TY - JOUR T1 - From Ad Hoc Bond-Risk Measures to Variance–Covariance Forecasts JF - The Journal of Fixed Income SP - 6 LP - 16 DO - 10.3905/jfi.2021.1.105 VL - 30 IS - 4 AU - Marielle De Jong AU - Frank J. Fabozzi Y1 - 2021/03/31 UR - https://pm-research.com/content/30/4/6.abstract N2 - It is common practice among investors to assess the risks of a fixed-income investment by looking at certain bond characteristics—and at two in particular. The duration provides insight on a bond’s contribution to the interest-rate risk of a portfolio, and the credit spreads provide information about credit risks. Linking these measures together in order to obtain one forecast for the overall price behavior of a bond and bond portfolio is not trivial, however. Although the popular method of taking products, the so-called duration-times-spread rule, is effective and superior to using spread duration, it stops short of delivering complete risk estimates. In this article, the authors propose a method that does provide a complete risk estimate and describe how bond characteristics can be converted into return variance and covariance forecasts.TOPICS: Fixed income and structured finance, risk management, portfolio construction, statistical methodsKey Findings▪ This article proposes a method to convert the durations and credit spreads of bond portfolios into return variance and covariance forecasts.▪ The conversion method is a direct extension of the duration-times-spread rule.▪ The method helps bridge the gap between front-office portfolio management tools and back-office risk control systems. ER -