Econometric analysis of European LIBOR-based options within affine term-structure models

This thesis attempts to bridge the gap between latent-factor models used by academic researchers and curve models used by market makers and risk managers in the LIBOR/Swap cash and options markets. The first group of models has consistent time-series properties but models of this category have various difficulties adapting to the instruments traded in the LIBOR/Swap market. The second group is designed specifically to easily price/calibrate to such derivatives, but they lack time-series consistency, are uninformative about the nature of risk premiums, and typically impose the counterfactual assumption of constant or deterministic volatility. This thesis develops a framework for pricing LIBOR-based derivatives that preserves temporal consistency of the first approach, while accommodating stochastic volatility and allowing identification of time-series of factor risk premiums. In this thesis we develop a series of techniques that greatly simplify estimation of multi-factor Affine Term-Structure Models (ATSMs) on the time series of Interest-Rate options. With these techniques in hand, we obtain maximum likelihood estimates of several 2- and 3-factor affine models on the weekly time series of US{dollar} LIBOR/Swap rates and Swaption volatilities. Time series of model-implied risk premiums and stochastic volatilities are extracted and examined.