Abstract

The investigations into the London Interbank Offered Rate (LIBOR) have highlighted that it is subject to manipulation. We examine a new method for constructing LIBOR that produces an unbiased estimator of the true rate. LIBOR itself is based solely on transactions. We allow for fines when a bank’s transaction is different than a comparison rate, which depends on the set of transactions and non-manipulated rates elicited by a revealed preference mechanism. These non-manipulated rates will always be used in the fines, but transactions may not. We address how this approach applies to potential replacements for LIBOR and other financial benchmarks, and how it works even in markets in which there are few transactions.

You do not currently have access to this article.