Expected loss and fair value over the credit cycle

Abstract

We present an easily applied method of risk-adjusting reduced-form models for changes in systematic risk over the credit cycle. Using an empirical approach, we model the probable changes in systematic risk over time, showing that investment-grade portfolios that are naive to changes in levels of systematic risk can significantly underestimate expected losses. We build a valuation model that risk-adjusts credit spreads for probable future levels of systematic risk. The model is used to construct low- turnover portfolios, and it is shown to outperform both a naive approach and also the Global Lehman Credit Index for the period 1981–2003.