Bankruptcy Prediction Models and the Cost of Debt
William F. Maxwell
SMU - Cox School
Andrew (Jianzhong) Zhang
University of Nevada, Las Vegas - Department of Finance
June 8, 2010
Journal of Fixed Income, Forthcoming
Financial institutions and academic researchers utilize bankruptcy prediction models to assess distress risk. However, predicting default can be problematic since (i) few firms actually experience default in any one year, (ii) the lag between practical and actual default can vary significantly, (iii) firms can strategically default, (iv) firms can rework their obligations outside of bankruptcy, and (v) default frequency varies significantly over economic life cycles. Thus, relying on bankruptcy data alone to calibrate and validate these models can be problematic. We take a simpler approach by relying on the firm’s cost of debt as a market proxy for distress risk. We then assess the validity of four widely used bankruptcy models including two accounting-based models (Altman’s, 1968; Ohlson’s, 1980), one reduced form model (Campbell, Hilscher, and Szilagyi, 2010) and one structural distance to default model (Merton, 1974). We find dramatically different assessment of risk based on the models used. The Campbell, Hilscher, and Szilagyi (2010) model has the most explanatory power on the cost of debt followed by the Merton model. The accounting based approaches of Altman (1968)’s Z-Score and Ohlson (1980)’s O-Score are highly ineffective. We caution researchers when using Z- and O-Scores and recommend the use of Campbell, Hilscher, and Szilagyi model to measure distress risk. We also demonstrate the problems of not controlling for industry and time variation in any of these measures.
Number of Pages in PDF File: 34
Keywords: Bankruptcy prediction models, cost of debt financing, distress risk
JEL Classification: C52, G13, G33, M41
Date posted: June 9, 2010 ; Last revised: February 28, 2012
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo6 in 0.328 seconds