Structural models and the prediction of default probabilities

In this paper, the three main structural models of default, the Merton model, Longstaff and Schwartz model and Leland and Toft model, are compared in terms of the real default probabilities they produce. I find that none of the models can accurately predict the default probabilities in all cases. Merton as well as Leland and Toft models underpredict default probabilities in all cases. The Longstaff and Schwartz model, although in some cases it produces EDFs that are close to the observed ones, suffers from important limitations. The model tends to overestimate the default probabilities of riskier bonds as well as the default probabilities of bonds with the same rating but higher equity volatility. Consistent with previous studies, I find that structural models tend to underestimate the default probabilities in early years.