Standby Letters of Credit and Demand Guarantees

The use of letters of credit in place of conventional guarantees was pioneered by American banks which were not permitted to issue conventional guarantees. By contrast, so called "demand guarantees" - performance bonds, advance payment guarantees and the like – which were originally used to provide protection against non-performance of contracts, are now also used as guarantees of payment. Martin Hughes investigates the legal nature of these instruments having first taken a look at the recent decision in Sirius International Insurance Company v FAI Insurance Limited [2004] UKHL 54.

When I told a colleague the other day that I was writing an article on standby letters of credit and demand guarantees he observed, somewhat sardonically, "that’ll really get them going, won’t it!". Yes, indeed, but like many of the topics I write about, if you don’t know your way around the area, there are some big holes to fall into and numerous "black holes" to be aware of.

Before going through some of the basic principles in this area, I will look at the decision in Sirius International Insurance Company (Publ) v FAI Insurance Limited and others [2004] UKHL 54. Although ostensibly about letter of credit issues – that is how it accidentally found its way to the House of Lords – the case represents a further step in the reshaping of English law as it relates to the interpretation of contracts. It is also yet another nice example of how randomly English law develops. I have said before that in order for points of law to be established we have to wait for the right facts to present themselves and for the right arguments to be run. It now seems that an alternative route is for a case to reach the House of Lords by mistake.

The House of Lords did not expect to find themselves dealing with a point of construction which did not involve a question of general public importance. As Lord Steyn put it at the end of paragraph 3 of the case report "The House would not ordinarily have given leave to appeal in such a one-off case. But the House is now seized with it, and the issue must be resolved". How did this come about? Because (quoting from the same paragraph) "When leave to appeal was granted by an Appeal Committee, it may have appeared that important issues regarding the autonomy principle applicable to letters of credit issued by banks would have to be resolved ……..In the result it has become clear that the appeal should be decided on the basis of the correct contextual (emphasis added) interpretation of two related documents …….[which] is, like the construction of all texts, a matter of law but it does not involve a question of general public importance.".

Lord Bingham also found himself somewhat out of sorts. He agreed that the appeal should be allowed but found it necessary to say (at paragraph 1): "I must acknowledge that the judge adopted the construction favoured by a majority of my noble and learned friends. My own reasons for favouring a different construction differ from those of the Court of Appeal. This being so, no purpose is served by expounding the interpretation which I myself would have put on the order".

So what was the case really about? An order – a so-called Tomlinson order, the one referred to by Lord Bingham - had been made in earlier proceedings and in paragraph 1 of this order (the terms of which were agreed by all the parties to the litigation, FAI included) it was stated that FAI "is indebted to the applicant [Sirius] in the sum of $22,500,000 and the applicant shall be entitled to prove in the liquidation or administration of FAI in the said sum of $22,500,000".

The question at issue was this: did that statement constitute consent by FAI to the payment of an insurance claim by Sirius, being a claim for which FAI’s consent was required under an earlier side letter between Sirius and FAI. The judge at first instance thought it did because that was the "commercial substance", but May LJ in the Court of Appeal disagreed, citing the agreement of Sirius’ counsel that "the literal words of [the side letter] do not express an agreement by FAI that Sirius should pay [the relevant] claim", stating that "creative construction or implication is required to interpret it as doing so" and disagreeing with Sirius’ counsel’s assertion that the agreement that Sirius should be entitled to prove in FAI’s liquidation or administration necessarily carried with it an agreement by FAI that Sirius should pay the relevant claim. "I do not think so" said May LJ.

In the view of Lord Steyn ( at paragraph 18) – and it is worth my quoting this passage at length – "the aim of the inquiry is not to probe the real intentions of the parties but to ascertain the contextual meaning of the relevant contractual language. The inquiry is objective: the question is what a reasonable person, circumstanced as the actual parties were, would have understood the parties to have meant by the use of specific language. The answer to that question is to be gathered from the text under consideration and its relevant contextual scene."

That inelegant final phrase – I am at a loss to know what the phrase "contextual scene" means – and the reference to a reasonable person who is "circumstanced" as the actual parties were - do not, to my mind, sit entirely happily with what I hope remains the locus classicus in this interpretative debate, the judgment of Lord Hoffman in Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896, in the course of which he said, at 913, "The meaning which a document (or any other utterance) would convey to a reasonable man is not the same thing as the meaning of its words. The meaning of words is a matter of dictionaries and grammars; the meaning of a document is what the parties using these words against the relevant background would reasonably have been understood to mean".

I can understand this proposition – and it is a very powerful proposition – which tells us that a document must be interpreted in a reasonable manner with due regard to the context in which it was created. This is not to say, however, that a person – you, me or the judge – who is construing a text has to pretend to be in the situation in which the parties found themselves. In other words, the enquiry really is objective. The question is not ‘what was it like to be there?’ but, given the circumstances, what do I, a reasonable person, think the parties achieved, what was the bargain they struck?

Coincidentally, I read the report of the Sirius decision - Sirius International Insurance Company is a leading Swedish insurer and reinsurer - while visiting Stockholm in order to give a presentation on legal writing to each of three very different groups: a group of colleagues; a group of in-house counsel at a leading Scandinavian financial institution; and a group of law students at Stockholm University under the auspices of ELSA (the European Law Students Association).

At all levels surprise was expressed at the shift in the English approach to the construction of contracts evident in Lord Hoffman’s judgment and in the Sirius decision, the universal view being that English law has historically been seen as the best choice for international commercial contracts precisely because of its literal approach to the construction of contracts.

From this perspective, Lord Steyn’s statement (in paragraph 19 of the Sirius case report) that "If possible it [literalism] should be resisted in the interpretative process" is quite shocking. It remains to be seen whether a development which seems, at least to me, to introduce into English law a healthily pragmatic approach to the construction of contracts will impair English law’s historical status as the pre-eminent law of choice for international commercial transactions.

And now it is time to go back to the beginning and to look at the functions and legal status of standby letters of credit and demand guarantees.

A standby letter of credit is a variant of the basic type of letter of credit, the documentary credit, the function of which is to finance international trade. In a "base case" documentary credit transaction, goods are shipped by the seller from their country of origin to a buyer in a different country and the seller – the beneficiary of the credit - delivers the shipping documents (and an invoice) to the seller’s bank – the issuer - which, if the documents are in order, pays the invoice and seeks reimbursement from the buyer (referred to in this context as the applicant). By contrast, the beneficiary of a standby letter of credit delivers to the issuer not shipping documents but a statement that (or to the effect that) a specified sum is due to it from another party under a contract by way of performance of the contract, eg repayment of a loan, or by reason of the other party’s not performing its obligations under the contract.

Thus, in Kvaerner John Brown Limited v Midland Bank plc [1998] CLC 446 – a "wholly exceptional case" in the words of Cresswell J, but not so because of the circumstances in which the letter of credit was issued – the supplier of materials and design services for the construction of a chemical plant procured the issue of an irrevocable standby letter of credit under which payment was to be made against sight drafts and a statement of the buyer certifying that the supplier had failed to comply with its contractual obligations and that he buyer had given the requisite contractual notice to the supplier

Unlike a conventional guarantee, the obligation of the issuer of standby letter of credit is a primary obligation and not a promise to answer for the obligations of another person – to borrow from the language of the Statute of Frauds 1677. Even if a standby letter of credit is used where a conventional guarantee could have been used, the issuer is not a surety and does not benefit from the rights a surety would have. For example, the issuer of a standby letter of credit will not acquires any right of subrogation when it pays the creditor and its right to be indemnified by the debtor arises as a matter of contract between it and the debtor (unlike the right of a guarantor to be indemnified which is an equitable right given legal force by the Mercantile Law Amendment Act 1856, s 5).

There is rarely any consideration to support the existence of a contractually binding obligation on the part of a bank which issues a letter of credit. It has been suggested that a letter of credit constitutes an offer which is accepted by the seller’s presenting documents (see Elder Dempster Lines Ltd v Ionic Shipping Agency Inc [1968] 1 Lloyds Rep 529) but this analysis does not explain why an irrevocable letter of credit is irrevocable as soon as it is issued. If the offer and acceptance analysis applied, an issuing bank could revoke the letter of credit at any time before acceptance of the offer by the beneficiary, but this is not the case.

The only convincing reason why letters of credit are contractually binding on the parties (and why they are irrevocable when issued, see Goode, Commercial Law, 3rd edition, 2004, p 970, note 77, or when advised to the beneficiary, see Documentary Credits, by Jack, Malek and Quest, 3rd Edition, 2001, p.95, para 5.3) is that this is the position which the courts have accepted for 200 years or so since ‘Lord Mansfield…fashioned the law merchant (lex mercatoria) to the needs of the 18th Century industrial revolution’ (per the US Court of Appeals for the 2nd Circuit in Alaska Textile Co v Chase Manhattan Bank 982 F 2d 813 (1992)).

Almost all documentary credits and many standby letters of credit are expressed to be subject to The Uniform Customs and Practice for Documentary Credits (1993 Revision) ICC Publication No 500, known as the UCP, which was first published in 1933. The UCP apply to documentary credits and, to the extent to which they may be applicable, to standby credits where they are incorporated into the text of the Credit (Article 1) and provide (Article 2) that a credit (documentary or standby) is an arrangement whereby an issuing bank at the request and on the instructions of a customer or on its own behalf: is to pay (or accept and pay a bill drawn by) the beneficiary; authorises another bank to do so; or authorises another bank to negotiate, in each case against stipulated documents, provided that the terms of the credit are complied with.

Since 1998 an alternative code, the International Standby Practices (ISP98), has been available. This code, which was endorsed by the ICC, provides a regime which is substantially similar to the UCP but is specifically designed for use in connection with standby letters of credit.

As Lord Steyn has told us, the House of Lords in the Sirius case had been expecting to deal with important issues regarding the autonomy principle applicable to letters of credit. ‘The fundamental principle governing documentary letters of credit…is that the obligation of the issuing bank…is independent of the performance of the underlying contract for which the credit was issued.’ (Bank of Nova Scotia v Angelica-Whitewear [1987], I SCR 59 at 81, Can SC per Le Dain J).

Another way of saying this is that the parties deal in documents and in documents alone. If the documents provided by the beneficiary to the bank conform to the terms of the credit, the bank must pay irrespective of the performance or failure to perform of the party on whose behalf the letter of credit has been issued. The main exception to this is fraud, whether relating to the letter of credit or the underlying contract. Where the bank is on notice of fraud on the part of the beneficiary or its agent, it is entitled (and obliged) to withhold payment. Other defences available to the bank are misrepresentation, mistake and illegality in the place for performance of the bank’s obligations.

A very clear statement of the "fraud exception" and the reason for it appears in the judgment of Sir John Donaldson MR in Bolivinter Oil S.A. v Chase Manhattan Bank [1984] 1 WLR 392 at 393: "The wholly exceptional case – and the Kvaerner case was one – where an injunction [restraining payment under a letter of credit] may be granted is where it is proved that the bank knows that any demand for payment already made or which may thereafter be made will clearly be fraudulent. But the evidence must be clear, both as to the fact of fraud and as to the bank’s knowledge. It would certainly not normally be sufficient that this rests upon the uncorroborated statement of the customer, for irreparable damage can be done to a bank’s credit in the relatively brief time which must elapse between the granting of such an injunction and an application by the bank to have it discharged".

It is a general principle that the documents presented by the beneficiary of a letter of credit must comply strictly with the terms of the credit – see Equitable Trust Co of New York v Dawson Partners Ltd (1926) Ll L Rep 49 at 52, where this fundamental principle was held to mean that a requirement for delivery of a confirmation from "experts who are sworn brokers" was not satisfied by a confirmation from one such broker, thus entitling the buyer to reject the documents and to decline to reimburse the issuing bank even though this requirement had been communicated by means of a code which made no distinction between the singular and the plural. The issuing bank was not helped by the coding error because it was committed by its correspondent bank.

(This and numerous other aspects of the operation of letters of credit are dealt with in Chapter 14 of my book, Selected Legal Issues for Finance Lawyers, Butterworths, 2003 – now published by Tottel - from which the foregoing paragraph is taken.)

Demand guarantees – so named to distinguish them from conventional guarantees – are instruments under which the issuer (a bank) will become obliged to make a payment to the beneficiary if the issuer’s customer defaults in the performance of a contractual obligation, usually physical performance rather than payment, the usual context for these instruments being construction contracts where the contractor is the bank’s customer.

There are three main types of demand guarantee, but first a warning: a similar type of protection for the beneficiary is sometimes given by a form of contract embodied in a deed under which a bank or, more likely, an insurance company will, together with its customer, enter into a suretyship guarantee - see Trafalgar House Construction (Regions) Ltd v General Surety & Guarantee Co Ltd [1996] AC 199, in which Lord Jauncey of Tullichettle endorses the observations of Lord Atkin in Trade Indemnity Co Ltd v Workington Harbour and Dock Board [1937] AC 1 where he says, at 209, ‘I find great difficulty in understanding the desire of commercial men to embody so simple an obligation in a document which is quite unnecessarily lengthy, which obfuscates its true purpose and which is likely to give rise to unnecessary arguments and litigation as to its meaning’.

(This is the judgment in which Lord Atkin makes the following memorable observation – one which, for me at least, bears frequent repetition.

For issuing banks the key issue is that the cleaner the demand guarantee, the clearer it is that the bank must pay against a simple demand and the less likely it is that the bank will be faced with an insoluble conflict between its need to be seen to honour its obligations and its desire to look after the interests of its customer. When, at the request and cost of its customer, a bank issues a demand guarantee, it is making a promise to pay to its customer’s counterparty a substantial portion of the contract price irrespective of the merits of the counterparty’s claim. There will always be a conflict of interest - in commercial terms – when a demand guarantee is called, but for the bank, the more clear cut its obligation to pay, the less guilt – or perceived guilt – it will have to suffer when it pays.

In I.E. Contractors Ltd v Lloyds Bank PLC and Rafidain Bank [1990] 2 Lloyd’s Rep 496, Staughton LJ observed, at 499, that "there is no reason why a performance bond should not depart from the usual pattern, and be conditioned upon the existence of facts rather than the production of a document asserting those facts" and went on to say that, if the meaning is plain, he would not view as irrebuttable the "bias or presumption" which "holds a performance bond to be conditioned upon documents rather than facts".

From an issuing bank’s perspective the better view is that of Ackner LJ in Esal (Commodities) Ltd and Reltor Ltd v Oriental Credit Ltd and Wells Fargo Bank NA [1985] 2 Lloyd’s Rep 546, CA at 549 where he observed that if the performance bond is so conditional the result would be "wholly inconsistent with the entire object of the transaction, namely to enable the beneficiary to obtain prompt and certain payment".

Instruments which are in substance ‘demand guarantees’ may be called guarantees or bonds. A Bid Bond will oblige the issuer to pay a percentage of the contract price (say 5%) if a successful bidder for a contract decides not to enter into the contract; an Advance Payment Guarantee will oblige the issuer to pay an amount equal to the advance payment if a contracting party which has received an advance payment fails to take the steps, e.g., mobilisation of labour and machinery, that it was required to take by the contract; and a Performance Bond will require the issuer to pay five, perhaps ten, per cent of the contract price if the contractor fails to perform its obligations properly.

The likelihood that there is no consideration for the issue of a demand guarantee is from the raises issues which are more or less the same as those already mentioned above in relation to standby letters of credit. However, since demand guarantees are of relatively recent origin, it seems difficult to resolve the difficulty posed by any lack of consideration by reference to lex mercatoria. Equally unhelpful, it seems to me, is the suggestion that the Contracts (Rights of Third Parties) Act 1999 could solve the problem. If the demand guarantee itself is not a contract, the only available contract is that between the issuer and the applicant under which the issuer’s obligation – to the applicant alone – is to issue the demand guarantee, not to pay the beneficiary.

This is in any event not a point which is likely to be taken. If preservation of the reputation of banks requires the fraud exception to be kept as narrow as possible, how much more incentive is there for English law to maintain that a bank is obliged to pay out under a document which by its terms states that it shall do so if the requisite documents are delivered. If for some reason the point were to be taken – perhaps by a liquidator – the court will have to find a solution to this problem. It is not obvious what form the solution would take, but it is more likely than not that the answer to this question will emerge in a random fashion when circumstances and allow and when counsel choose to run the argument – how else does English law develop (otherwise than by statutory intervention)?

In practice, many demand guarantees are governed by a law other than English law and such instruments are routinely issued without there being an express choice of law, the working assumption being that the governing law is the law of the place where the contract is to be performed. If a beneficiary has a concern about the enforceability of a demand guarantee which is governed by English law, the simple expedient of having the instrument executed as a deed will address that concern.

Most of the rules that apply to standby letters of credit apply to demand guarantees although there are some apparently conflicting judicial observations about the need for "strict compliance" as discussed above. This issue is, however, more apparent than real. It is not so much that there is one rule for letters of credit and another for demand guarantees, but rather that demand guarantees tend to have less exacting provisions with regard to precisely what documents are to be submitted in order to make the issuer liable to pay. In Siporex Trade S.A. v Banque Indosuez [1986] 2 Lloyd’s Rep 146, Hirst J refers at 159 to the absence of a need for precise wording in a performance bond "particularly where the bond itself specifies no more than a requirement for "your declaration to that effect" (his emphasis )".

More accurately, Staughton LJ observed in I.E. Contractors Ltd v Lloyds Bank PLC and Rafidain Bank [1990] 2 Lloyd’s Rep 496 at 501 that "The degree of compliance required by a performance bond may be strict, or not so strict. It is a question of construction of the bond. If that view of the law is unattractive to banks, the remedy lies in their hands". Later in his judgment, responding to a suggestion in Esal - an earlier case in the Court of Appeal - that the need for a beneficiary to inform the bank of the true basis upon which it is making its demand may be "very salutary", Staughton LJ remarked that, like Mr Justice Leggatt (the Judge at first instance in the IE Contractors case), "I doubt whether it is a sufficient reason to adopt a particular construction that the result would be desirable or salutary".

There is definitely some tension here between the absence of any interest in whether a particular construction is "desirable" and the decision of the judge at first instance in the Sirius case that the correct construction was that which reflected the "commercial substance" of the transaction. We can probably all agree, however, that there is no need for the construction of a contract to be affected by the perception that a particular result would be "very salutary".

The principle of autonomy also applies to demand guarantees. In Edward Owen Engineering Ltd v Barclays Bank International Ltd [1978] QB 159, the Court of Appeal, led by Lord Denning MR, confirmed that even though the beneficiary had committed a serious breach of contract – by failing to ensure that the letter of credit whereby the price was to be paid was confirmed – Barclays’ customer was not entitled to an injunction which would prevent Barclays from paying under its performance guarantee. If you are a bank which issues a demand guarantee, your word is your bond.

Subsequent case law has only served to confirm Lord Denning’s penetrating analysis, in July 1977, of the role played by demand guarantees in international trade, an analysis provided in a judgment at the beginning of which (at C on page 164) he observed that "This case concerns a new business transaction called a ‘performance guarantee’ or ‘performance bond’" and towards the end of which he observed (at H on page 170) that "as one takes instance after instance, these performance guarantees are virtual promissory notes payable on demand. So long as the [customers] make an honest demand, the banks are bound to pay".

That says it all.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.