Purpose – Most of the credit spread literature deals with the determinants of credit spread changes for individual bonds. The purpose of this paper is to investigate the explanatory power of credit spread changes and their determinants for portfolios. Design/methodology/approach – Using ordinary least squares (OLS) regressions and monthly data from 1990 to 1997, this paper tests several new potential determinants (e.g. portfolio diversification) and expectations (and realizations) for some previously identified determinants (e.g. gross domestic product (GDP)) of credit spread changes for portfolios of financials as derived from spot curves. Findings – Strong empirical support is reported that default risk and undiversified risk are priced in credit spreads. The paper finds that forecasts for GDP and inflation are better determinants of credit spread changes than the realized values previously used in the literature, which is consistent with the notion that term structures convey expectations about future interest rates. Research limitations/implications – Interesting issues for future research include the sensitivity of the results to the use of other procedures for deriving zero-coupon spot rates, and whether forecasts of macrovariables (such as GDP) are better determinants of credit spreads for other industrial categories, such as utilities and industrials. Practical implications – The findings provide guidance for the management of risk for fixed income portfolios, for the pricing of fixed income securities differentiated by the difficulties encountered in achieving well-diversified portfolios, and for assessing the performance of credit spread portfolios managed by financial institutions. Originality/value – The empirical model, which achieves substantial explanatory power while being parsimonious, is the first to support the usage of forecasts instead of realized values in determining credit spreads, and to show that undiversifiable risk is an important component of the credit spreads of portfolios.
