44 Pages Posted: 28 Aug 2012 Last revised: 16 Feb 2017
Date Written: February 12, 2017
Recent LIBOR manipulation scandal inspired discussions of the proper design of survey-based benchmarks. I solve a version of the survey as a game of incomplete information in which the benchmark administrator is unaware of the distribution of private signals. I characterize a standard survey in which the expected benchmark bias is distribution-free and corrigible. I then study the LIBOR benchmarking as a particular case of standard survey. Banks’ reporting errors decrease with the panel size and the number of quotes used for calculation. The expected benchmark bias is no longer distribution-free under collusion or signaling, but embargo fixes this problem.
Keywords: LIBOR; Interbank lending; Benchmarking; Strategic reporting; Mechanism design
JEL Classification: D43, D44, D47, G10, G21
Suggested Citation: Suggested Citation