Friday, 21 October 2011: 8:30 AM
In this paper the term structure of LIBOR US dollar interest rates is examined with the use of regression based tests. Following the methodology of Pagan and Ullah (1988) that allows for modeling risk terms a conventional perfect foresight spread (PFS) equation is augmented by a proxy of the time-varying term premium and estimated by the GMM. The monthly sampled data span from January 1987 to December 2009 and cover the whole spectrum of maturities ranging from 1 to 12 months. Having run the restricted PFS regression on the term spread only it is found that the hypothesis stating the slope of the term structure equals unity as predicted by the expectations hypothesis is not rejected for all maturities but that of 2 month. It is rejected for all maturities under 8 months, however, while allowing for the term premium to vary across time and approximating it by the volatility of one month rate.