Using proxies for the short rate: when are three months like an instant?

The dynamics of the unobservable short rate are frequently estimateddirectly using a proxy. We examine the biases resulting fromthis practice (the 'proxy problem'). Analytic results show thatthe proxy problem is not economically significant for single-factoraffine models. In the two-factor affine model of Longstaff andSchwartz (1992), the proxy problem is only economically significantfor pricing discount bonds with maturities of more than fiveyears. We also describe two different numerical procedures forassessing the magnitude of the proxy problem in a general interestrate model. When applied to a nonlinear single-factor model,they suggest that the proxy problem can be economically significant.