Letters of credit as signals.

• Author: Gillette, Clayton P.
• Position: Response to article by Ronald J. Mann in this issue, p. 2494

Comments on Ronald Mann's `The Role of Letters of Credit in Payment Transactions'

Why would buyers and sellers transact with each other through a third party that charges a significant fee for its services and that typically is authorized to make payment notwithstanding noncompliance with the very prerequisites that it has been engaged to monitor? This is the puzzle that Ronald Mann's provocative and nuanced article purports to explain.(1) Under the traditional story about the esoteric world of letters of credit, these transactions allow distant buyers and sellers to circumvent obstacles that would otherwise frustrate long-distance transactions. The traditional story explains that these credits induce buyers to approve payment prior to receiving conforming goods because the transactional structure provides buyers with documents that testify conforming goods are en route. Similarly, credits induce sellers to ship goods prior to payment because that same transactional structure assures that payment is forthcoming from a credible and credit-worthy source.(2) Mann suggests that the story is incorrect or at least incomplete. For him, the real function of the letter of credit is to solve informational asymmetries concerning the parties involved in the transaction by allowing an issuer with superior information to verify a buyer's legitimacy to the distant seller or to the buyer's government.

I find the claim that letters of credit fill informational gaps highly plausible. Indeed, I take it to be wholly consistent with the traditional story that banks are asked to issue letters of credit because they have an informational advantage about the financial status of their customers, the applicants. It is less clear to me that what the beneficiary learns from the issuer's conduct should be vaulted into a primary explanation for letters of credit. I want, therefore, to raise two issues with respect to the story that Mann tells us. One issue concerns his substantive claims about the role of letters of credit. The second is more directly related to the general theme of this conference and reflects on how Mann's article may be refined, validated, or refuted by empirical work that has not yet been done.

Let me begin with the substance of Mann's claims. Mann first informs us that discrepancy rates in letter-of-credit transactions belie the traditional story that payment is made only when there is strict compliance with the documents required under the credit. In fact, he discovers, discrepant documents are not only common, but are also typically waived. These are fascinating findings, notwithstanding that Mann indicates that the phenomenon is well appreciated by those in the letter of credit industry.(3) At first glance, they seem wholly inconsistent with the traditional story, which appears to rely so heavily on seller compliance with the exacting terms of the credit.

On reflection, however, we should perhaps be less surprised at the frequency of discrepancies and less certain of what their existence signifies. In the first instance, the level of compliance required for payment may be more flexible than its textbook statement suggests. Mann reviews international transactions, presumably governed by the International Chamber of Commerce Uniform Customs and Practices. That document provides as the standard for compliance not the "strict compliance" language of the U.C.C., but rather a requirement that "[c]ompliance of the stipulated documents ... be determined by international standard banking practice...."(4) The U.C.P. standard, therefore, arguably provides more flexibility than the U.C.C.'s requirement that an issuer honor a presentation that appears "on its face strictly to comply with the terms and conditions of the letter of credit."(5) One might predict that the vague standard of the U.C.P. would be more susceptible to claims by issuers that discrepancies exist. In close cases, issuers have incentives to classify documents as discrepant rather than complying because their conservative conclusion can easily be overridden by a customer where the customer considers discrepancies to be minor.(6) Conversely, an issuer that liberally interprets the requirements under the credit risks complaints from a customer with whom the issuer has an ongoing relationship and who regrets that the bank has paid. Of course, that same narrow interpretation of compliance will cause similar arguments between banks and beneficiaries who contend that the variations do not constitute discrepancies under Article 13. But presumably banks would prefer the latter arguments to the former, because they want to obtain reimbursement from their customer and they have more at stake with respect to repeat business with the customer than with the foreign beneficiary.

But even if we anticipate discrepancies, does it follow that the letter of credit transaction is not intended primarily as a mechanism for assuring payment? Let me suggest an explanation more in keeping with the traditional story. Keep in mind that while Mann asserts that the traditional story involves assurance for sellers that payment is forthcoming, that account provides a parallel assurance for the buyer -- the documents indicate that goods conforming to the contract have been shipped. This assurance is what ultimately makes the letter of credit beneficial to both parties, as opposed to a mere guaranty of payment, which would benefit the seller without providing any assurances of product quality to the buyer.

One would imagine that the great majority of transactions are completed without any deviation from expected product quality, even if there is a discrepancy in the underlying documents. Indeed, that is precisely what Mann discovers. The great majority of discrepancies he reports appear to entail little risk of seller default on the quality of the goods in the underlying contract. Departures from expected product quality are the least frequent discrepancies that he reports.(7) But if that is the case, then neither the existence nor the honor of discrepant documents should be surprising. From the buyer's perspective, there is little need to insist on strict compliance with the technical terms of the credit as long as the discrepancy does not portend ultimate delivery of nonconforming goods. From the seller's perspective, once it is confident that it has shipped conforming goods, there is little need to invest much in ensuring strict compliance as long as minor discrepancies can be corrected and documents resubmitted in the rare case of an intransigent buyer.(8)

While this scenario suggests that applicants and issuers do not always rely on the rights that they have to delay payment and that beneficiaries do not heavily invest in ensuring conformity to obtain payment, it does not suggest that those rights are of insufficient...