Does a Beneficiary’s Reckless Presentation under Letters of Credit amount to Fraud? – A critique of Credit Agricole v PPT Energy
Letters of credit (“LCs”) are the lifeblood of international trade. The policy considerations of ensuring that such documentary credits serve the function of cash means that banks must pay against a conforming presentation under an LC; its role limited to dealing with the documents on its face and not the underlying transaction, the underlying goods or the underlying motivations of parties to the trade. Fraud is the one exception to payment under a letter of credit, but the law is clear that only the fraud of the beneficiary can vitiate a presentation. But what about the conduit trader who allows its name to be inserted into a chain for a profit margin – indifferent as to the purpose of the trade, whether the documents or title it is passing through is legitimate. Not fixed with guilt of actual dishonesty but not innocent in allowing itself to obfuscate trading structures, should the law/public policy reward such traders to take part in dubious trading structures while pushing the risk of fraud onto the bank required to pay under letters of credit.
The opportunity to consider this arose in the case of Crédit Agricole Corporate & Investment Bank, Singapore Branch v PPT Trading Energy Co. Ltd.  SGHC(I) 1 (“Credit Agricole”). The Singapore International Commercial Court (SICC) in Credit Agricole looked at various key issues including circular trading, the passing of title in oil trades without original bills of lading (BLs) and the fraud exception when seeking payment under a letter of credit.
The case involved a string of FOB oil trading contracts originating from Total Oil (Totsa) to Zenrock, who then created a circular trade involving Shandong, PPT and back to Zenrock before the goods were then sold back to Totsa (see blue portions in diagram). Zenrock applied for a letter of credit with its bank, Credit Agricole (CACIB), who in turn opened a LC in favour of PPT payable upon presentation of amongst other things, an invoice and a letter of indemnity (in lieu of original BLs). In applying for the LC, Zenrock showed to CACIB its cargo purchase contract with PPT well as a fabricated sales contract with Totsa (the “Fabricated Totsa Contract”). The sales and purchase contracts showed that Zenrock was making a profit and grossly inflated the value of the cargo at a price of Platts plus a premium of USD 3.60 per barrel (when in fact the actual pricing of the cargo at the time was negative during the famous oil crash of March 2020). Unbeknown to CACIB, Zenrock had already financed and sold the cargo to Totsa under its true sales contract at the price of Platts minus USD3.60. By inversing of the price and creating the Fabricated Totsa Contract, Zenrock gave CACIB the impression that the trade flow it was financing was profitable.
PPT presented its invoice and LOI to CACIB for payment under the LC, who did not reject the documents within 5 days as required under the UCP 600. CACIB became suspicious of the possibility of double financing when it received information from Totsa that Zenrock had assigned the receivables from the Zenrock-Totsa contract twice (see section in green).
CACIB then sought an interim injunction to restrain payment of over US$20 million dollars under the LC. The matter went before the SICC with CACIB, alleging that PPT had actual knowledge or were willfully blind to the fraudulent scheme to inflate the value of the cargo and to procure double financing for it.
The SICC found that PPT did not act fraudulently and was entitled to payment under the LC. While PPT knew of the round-tripping, it did not know of the Fabricated Totsa Contract and so could not be said to be dishonest. Critical to the SICC’s findings was that PPT did not know of the inflated pricing at which it was offered the deal by Zenrock (some 87% higher than market value). Notwithstanding some red flags, the Court took the view that a reckless failure to verify representations made in the documents, was not sufficient to amount to fraud so that CACIB could refuse payment under the LC. As to the warranties as to title contained in the LOI, the SICC correctly pointed out that these warranties was not triggered as the language of the LOI expressly stated that these would be given in consideration of payment, which CACIB did not make at the relevant time.
The test for fraud – different horses for different courses?
The case demonstrates the entrenched judicial views on injuncting payments under LCs, with assurance of payment being vital for LCs. Our commentary here is limited to what appears to be different tests for fraud with respect to the element of recklessness but a full analysis can be found here.
In an attempt to reconcile two differing restatements of fraud in previous Singapore cases, the SICC drew an unnecessary distinction between demand guarantees and LCs. Applying the classic formulation of the autonomy principle, the Singapore Court of Appeal in Brody, White and Co v Chemet Handel Trading followed the English law reasoning that it is only fraud in the presentation of the documents that can taint a claim under an LC and not fraud in the underlying transaction. However, this narrower formulation is at odds with another Singapore Court of Appeal case of Arab Banking v Boustead Singapore which held that reckless, i.e., indifference to the truth of statements made can constitute fraud and this exception applies equally to LCs as it would demand guarantees. It is uncontroversial this wider formulation of recklessness applies to the tort of deceit and so, this distinction therefore creates different standards of fraud (and hence different outcomes), depending on when the fraud is discovered – a narrower standard where fraud is discovered before payment is made under the LC and a broader standard that includes recklessness for after payment is made, and the bank seeks to pursue an action against the beneficiary under the common law tort of deceit. We find this distinction unsustainable.
If the fraud exception encompassed reckless conduct, would the outcome be different in the Credit Agricole case? We suspect so: in that regard, the evidence surrounding an earlier attempted trade flow involving Trafigura is telling. PPT’s rationale for its business model as a conduit trader was at odds with the attempted trade involving Trafigura – where PPT did not serve the function of a credit sleeve, provide credit or anonymity. Within minutes of Trafigura pulling out of the deal, Zenrock swapped its role from seller to buyer when dealing with PPT, indicating the synthetic nature of the trade. Perhaps the most damning evidence in the attempted deal was communication from Zenrock’s traders to PPT instructing them not to use Credit Agricole as their financing bank to avoid the full chain being visible.
In an oil trade where no original BLs were to be passed, and where it appears that PPT did not take any steps to verify title or whether it even knew its supplier Shandong, it appears that PPT was indeed indifferent as to whether it was receiving and passing good title. In a pre-structured round-tripping deal where PPT could not justify its purpose in the trade, knew their counterpart wanted to hide aspects of the trade and remarkably did not notice the pricing of its cargo was 87% higher than market prices, should a participant be rewarded for failing to verify and authenticate the trade? For decades, case law has been heavily influenced by the need for LCs to perform its purpose, but this now needs to be counterbalanced against the practice of dubious traders creating synthetic structures and pushing the risk of fraud to the bank as we have seen in the 2020 trade fraud cases.