Bunge SA v Nidera BV

This is an arbitration appeal which concerns the proper construction and application of the GAFTA Prohibition Clause and Default Clause.

By GAFTA Appeal Award No.4283 dated 22 June 2012 the GAFTA Board of Appeal upheld the Defendant Buyers’ claim for substantial damages for the wrongful repudiation by the Claimant Sellers of a contract for the sale of 25,000 m.t. (± 10% in Buyers’ option) of Russian milling wheat, 2010 crop, f.o.b. Novorossyisk. The contract, concluded on 10 June 2010 via brokers in Milan, incorporated the GAFTA 49 contract form, a standard set of f.o.b. contract terms designed for contracts “for the delivery of goods from central and eastern europe in bulk or bags”.

The Sellers contend that the contract was automatically cancelled under the GAFTA Prohibition Clause on the announcement of an export ban by the Russian Government on 5 August 2010. The Board disagreed and found that the Sellers’ claim for cancellation was a repudiation. This is the liability issue on the appeal.

The Board awarded the Buyers damages in accordance with the GAFTA Default Clause. The Sellers contend that only nominal damages should have been awarded as no loss was suffered at common law by reason of the principles set out in The Golden Victory [2007] 2 AC 353 and/or as a matter of mitigation. This is the damages issue on the appeal.

Leave to appeal under s.69 of the Arbitration Act 1996 was granted by Andrew Smith J. on 10 October 2012 on the following four questions of law:

The Board found that on 5 August 2010, the contractual delivery period having been narrowed to 23-30 August 2010, the Buyers nominated a vessel to take delivery. However, by Russian Government Resolution No.599, published that same day by the Prime Minister, Vladimir Putin, the export of wheat from the territory of the Russian Federation between 15 August and 31 December 2010 was prohibited (“the export ban”).

The Sellers contended and contend on appeal that under the GAFTA Prohibition Clause, the contract was thereby cancelled.

The Prohibition Clause provides that:

“PROHIBITION

In the case of export, blockade or hostilities or in case of any executive or legislative act done by or on behalf of the government of the country of origin of the goods, or of the country from which the goods are to shipped, restricting export, whether partially or otherwise, any such restriction shall be deemed by both parties to apply to this contract and to the extent of such total or partial restriction to prevent fulfilment whether by shipment or by any other means whatsoever and to that extent this contract or any unfulfilled portion thereof shall be cancelled. Sellers shall advise Buyers without delay with the reasons therefore and, if required, Sellers must produce proof to justify the cancellation.”

The GAFTA Board of Appeal found that (para.5.3):

“….we agree with Buyers, and also with the first tier Tribunal, that to enjoy the protection that the Prohibition Clause affords, it is necessary for the seller to show that the prohibition prevents the seller from performing. As of 11 August, when the contract came to an end by reason of acceptance by Buyers of what they said was Sellers’ anticipatory repudiation, it could not be said, with certainty, that the export ban would prevent Sellers from performing. We accept that the ban was temporary, in the sense that it was of defined rather than indefinite duration but this does assist Sellers since it was always possible that before the delivery period under the Contract expired the export ban might be revoked or modified in some material way so as to permit performance. That this was a possibility is borne out by what has happened with export bans in the past. The US soybean meal embargo is a good example. Export bans are introduced by governments for domestic policy reasons and the wider international ramifications are not always fully thought through. In this case the initial ban was in fact varied, by extending it into 2011. The initial ban could have been curtailed later in August.”

In the light of their finding that it was always possible that before the delivery period under the contract expired the export ban might be revoked or modified so as to permit performance, the Board concluded that the Prohibition Clause did not operate to cancel the contract and so the Sellers’ claim of cancellation was repudiatory.

The Sellers argue that as soon as the prohibition of export was announced the contract was automatically cancelled. On its terms the prohibition applied to goods of the contractual description and to the entirety of the contractual shipment period. As such the contract was there and then cancelled. The clause is forward looking. It is unnecessary to prove or inquire into whether the prohibition in fact had any effect on the sellers’ ability to perform the contract. It is deemed to do so under the clause.

The consequence of the Sellers’ argument is that it would make no difference if the prohibition had been lifted before the shipment period and if it in fact had no impact on the sellers’ ability to perform. The contract is cancelled regardless. There is no requirement to show any causal connection between the prohibition and the failure of performance.

The Buyers argue that it would be remarkable if the clause operated so as to cancel the contract in advance so as to excuse further performance, and indeed disentitle further performance, merely because there was a doubt as to whether the Sellers would be able to perform, which is the effect of the Board’s findings. They submit that the GAFTA Prohibition Clause, in common with virtually all like clauses which have been considered in the authorities, does require proof of a causal connection.

Both parties have addressed the issue by reference to (1) the wording; (2) the authorities, and (3) commercial considerations.

The Sellers submit that:

The Buyers submit that this construction ignores the need to show that the prohibition (or other specified event) was one “restricting export”. It will only do so if it in fact restricts export. On the Board’s findings, at the material time (claimed cancellation), the prohibition was one which might or might not restrict export. It depended upon whether it stayed in place during the contractual shipment period and, on the Board’s findings, it was always possible that it would not do so. For the prohibition to be one “restricting export” it is necessary to show that the ban in fact prevents performance in the sense of being in place for the duration of the time for performance; it is not enough to show that if it remained in place it would prevent performance.

The Buyers further submit that most of the Sellers’ detailed points on construction are neutral. The wording addressed in points (4) and (5), for example, applies meaningfully on either side’s construction of the clause.

There is no case which decides the present issue. However, the Buyers can point to two cases in which essentially the same argument as that advanced by the Sellers was rejected, and in the later case in relation to the same Prohibition Clause. They submit that the reasoning of the court in so doing is highly persuasive.

In Sanday v Cox McEuen (1922) 10 Ll. L.R.459 Bankes LJ rejected the sellers’ argument that there was no need to show a causal connection in relation to a prohibition clause (at p.460rhc):

“Now, as I understand, the sellers' argument is this. It is said that the change of language is deliberate and whereas the clause as printed ran "Should shipment be prevented by prohibition," the substituted clause was, "In the event of prohibition of export," and it is said that the alteration of language was deliberate and intentional for the purpose of avoiding any dispute as to whether the shipment was or was not prevented, and that the intention was by the substitution of the words, "In the event of prohibition of export," to insert a clause the effect of which would be that upon the happening of the event the obligation to ship ceased.

Now, if one were to accept the contention of the sellers it would mean this, that in the event of prohibition by the Argentine Government, whatever the effect of it in reference to a particular cargo might be, the clause is to operate automatically, and therefore, for instance, if the Argentine Government issued a prohibition on the morning of some particular day and then three or four hours later withdrew it, the clause would have automatically operated to the great detriment of the buyer. Now, I cannot accept that contention.

It seems to me that the words: "In the event of prohibition of export," must refer to an effective prohibition, and if you once admit that it must be effective, then there is no material difference between saying, "In the event of effective prohibition" and saying "In the event of a prohibition preventing export," and I am not prepared to attach the meaning to the change of language which is contended for by the sellers.”

Having recognised that the Sanday case treated the prohibition clause as one requiring proof of a causal connection Benjaminon Sale of Goods [8^th ed.) summarises its effect as follows at para.18-388:

“The party claiming to be discharged under such a [prohibition] clause must show that there was an effective prohibition, and not merely a technical one. To be “effective” for this purpose, the prohibition must prevent the seller from exporting goods of the contract description within the shipment period. Thus a seller cannot rely on the [prohibition] clause if the prohibition is imposed only for a short time and then withdrawn when shipment in accordance with the contract is still possible…”

In Pancommercev Veecheema [1983] 2 Lloyd’s Rep.304, although the factual issues were different, the sellers advanced essentially the same argument, namely that the GAFTA Prohibition Clause “is designed to make it unnecessary to enquire into whether the governmental restriction in fact had any effect upon the sellers’ ability to deliver” (p306rhc).

The Court of Appeal rejected this argument and found it to be “wholly unconvincing”. As Sir John Donaldson MR explained (p306-7):

“It is an interesting argument, but one which I find wholly unconvincing. As Mr Justice Bingham pointed out, the commercial consequences would be startling in the extreme. Prohibitions of export are a sellers’ nightmare, but the clause, so construed, would convert them into a sellers’ dream. Possessed of goods which they had contracted to sell and a licence for a sufficient quantity to export them all, sellers would be able to re-negotiate the price to reflect the effect of the prohibition of export not withstanding that they could honour their contracts without let or hindrance.”

The Court of Appeal concluded that “If and in so far as this [the ban] prevents shipment, the contract is cancelled. But on the facts it did not prevent shipment. Accordingly the sellers are in breach” (p.307lhc). The Buyers submit that this conclusion reflects recognition of the need to show a causal connection.

This submission derives some support from how the case is treated in Benjamin at para. 18-395:

“Taken literally, these words [the deeming provision in the GAFTA Prohibition Clause] might be thought to have been intended to apply irrespective of any causal connection between the prohibition and the seller’s non-fulfilment. But in Pancommerce SA v Veecheema BV the Court of Appeal rejected the argument that the sellers could rely on this provision merely because a prohibition had been imposed, when in fact that prohibition had not prevented them from performing”.

The Sellers submit that Bankes LJ’s observations in Sanday were obiter and concerned a differently worded clause. They point out that neither Sanday nor Pancommerce concerned a prohibition which by its terms prevented performance of the contract and therefore did not address the specific issue which arises in this case.

The Sellers place reliance on the case of Ford & Sons (Oldham) Ltd v. Henry Leetham & Sons Ltd (1915) 21 Com Cas 55. This is cited by Benjamin as an example of a prohibition clause which did not require proof of a causal connection – see para. 18-387. The clause in question also contained similar introductory words – “in case of…” However, the clause was not linked to contractual performance. It was a domestic sale of goods case which provided the sellers with an option to cancel “in case of prohibition of export” to this country, most probably because of the potential effect on prices. Since it was not a contract which required performance by export (or import) it is not surprising that the Judge found that there was no need to show a causal connection with performance. It is nevertheless striking that it is the only example which the Sellers could identify of a prohibition clause which was held not to require a causal link. Benjamin notes that this is a case “not now commonly relied upon” (para. 13-395) and neither party have found any case in which it has been.

The Sellers also place reliance on certain passages in Ross T Smyth v. Lindsay [1953] 1 WLR 1280 (Devlin J) and Bremer Handelsgesellschaft mbH v. Vanden Avenne Izegem PVBA [1978] 2 Lloyd’s Rep. 109 (Lord Wilberforce at p114 lhc; Viscount Dilhorne at p121 lhc) which they submit support the concept of prospective cancellation under a prohibition clause operating from the time of the announcement of the prohibition. However, in neither of those cases did the present issue arise or need to be addressed and, as the Buyers point out, there are other passages in the judgments which suggest the need for a causal connection (for example, Devlin J in Smyth v Lindsay at p1284 and Lord Salmon in Bremer v Vanden Aven at p128).

The main consideration which the Sellers emphasise is that on the Board’s approach, buyers and sellers have to continue to conduct themselves (and incur costs) on the footing that performance may ultimately prove possible because, despite the announcement, the ban may in fact be lifted/revised. Much better, the Sellers say, to have the certainty of a cancelled contract immediately upon the announcement than having to wait and see whether the ban remains at the time of performance. However, there are many contractual situations in which parties must continue to conduct themselves on the footing that the contract will, in the future, be performed by the other party despite the possibility, even likelihood, that it will not be as, for example, where there is a likely but not yet unequivocal repudiation.

The Buyers stress the examples given by Bankes LJ and Donaldson LJ set out above. They submit that it is inherently improbable that the parties would intend the contract to be cancelled in respect of a ban which has no impact on performance. Moreover, this would still be the case even if it was found at the time of the announcement of the prohibition that the ban would almost certainly be lifted before the relevant shipment period.

They point out that the consequence of the Sellers’ argument is that a contract which could in fact have been performed is cancelled with unnecessary financial detriment to the seller and unwarranted benefit to the buyer whenever the market is below the contract price at the time of the announcement; and unwarranted financial benefit to the sellers and corresponding unnecessary financial detriment to the buyers whenever the market is above the contract price at the time of the announcement.

They further point out that in the event of a ban of unspecified duration all contracts still to be performed would be automatically cancelled regardless of how long in the future the shipment period(s) might be.

Finally they submit that the Court should give due weight to the conclusions of an arbitral trade tribunal, chosen by the parties, as to the construction of trade terms in relation to which commercial considerations may have a bearing (see Kershaw Mechanical v Kendrick Construction [2006] 4 All ER 79 at [55]-[57] ).

In my judgment the GAFTA Prohibition Clause requires proof of a prohibition “restricting export”. That means establishing that the prohibition does in fact restrict export of goods of the contractual description during the contractual shipment period. A causal connection has to be proven.

On the findings made by the Board it could not be said on 9 August 2010, when Sellers purported to declare the contract cancelled, that there was a prohibition restricting such export. On their findings there was at the time a prohibition which might or might not do so. It depended on whether it was revoked or modified so as to permit some element of performance before the shipment period expired, something which was “always possible”.

The deeming provision in the Prohibition Clause means that it is not necessary for the sellers to show that they had or would have had goods available and could have performed but for the event in question. It does not, however, mean that there is no requirement to prove that export of goods of the contractual description during the contractual shipment period is actually restricted, either totally or partially.

The background to the introduction of the deeming provision was the particular issues which had arisen in relation to cases concerning the 1973 US soyabean embargo and the difficulties of proving prevention in relation to string and “loophole” goods – see Benjamin at para. 18-395 and the Pancommerce case at p306. It meant that there was no longer a need for an inquiry into whether, but for the prohibition, the sellers could have performed the contract, and whether their position was short or long and whether they could or should have obtained goods elsewhere etc. It did not, however, remove the need to show any causal connection between the prohibition and the ability to export goods of the contractual description during the contractual shipment period.

That conclusion is supported by the authorities and the commercial considerations relied upon by the Buyers. In particular:

For all these reasons I conclude that the Board was right to conclude as they did on the liability issue.

The GAFTA Default Clause provides as follows:

“In default of fulfilment of contract by either party, the following provisions shall apply:

(a)

The party other than the defaulter shall, at their discretion have the right, after serving notice on the defaulter, to sell or purchase, as the case may be, against the defaulter, and such sale or purchase shall establish the default price

(b)

If either party be dissatisfied with such default price or if the right at (a) is not exercised and damages cannot be mutually agreed, then the assessment of damages shall be settled by arbitration.

(c)

The damages payable shall be based on, but not limited to, the difference between the contract price and either the default price established under (a) above or the actual or estimated value of the goods on the date of default established under (b) above.

(d)

In all cases the damages shall, in addition, include any proven additional expenses which would directly and naturally result in the ordinary course of events from the defaulter’s breach of contract, but shall in no case include loss of profit on any sub-contracts made by the party defaulted against or others unless the arbitrator(s) or board of appeal, having regard to special circumstances, shall in his/their sole and absolute discretion think fit.

(e)

Damages, if any, shall be computed on the quantity called for, but if no such quantity has been declared then on the mean contract quantity and any option available to either party shall be deemed to have been exercised accordingly in favour of the mean contract quantity.”

The Board awarded damages to the Buyers in accordance with the Default Clause. The Sellers contend that they were wrong so to do because at common law no loss had been suffered (1) because of the effect of the House of Lords decision in The Golden Victory [2007] 2 AC 353 and (2) as a matter of mitigation.

The Sellers contend that the The Golden Victory establishes that whether loss has been suffered by reason of the early termination of a contract depends on what would have happened absent termination, decided on the balance of probabilities and with the benefit of hindsight. Applying that principle there could only be nominal damages. The Board of Appeal found that had the sale contract continued, it would have been cancelled without liability under the GAFTA Prohibition Clause.

As to mitigation, the Sellers contend that likewise there could only be nominal damages. The Sellers offered to reinstate the sale contract in full, on its original terms. They thus offered to restore to the Buyers precisely what any damages claim would assert had been lost. There was no good reason for rejecting that offer; rejecting it was therefore a failure properly to mitigate under Payzu Ltd v. Saunders [1919] 2 KB 581.

In relation to both arguments the Board held that they did not preclude the Buyers from recovering damages in accordance with the Default Clause, the contractually agreed scheme for assessing damages in the case of default by either party.

They found that:

“…A very large number of default cases come before GAFTA arbitrators and GAFTA Appeal Boards. The GAFTA Default Clause is a clause with which everyone in the trade is fully familiar. (para.5.17)

…

On the date of default in sub clause (c) of the Default Clause means exactly what is says. The Default Clause is intended to be easily understood and readily applied by traders and trade arbitrators alike. In a different context in their submissions in this case, Buyers stressed the need for certainty. Certainty is what the clause is designed to achieve and does achieve. This need for certainty is underscored by (e) of the default clause. It avoids any enquiry as to the quantity that would or might have been shipped but for the default. It is also underscored by the second part of sub-clause (d). Subject to the Tribunal’s discretion loss of profits are not to be taken into account. (para.5.20)

…

On mitigation we agree with Buyers that Sellers’ case is misconceived, for several reasons. First, the Default Clause does not require a defaulted against party to attempt to mitigate loss by accepting an offer to reinstate the contract. Second, as of 11 August Buyers had an accrued entitlement to very substantial default damages. Accepting Sellers’ proposal that Buyers withdraw their default declaration would have provided no guarantee to Buyers that they would receive any goods and in the event that they would not have received any goods; because the ban remained in place. Third, delivery would have been possible, albeit not on the delivery terms of the Contract, before the ban came into force but Sellers never came forward with any proposal in that regard.” (para.5.23)

The Sellers submit that to award damages, there must be (1) a decision that loss has been suffered by reason of the default and (2) an assessment of the amount of that loss. They submit that both their common law arguments went to issue (1), whereas the Default Clause was only concerned with issue (2). There was accordingly no reason why common law principles should not be applied regardless of the terms of the Default Clause.

I do not consider, however, that there is a clear dividing line between the Sellers’ two issues. Both are concerned with the assessment of damages. An assessment of damages is being made even if the conclusion is that no damages have been suffered. This is illustrated by the Sellers’ own definition of the questions of law arising out of the award which is put in terms of whether the Default Clause excludes the operation of common law principles in relation to “the assessment of damages”. It is also borne out by the Default Clause itself. Under sub-paragraph (b), where the innocent party has exercised his right under sub-paragraph (a) or one party is dissatisfied with the default price, “then the assessment of damages shall be settled by arbitration”. That must include all aspects of the assessment of damages, including arguments that there had been no loss. If it were otherwise there would be no mechanism being provided for determining such damages issues, which would be nonsensical.

The Sellers further submit that in any event the Default Clause does not exclude the operation of common law principles of damages. They stress that sub-paragraph (c) states that “The damages payable shall be based on, but not limited to …” the damages then set out. They submit that the sense of the words is not that damages equal to the price/value difference referred to are to be paid whether or not loss has been suffered and that the words are nothing like clear enough to exclude the basic compensatory principle of damages under English law, or to exclude the doctrine of mitigation from a consideration of whether damages are to be awarded. In this connection, they rely by analogy on the decision in The Selda [1998] 1 Lloyd’s Rep. 416 (Clarke J.), [1999] 1 Lloyd’s Rep. 729 (C/A) that the words were not clear enough – even without the phrase “but not limited to” that now appears – to exclude other heads of recoverable loss, if suffered, from being awarded as damages.

The Selda raised a different issue. It was a case where the party in default sought to say that the innocent party’s remedy was, under the Default Clause, more limited than it would be at common law. Unsurprisingly it was held that if the clause was to exclude a common law remedy, it needed to make that quite clear. By contrast, here the party in default seeks to say that the innocent party’s remedy as it clearly appears to be under the Default Clause should be limited by the common law.

In The Selda it was recognised that the regime under the Default Clause is in some respects different to the common law. The parties are free so to agree – see, for example, Fleming v Sanofi [2003] 2 Lloyd’s Rep.473 at para.47.

The Sellers’ argument in this case is not that the Default Clause should be supplemented by common law principles, but that such principles should override the contractually agreed damages scheme so as to produce a very different outcome.

The Default Clause sets out a simple and clear scheme that damages “shall” be based upon. That is the parties’ mutual bargain. There is nothing unusual about contracting parties seeking to set out the measure of damages in advance and being confined, for good or bad, to that measure even if does not reflect the measure that would be available at common law. The words “based on” reflect the fact that the clause is setting out the basic measure of damages recoverable, whilst recognising that additional heads of loss may also be so. They do not mean that that measure is only provisionally recoverable and may be displaced in certain, unspecified circumstances.

There is no principled reason for cutting across the parties’ agreed contractual damages scheme. Indeed there is every reason for not so doing, as the Board made clear. Keeping to the agreed contractual scheme promotes simplicity and certainty. As Lord Bingham stated in his powerful dissenting speech in The Golden Victory at p378C:

“The importance of certainty and predictability in commercial transactions has been a constant theme of English commercial law at any rate since the judgment of Lord Mansfield in Vallejo v Wheeler (1774) 1 Cowp 143, 153, and has been strongly asserted in recent years…”

For all these reasons, in my judgment the Board were correct to award damages in accordance with the Default Clause, and regardless of the Sellers’ no loss arguments at common law.

In those circumstances it is not necessary to decide Question 3. However, like the Board, I would not accept that it is settled law that the The Golden Victory approach applies to a one off sale of goods contract such as this. The majority in The Golden Victory recognised that they were departing from the general rule that damages in respect of a marketable commodity fall to be assessed by reference to the available market price at the date of breach, but considered that the compensatory principle justified them so doing in the circumstances. However, The Golden Victory concerned a period contract and the departure from the general rule was only adopted in relation to the period element of the damages claim, not the applicable hire rate. Further, as the Board observed, Lord Scott at paragraph 34 recognised that the assessment at the date of breach rule “is particularly apt” in sale of goods cases, as is reflected in the Sale of Goods Act. At paragraph 35 he drew a distinction between a one-off sale and “a contract for the supply of goods over some specified period”. It is also to be noted that Benjamin treats The Golden Victory as being relevant to sale of goods cases because its “reasoning could apply to long term contracts for the sale of goods” – para.19-170.

Although there are passages in the majority judgments in The Golden Victory which are put in very general terms, I would for my part regard it as very much an open question whether The Golden Victory approach would apply in a one off sale of goods contract where there is an available market and damages fall to be assessed in accordance with the Sale of Goods Act. There is in such a case no difficulty about valuing what has been lost. The innocent party is compensated for the value of what he has lost at the time he loses it. Having had a mitigation opportunity which can be valued without difficulty by reference to the market there is no need or warrant to consider subsequent events. Fixing the damages by reference to market value promotes certainty and predictability, and helps inform the innocent party’s decision whether or not to terminate. However, the matter was not fully argued and since it is not necessary to decide it on this appeal I do not propose to do so.

Finally, in relation to the mitigation argument, I would further observe that the Default Clause itself sets out the required mitigation, either actual (under sub-paragraph (a)) or deemed (under sub-paragraph (c)). In any event I accept the Buyers’ submission that the Board has decided the issue against Sellers not only on the basis that the Default Clause renders the argument irrelevant, but also on the basis that the Buyers’ conduct was not unreasonable in all the circumstances. That is a finding of fact which cannot be appealed.

For all these reasons I conclude that the Board was also right in its conclusion on damages.

The Board reached the correct conclusion and the appeal must be dismissed.