Essays on term structure models

Abstract

Estimating risk premia has been at the forefront of the financial economics’ literature due to their informational content. Risk premia are of particular interest to academics, policymakers and practitioners given the information they disclose on expected asset returns for a given level of risk, their contribution in asset pricing and their ability to disentangle the different sources of risk. However, risk premia are unobserved and their estimates strongly differ from one study to another, as they are highly sensitive to the specification of the underlying model, sparking hence a strong interest in their analysis. The aim of the thesis is to estimate risk premia in a dynamic term structure model setting. The first part of the thesis comprises of an overview of a particular class of dynamic term structure models, namely affine term structure models. The overview will include important concepts and definitions. The second part of the thesis uses a risk-averse formulation of the uncovered interest rate parity to determine exchange rates through interest rate differentials, and ultimately extract currency risk premia. The method proposed consists of developing an affine Arbitrage-Free class of dynamic Nelson-Siegel term structure models (AFNS) with stochastic volatility to obtain the domestic and foreign discount rate variations, which in turn are used to derive a representation of exchange rate depreciations and risk premia. The third part of the thesis studies both the nominal and real UK term structure of interest rates using a Gaussian dynamic term structure model, which imposes the non-negativity of nominal short maturity rates. Estimates of the term premia, inflation risk premia and market-implied inflation expectations are provided.