Credit spreads between German and Italian sovereign bonds: Do one-factor affine models work?

In this paper we analyze the credit spread between Italian and German gover nment bonds after the exchange-rate agreement in May 1998. We estimate the parameters of two mean-reverting affine models for the German term structur e and the spread process-the Gaussian Vasicek and the square-root Cox-Inger soll-Ross (CIR) model. Similar to Pearson and Sun (1994) we combine cross-s ectional and rime-series information of daily observations to estimate the process parameters employing a maximum likelihood method. Our empirical res ults show that the Vasicek and CIR model describe the German term structure dynamics equally well. Both models fail to account for all observed shapes of the credit spread structure whereas the spread residuals in the Vasicek case seem to be less volatile. Our results suggest application in the area of pricing credit-sensitive instruments such as credit derivatives or the management of credit risk, especially for European government debt.