wider test applies in an action for deceit.
The dicta in all three judgments, as well as the actual calculation of damages in Doyle v Olby (Ironmongers) Ltd., make clear that the victim of the fraud is entitled to compensation for all the actual loss directly flowing from the transaction induced by the wrongdoer. That includes heads of consequential loss.
Significantly in the present context the rule in the previous paragraph is not tied to any process of valuation at the date of the transaction. It is squarely based on the overriding compensatory principle, widened in view of the fraud to cover all direct consequences. The legal measure is to compare the position of the plaintiff as it was before the fraudulent statement was made to him with his position as it became as a result of his reliance on the fraudulent statement.
Doyle v Olby (Ironmongers) Ltd. was subsequently applied by the Court of Appeal in two Court of Appeal decisions: East v Maurer  1 W.L.R. 461 and Smith Kline & French Laboratories Ltd. v Long  1 W.L.R. 1. East v Maurer is of some significance since it throws light on a point which arose in argument. Counsel for Citibank argued that in the case of a fraudulently induced sale of a business, loss of profits is only recoverable on the basis of the contractual measure and never on the basis of the tort measure applicable to fraud. This is an over simplification. The plaintiff is not entitled to demand that the defendant must pay to him the profits of the business as represented. On the other hand, East v Maurer shows that an award based on the hypothetical profitable business in which the plaintiff would have engaged but for deceit is permissible: it is classic consequential loss. Turning to the Smith Kline case it has been suggested that the Doyle v Olby (Ironmongers) Ltd. rule was wrongly applied: Burrows, Remedies for Torts and Breach of Contract, 2nd ed. (1994), pp. 173-174. The correctness of that comment I need not examine. In my view it is sufficient to say that the principles emerging from Doyle v Olby (Ironmongers) Ltd. are good law.
At the risk of being side-tracked I must now refer to two Court of Appeal decisions which were discussed in argument. In Royscot Trust Ltd. v Rogerson  2 Q.B. 297 the Court of Appeal held that under section 2(1) of the Misrepresentation Act 1967 damages in respect of an honest but careless representation are to be calculated as if the representation had been made fraudulently. The question is whether the rather loose wording of the statute compels the court to treat a person who was morally innocent as if he was guilty of fraud when it comes to the measure of damages. There has been trenchant academic criticism of the Royscot case: see Richard Hooley, "Damages and the Misrepresentation Act 1967" (1991) 107 L.Q.R. 547-551. Since this point does not directly arise in the present case, I express no concluded view on the correctness of the decision in the Royscot case. The second case is the decision of the Court of Appeal in Downs v Chappell  3 All E.R. 344. The context is the rule that in an action for deceit the plaintiff is entitled to recover all his loss directly flowing from the fraudulently induced transaction. In the case of a negligent misrepresentation the rule is narrower: the recoverable loss does not extend beyond the consequences flowing from the negligent misrepresentation: see Banque Bruxelles Lambert S.A. v Eagle Star Insurance Co. Ltd.  3 W.L.R. 87. In Downs v Chappell  3 All E.R. 344, 361, Hobhouse L.J. applied this narrower rule to an action for deceit. He enunciated the following "qualification" of the conventional rule [361h]:
In my judgment, having determined what the plaintiffs have lost as a result of entering into the transaction - their contract with Mr. Chappell - it is still appropriate to ask the question whether that loss can properly be treated as having been caused by the defendants' torts, notwithstanding that the torts caused the plaintiffs to enter into the transaction.
That led Hobhouse L.J. "to compare the loss consequent upon entering into the transaction with what would have been the position had the represented, or supposed, state of affairs existed:" at p. 362a. The correctness of this proposition in a case of deceit was debated at the bar. Counsel for Citibank in whose interest it was to adopt this proposition felt some difficulty in doing so. In my view the orthodox and settled rule that the plaintiff is entitled to all losses directly flowing from the transaction caused by the deceit does not require a revision. In other words, it is not necessary in an action for deceit for the judge, after he had ascertained the loss directly flowing from the victim having entered into the transaction, to embark on a hypothetical reconstruction of what the parties would have agreed had the deceit not occurred. The rule in deceit is justified by the grounds already discussed. I would hold that on this point Downs v Chappell was wrongly decided.
The date of transaction rule
That brings me to the perceived difficulty caused by the date of transaction rule. The Court of Appeal referred to the rigidity of "the rule in Waddell v Blockey (1879) 4 Q.B.D. 678, which requires the damages to be calculated as at the date of sale"  1 W.L.R. 1271, 1283G. No doubt this view was influenced by the shape of arguments before the Court of Appeal which treated the central issue as being in reality a valuation exercise. It is right that the normal method of calculating the loss caused by the deceit is the price paid less the real value of the subject-matter of the sale. To the extent that this method is adopted, the selection of a date of valuation is necessary. And generally the date of the transaction would be a practical and just date to adopt. But it is not always so. It is only prima facie the right date. It may be appropriate to select a later date. That follows from the fact that the valuation method is only a means of trying to give effect to the overriding compensatory rule: Potts v Miller, 64 C.L.R. 282, 299, per Dixon J.; and County Personnel (Employment Agency) Ltd. v Alan R. Pulver & Co.  1 W.L.R. 916, 925-926, per Bingham L.J. Moreover, and more importantly, the date of transaction rule is simply a second order rule applicable only where the valuation method is employed. If that method is inapposite, the court is entitled simply to assess the loss flowing directly from the transaction without any reference to the date of transaction or indeed any particular date. Such a course will be appropriate whenever the overriding compensatory rule requires it. An example of such a case is to be found in Cemp Properties (U.K.) Ltd. v Dentsply Research & Development Corporation  2 E.G.L.R. 197, 201, per Bingham L.J.
There is in truth only one legal measure of assessing damages in an action for deceit: the plaintiff is entitled to recover as damages a sum representing the financial loss flowing directly from his alteration of position under the inducement of the fraudulent representations of the defendants. The analogy of the assessment of damages in a contractual claim on the basis of cost of cure or difference in value springs to mind. In Ruxley Electronics & Construction Ltd. v Forsyth  A.C. 344, 360G, Lord Mustill said: "There are not two alternative measures of damages, as opposite poles, but only one; namely, the loss truly suffered by the promisee." In an action for deceit the price paid less the valuation at the transaction date is simply a method of measuring loss which will satisfactorily solve many cases. It is not a substitute for the single legal measure: it is an application of it.
So far I have discussed in general terms the scope of a fraudster's liability in accordance with the rule identified with Doyle v Olby (Ironmongers) Ltd.  2 Q.B. 158. It is now necessary to consider separately the three limiting principles which, even in a case of deceit, serve to keep wrongdoers' liability within practical and sensible limits. The three concepts are causation, remoteness and mitigation. In practice the inquiries under these headings overlap. But they are distinct legal concepts. For present purposes causation is the most important. The major issue in the present case is whether there is a causal link between the fraud and the loss arising by reason of the pre-existing fraud perpetrated on Ferranti. How should this matter be approached?
The development of a single satisfactory theory of causation has taxed great academic minds.: see Hart and Honoré, Causation in the Law, 2nd ed. (1985) and Honoré, "Necessary and Sufficient Conditions in Tort Law," in Owen, Philosophical Foundations of Tort Law, (1995), p. 363. But, as yet, it seems to me that no satisfactory theory capable of solving the infinite variety of practical problems has been found. Our case law yields few secure footholds. But it is settled that at any rate in the law of obligations causation is to be categorised as an issue of fact. What has further been established is that the "but for" test, although it often yields the right answer, does not always do so. That has led judges to apply the pragmatic test whether the condition in question was a substantial factor in producing the result. On other occasions judges assert that the guiding criterion is whether in common sense terms there is a sufficient causal connection: see Yorkshire Dale Steamship Co. Ltd. v Minister of War Transport  A.C. 691, 706, per Lord Wright. There is no material difference between these two approaches. While acknowledging that this hardly amounts to an intellectually satisfying theory of causation, that is how I must approach the question of causation.
REMOTENESS AND MITIGATION
The second limiting principle is remoteness. I have already discussed the special rule of remoteness developed by the courts in the context of deceit. This requirement is in issue in the present case: if there is a sufficient causal link it must still be shown that the entire loss suffered by Smith is a direct consequence of the fraudulently induced transaction. The third limiting principle is the duty to mitigate. The plaintiff is not entitled to damages in respect of loss which he could reasonably have avoided. This limiting principle has no special features in the context of deceit. There is no issue under this heading and I need say no more about it.
It is now necessary to take stock of the case. The distinctive features of the case are-
that the fraud of Mr. Roberts induced Smith to buy the Ferranti shares, the value of which were already at the date of sale doomed to collapse due to the fraud practised on the company by a third party; and
that by reason of the fraud of Mr. Roberts, Smith was induced to buy the shares as a market making risk, i.e. to hold on to the shares for sale at a later stage.
In these circumstances Smith was truly locked into the transaction by reason of the fraud perpetrated on it. And the causative influence of the fraud is not significantly attenuated or diluted by other causative factors acting simultaneously with or subsequent to the fraud. The position would have been different if the loss suffered by Smith arose from a subsequent fraud. That would be a case like the misrepresented horse in Cockburn C.J.'s example in Twycross v Grant 2 C.P.D. 469, 544-545, where the buyer plainly cannot recover the entire value of the horse if it subsequently catches a disease and dies. In the actual circumstances of this case I am satisfied that there was a sufficient causal link between the fraud and Smith's loss. Moreover, for substantially the same reasons, I would hold that Smith's losses, calculated on the basis of the difference between the price paid and the proceeds of subsequent realisations, flow directly from the fraud. In my view Smith would on this basis be entitled to recover the sum of about £11.3m. Smith merely seeks restoration of the order for payment of £10,764,005 which Chadwick J. made on a different basis. In law Smith are entitled to succeed on this appeal to that extent.
I have not lost sight of counsel for Citibank's argument that, given the way the case was pleaded, Smith should not be allowed to succeed on this legal basis. In my view the argument that Citibank was prejudiced is quite unreal. I reject it.
I would allow Smith's appeal on damages and restore the order of Chadwick J.
Arkwright v Newbold (1881) 17 Ch.D. 301; Attorney-General of Hong Kong v Wong Muk Ping  A.C. 501; Banque Bruxelles Lambert S.A. v Eagle Star Insurance Co. Ltd.  3 W.L.R. 87; Broome v Speak  1 Ch. 586; Cemp Properties (U.K.) Ltd. v Dentsply Research & Development Corporation  2 E.G.L.R. 197; Clark v Urquhart  A.C. 28; County Personnel (Employment Agency) Ltd. v Alan R. Pulver & Co.  1 W.L.R. 916; Davidson v Tullock (1860) 36 L.T. 97; Dodd Properties Ltd. v Canterbury City Council  1 W.L.R. 433; Doyle v Olby (Ironmongers) Ltd.  2 Q.B. 158; Downs v Chappell  3 All E.R. 344; East v Maurer  1 W.L.R. 461; I.B.L. Ltd. v Coussens  2 All E.R. 133; In re H. (Minors) (Sexual Abuse: Standard of Proof)  A.C. 563; Johnson v Agnew  A.C. 367; Livingstone v Rawyards Coal Co. (1880) 5 App.Cas. 25; McConnel v Wright  1 Ch. 546; Overseas Tankship (U.K.) Ltd. v Morts Dock & Engineering Co. Ltd. (The Wagon Mound)  A.C. 388; Peek v Derry (1887) 37 Ch.D. 541; Potts v Miller (1940) 64 C.L.R. 282; Royscot Trust Ltd. v Rogerson  2 Q.B. 297; Shepheard v Broome  A.C. 342; Smith Kline & French Laboratories Ltd. v Long  1 W.L.R. 1; Twycross v Grant (1877) 2 C.P.D. 469; Waddell v Blockey (1879) 4 Q.B.D. 678; Yorkshire Dale Steamship Co. Ltd. v Minister of War Transport  A.C. 691
Authors and other references
Burrows, Remedies for Torts and Breach of Contract, 2nd ed. (1994)
Hart and Honoré, Causation in the Law, 2nd ed. (1985)
Honoré, "Necessary and Sufficient Conditions in Tort Law," in Owen, Philosophical Foundations of Tort Law, (1995)
Wendell Holmes, The Common Law (edited by M. De W. Howe), (1968)
all rights reserved
all rights reserved