Modeling the Term Structure of Interest Rates: Where Do We Stand?

Abstract

No-arbitrage term structure models are becoming increasingly important to policy makers and practitioners alike. Several factors justify this trend. First, modeling progress has been tremendous over the last years, allowing a much better fit of actual yield curve dynamics and increased model realism (see Dai and Singleton (2002a,b)). Second, increases in computing power allow the efficient panel estimation of term structure models. Given that term structure models have implications for both the cross-section and time series dimension of yields, panel estimation techniques are to be preferred over either crosssection or time series techniques. Third, term structure models have recently been extended in ways that are of direct interest to policy makers. Example given, Dewachter and Maes (2001) model the international term structure of interest rates, taking into account the role of the exchange rate in a no-arbitrage economy, while amongst others Hordahl et al. (2002) and Dewachter et al. (2002) jointly model the term structure of interest rates with the dynamics of macroeconomic variables. The latter approach allows to study (i) the driving factors behind the term structure and the risk premia in terms of clearly interpretable macroeconomic variables and their determinants, and (ii) the effects of monetary policy on the term structure of interest rates and macroeconomic variables within a consistent no-arbitrage framework.