A Copula Approach to Value-at-Risk Estimation for Fixed-Income Portfolios

This paper introduces a multivariate copula approach to Value-at-Risk estimation for fixed income portfolios. Using a parsimonious model to extract time-varying parameters used as proxies for factors affecting the shape of the yield curve, and a Student copula to model the dependence structure of these factors, we are able to generate VaR estimates that strongly dominate standard VaR estimates in formal out-of-sample tests. A revisited version of this paper was published in the Summer 2007 issue of the Journal of Fixed Income.