Calibration of the Structural Model of Corporate Bond Spreads

It has been long recognized that endogenous default probabilities cannot explain spreads between corporate and the riskless bonds. Recently, this issue has been subjected to rigorous scrutiny. Previous studies have found that for investment-grade debt, structural models explain only 15-25% of the observed spreads. On the other hand, for the high-yield debt, the structural models exaggerate actual spreads 1.5-2 times. These findings are perplexing because, while one could argue that factors other than default risk, e.g. illiquidity, can influence the spread for investment-grade bonds, it is difficult to justify the findings for junk bonds. In this paper, we offer an explanation to these puzzling results. Specifically, we account for the differential tax treatment of regular income and capital gains. In addition, we consider the uncertainty of the residual assets due bondholders, as well as the tax liability of the residual asset. We argue that the uncertainty in the claims on the future assets of the company in the case of potential default drives the spreads of an investment-grade debt as much as, if not more than, the probability of default.