FACULTÉ & RECHERCHE

This article describes an analytical approach to examine the credit-risk behavior of a homogeneous portfolio. The authors demonstrate the usefulness of the approach using a synthetic index linked to high-yield corporate bonds (which resembles a synthetic CDO) and then analyze an actual synthetic CDO transaction. They show that the conventional approach to analyze these structures (Monte Carlo simulations combined with the Gaussian copula) fails to account for the tri-modal nature of the underlying portfolio default distribution, and consequently, risk assessments based on this method give a misguided view of the risk–reward profile of such portfolios. The authors further show that the benefits of portfolio diversification in the context of credit-risk portfolios are limited in high-correlation scenarios. These findings have important implications for risk managers and financial regulators.