The Judicial Committee of the Privy Council Decisions

1. This appeal raises a point of principle about the extent of the liability of a vendor of shares who gives a limited warranty as to the preparation of a profit forecast upon which the purchaser relies.

2. The vendor was a subsidiary of a listed New Zealand company called Lion Nathan Limited, which guaranteed the vendor's obligations. In the contract, Lion Nathan Limited is called "the Guarantor" but for present purposes it will be convenient to call it "the vendor". By a contract dated 13th May 1989 the vendor agreed to sell the entire issued share capital of Oasis Industries Limited ("Oasis") to a substantial Australian company called C.C. Bottlers Limited ("the purchaser"). The price was NZ$250 million, subject to adjustment. Oasis carried on business bottling

3. The contract was negotiated with some haste and secrecy because neither party wanted other players in the soft drinks market to know that Oasis was being sold. This meant that the purchaser could not be given access to the company's books and had to rely upon the information provided by the vendor. The price was calculated by applying a multiple of 20 to the vendor's forecast of the expected profits after tax in the company's 1988-89 year of account, which ran from 1st September 1988 to 2nd September 1989. The vendor was willing to warrant the accuracy of its forecast up to the date of completion, which was fixed for 3rd July 1989. The contract provided for an adjustment of the price to reflect any shortfall in the forecast profit up to that date. But because thereafter the company was expected to be under the control of the purchaser, the vendor was willing to offer only a more limited warranty in respect of the earnings during the remaining two months or so of the financial year. It supplied a Projected Revenue Statement ("PRS") which forecast that the company's earnings before interest and tax ("EBIT") during that period would be $2.223m. There was no provision for adjustment of the price by reference to this forecast but Schedule 4, paragraph 32 of the contract contained a warranty in the following terms:-

4. In the event, there was a substantial shortfall in earnings over the whole of the financial year. This led, in accordance with the contract, to an adjustment of the price by reference to the shortfall in respect of the first ten months. As to these figures, there is no dispute. The action arises out of the forecast of $2.223m for the last two months. The actual earnings during that period are agreed to have been $1.233m.

5. Blanchard J. held that there had been a breach of warranty. He found that the forecast had not been calculated on a proper basis and that on then current trends and performance the forecast results were not reasonably achievable. The Court of Appeal upheld the judge's finding on liability and there is no appeal against that decision. The issue is over the measure of damages.

6. The judge construed paragraph 32 to mean, putting the matter very shortly, that the vendor warranted that reasonable care had been taken in the preparation of the forecast. He said that the requirement that the forecast should be "on a proper basis" meant that it had to be one which -

7. As to the requirement that the results in the forecast were "achievable", the judge said that:-

8. The Court of Appeal agreed with the judge's view that the first part of the warranty was concerned solely with the way in which the forecast had been made. But they said that the second part, warranting that the forecast results were achievable, was the equivalent of the warranty as to quality. It was a warranty that Oasis was a company reasonably capable of achieving earnings of $2.223 in the period in question.

9. This difference over construction has an important effect on the way in which damages are calculated. In the case of a warranty as to the quality of the goods, the purchaser is prima facie entitled to the difference between what the goods as warranted would have been worth and what they were actually worth. If the vendor had warranted that the earnings in the last two months would be $2.223m, there would have been an analogy with a warranty of quality and the damages would prima facie have been the difference between what the shares would have been worth if the earnings had been in accordance with the warranty and what they were actually worth. The Court of Appeal was saying that although the vendor had not warranted that the earnings would be $2.223m, he had effectively warranted that the company could be valued on the assumption that they would be in the region of $2.223m. As the region would be a range above and below the figure of $2.223m, the reasonable buyer would value such a company, as the actual purchaser had

10. On the other hand, if one construes paragraph 32 as a warranty that reasonable care has been taken in the preparation of the forecast, there is no analogy with a warranty of quality. The forecast, though prepared with reasonable care, may on account of unknown or unforeseeable factors turn out to be substantially inaccurate. It therefore does not warrant that the company has any particular quality. The prima facie rule for breach of a warranty of quality of goods cannot be applied. One must therefore return to the general principle of which that rule is only one example, namely that damages for breach of contract are intended to put the plaintiff in the position in which he would have been if the defendant had complied with the terms of the contract. In this case the vendor represented to the purchaser that $2.223m was a figure upon which he could rely in calculating the price. The figure was in fact used in the calculation of the price. If the vendor had made a forecast in accordance with the terms of the warranty, he would have produced a lower figure and the price would have been correspondingly lower. The damages are therefore the difference between the price agreed on the assumption of NZ$2.223m earnings and what the price would have been, using the same method of calculation, if the forecast had been properly made.

11. This was the principle applied by Blanchard J. and accepted by the Court of Appeal as correct in respect of the first part of the warranty. But their Lordships think, with respect to the Court of Appeal, that the judge was also right in construing the second half as having the same meaning as the first, namely that reasonable care had been taken in the preparation of the forecast. The construction given to the second half by the Court of Appeal amounts in substance to a warranty that for the purposes of calculating the value of the company, it could be assumed that the forecast would be correct. Their Lordships think that this is exactly the form of warranty which the vendor did not wish to give. They therefore agree with the judge that the crucial question in this case is the ascertainment of what a properly prepared forecast would have been.

12. The judge decided that a properly prepared forecast would have been $1.6m. He assessed the damages by making a deduction for tax and calculating what the price would have been if the agreed multiple of 20 had been applied to the resulting figure. The Court of Appeal held that the figure of $1.6m was too high and that a

13. The judge heard a great deal of expert and other evidence in the course of a trial which lasted 25 days. Their Lordships therefore accept that his finding should not be disturbed unless it can be shown to be founded upon some error of principle.

14. In order to see the principle upon which the judge chose the figure of $1.6m, it is necessary to summarise and quote from his careful and detailed judgment. In dealing with the question of liability, he found that the forecast was not, as one would have expected, based upon induction from the information available at the time it was made. Instead, it started from the desired result, namely to produce a forecast which, together with the results and projections for the rest of the year, would justify a price of $250 million. The forecasts of expected sales volumes, discounts and expenses were made to yield that result. In some cases this was done without attempting to obtain information which was available and would have shown that the assumptions necessary to support the forecast were unsustainable. In other cases, it was done in the teeth of such information. Their Lordships quote, with some annotation, the judge's summary of his conclusions:-

15. So much for the findings as to why the forecast failed to comply with the warranty. They plainly show a radical departure from acceptable methods of forecasting. The judge's reasoning on the assessment of damages is contained in the following passage:-

16. What did the judge mean by saying "highest tolerable forecast" and "the relevant side of the ballpark"? Their Lordships think that the concept of the ballpark refers back to the passage quoted earlier in which the judge construed the word "achievable" in paragraph 32 as meaning that the actual figure would be "in the same ballpark", followed by the reference to the Adelaide Oval, which their Lordships understand to be a relatively small cricket ground. The concept which the judge seems to have had in mind is the range of deviation from the mean which is a necessary element in any forecast. A forecaster who predicts that profits in a given period will be, say, $2.223 million, is not doing anything so silly as to say that in his opinion the profits will be precisely that figure. He is saying that $2.223 million is in his opinion the most probable outcome, but that figures slightly higher or lower are almost equally probable and that on either side of them there is a range of possible figures which become increasingly less probable as they deviate from the mean. The forecaster, if asked, should be able to supplement the bald figure with a statement of the limits of deviation to which confidence can be attached. In some cases, the information which he has may enable him to say that the probability of deviation outside fairly narrow limits is very small. In other cases, the possibility of the outcome being affected by unknown or unforeseeable factors may be much greater and the limits of foreseeable deviation therefore much wider. The same is true of a valuation of property, which is no more than an estimate of what a property would fetch on a given date, based upon induction from information about what similar property has fetched.

17. Their Lordships therefore understand the judge to mean that if the vendor had taken reasonable care in the preparation of his

18. Their Lordships can for the moment postpone consideration of this question because in their opinion the judge's use in this context of the concept of reasonable foreseeable deviation was in itself erroneous. As has been said, a forecast is always the forecaster's estimate of the most probable outcome, the mean figure within the range of foreseeable deviation. The judge appears to have assumed that if a figure would have been within the range of foreseeable deviation from the mean of a properly prepared forecast, it must follow that it would have been proper to put that figure forward as the mean. This proposition has only to be stated to be seen to be fallacious. There is no connection between the range of foreseeable deviation in a given forecast and the question of whether the forecast was properly prepared. Whether a forecast was negligent or not depends upon whether reasonable care was taken in preparing it. It is impossible to say in the abstract that a forecast of a given figure "would not have been negligent". It might have been or it might not have been, depending upon how it was done.

19. Assume, for example, that the vendor had forecast $1.25 million and that the limits of foreseeable deviation would have been regarded as $50,000 either way. Assume that the forecast was unexceptionable in every respect but one: there had been a careless double counting of sales which, if noticed, would have reduced the estimate by $25,000. To that extent, the estimate has not been made with reasonable care. If on account of some compensating deviation the outcome is $1.25m or more, the purchaser will have suffered no loss and the vendor will incur no liability. But if the outcome is less than $1.25m, their Lordships think that the purchaser is entitled to say that if the estimate had been made with reasonable care, the figure put forward by the vendor as the mean and upon which he relied in fixing the price, would have been $25,000 lower. To this extent, he has suffered loss by reason of the breach of warranty. It is nothing to the point that the

20. Their Lordships think that the judge confused the concept of potential deviation between the forecast mean and the eventual outcome with a related but different concept, namely the variation which may exist in forecasts of the mean by different forecasters, all of whose forecasts are made with reasonable care. In other words, we are concerned here not with the possible variance between forecast and outcome, but with possible variances between one reasonable hypothetical forecast and another. Because the prediction of the mean involves questions of judgment and degree, there will obviously be a range of possible forecasts which can all be regarded as reasonable interpretations of all the relevant information. The judge of course had this well in mind when he considered whether the actual forecast which the vendor had made complied with the warranty. Having made the finding that the vendor had not taken reasonable care, he had then for the purposes of assessing the damages to decide what figure should actually have been put forward as the most likely outcome.

21. In a case in which it is possible to isolate the negligent error from the rest of the forecast, as in the example of the single $25,000 mistake, it would be reasonable to say that in other respects the forecast would have been the same. All that is necessary is to adjust the figure by $25,000. But in this case, the breach of warranty went to the whole methodology of the forecast. It is not possible to say that in any particular respect, this vendor would, if he had taken proper care, have produced elements of the same forecast. It is therefore necessary to approach the question objectively and ask what a reasonable forecast would have been. This in turn involves choosing from within the range of forecasts, all of which would have been reasonable interpretations of the information then available to the vendor. Where within this range should the court choose?

22. Mr. Sumption Q.C., who appeared for the appellant, said that the court should choose the highest figure, which on the

23. All this makes perfectly good sense when damages depend upon a prediction of how the defendant would have performed outstanding contractual obligations which gave him a choice of what to do. But this is not such a case. Paragraph 32 did not allow the vendor to choose from the range of figures which would have counted as reasonable forecasts and put forward the highest figure in the range in order to obtain the highest possible price. The PRS had to be "calculated in good faith" and therefore had to be a bona fide estimate made without regard to whether it would have produced a higher or lower price. There is accordingly no basis for calculating the damages on the assumption that the vendor was contractually entitled to choose the highest figure. All that can be said is that there would have been a range of possible figures. But there is no legal basis for assuming that the hypothetical figure would have been at the upper rather than the lower end of the range. So far as the estimates which could reasonably have been made differed from the mean of those estimates, it would have been on a random basis and not because the vendor was entitled to choose the highest figure. In those circumstances, the only rational course open to a court is to choose the figure which it considers that a forecast made with reasonable care was most likely to have produced.

24. The Court of Appeal held that the judge had found this figure to be $1.2m and had then added 33 a%. It would therefore have substituted $1.2m for his figure but for the fact that the plaintiffs' claim was based on the assumption that the right figure would have been $1.223m. It therefore used this figure instead.

25. Their Lordships find it difficult to read the passage in the judgment, which has already been cited, as a finding that $1.2m was the figure which a forecast made with reasonable care was most likely to have produced. It is possible that this is what the judge meant, but the judgment goes no further than to ascribe it to the plaintiffs' experts and to say that the "highest tolerable forecast" would have been 33 a% higher. Unless, therefore, the judgment provides material upon which a finding of the correct figure can be made, the case would have to be remitted to the judge.

26. Their Lordships think that in considering this matter, one may start with a prima facie assumption that the range of reasonable possible forecasts will be distributed around the figure which was the actual outcome. The uncertainty inherent in the process of forecasting may have led to reasonable forecasts both higher and lower than the actual outcome. But since those uncertainties tend in both directions, the only way in which a court, required to find a particular figure, can deal with the matter is to regard the unpredictable factors as cancelling each other out. The actual outcome is therefore prima facie likely to have represented the mean and therefore the figure most likely to have been put forward. This prima facie assumption may however be displaced by evidence that the outcome was affected, in one particular direction, by a factor which could not have been reasonably foreseen. For example, if demand was reduced by the imposition of an unforeseen tax after the date of the forecast, the mean of the whole range of reasonable forecasts will to that extent be higher than the actual outcome. But the mere existence of uncertainties which could have affected the outcome either way is no more than the reason why there would have been a range of reasonable forecasts and does not affect a conclusion that the most likely mean figure would have been the actual outcome.

27. The learned judge makes two references in his judgment to matters which occurred after the forecast and would or might have been unforeseeable at the time. One is a passing reference, in the passage already quoted, to the possibility that the level of competition offered by Pepsico "could not have been fully foreseen". The other, earlier in the judgment, was to the increased sales in August in advance of an announced price rise taking effect at the beginning of September, which the judge said were not predictable. The amount of the increase was unknown.

28. Their Lordships think that these two factors may, at best for the vendors, be regarded as cancelling each other out. There is accordingly no evidence to displace the prima facie assumption that the most likely forecast would have reflected the actual

29. © CROWN COPYRIGHT as at the date of judgment. Licensed for Internet publication on terms that no alteration may be made to the wording of this text, and any reproduction must include the text of this notice