Measuring Risk in Fixed Income Portfolios Using Yield Curve Models
28 Pages Posted: 18 Aug 2013
Date Written: June 3, 2013
We propose a novel approach to measure risk in fixed income portfolios in terms of value-at-risk (VaR). We use closed-form expressions for the vector of expected bond returns and for the covariance matrix of bond returns based on a general class of well established term structure factor models, including the dynamic versions of the Nelson-Siegel and Svensson models, to compute the parametric VaR of a portfolio composed of fixed income securities. The proposed approach is very flexible as it can accommodate alternative specifications to model the yield curve and also alternative specifications to model the conditional heteroskedasticity in bond returns. An empirical application involving a data set with 15 fixed income securities with different maturities indicate that the proposed approach delivers very accurate VaR estimates.
Keywords: backtesting, dynamic conditional correlation (DCC), forecast, maximum likelihood, value-at-risk
JEL Classification: C53, E43, G17
Suggested Citation: Suggested Citation