 |
An International Meeting of American Accounting
Association
|
The Impact of Incorporating the Cost of Errors into Bankruptcy Prediction Models |
Lawrence A Weiss
HEC - University of Lausanne
Vedran Capkun
HEC - University Of Lausanne
Abstract: The current methodology to use and evaluate default and bankruptcy prediction mod- els is to determine their precision - the percentage of ¯rms predicted correctly. In this study we develop a framework for incorporating Type I (the amount lost from lending to a ¯rm which goes bankrupt) and Type II (the opportunity cost of not lending to a ¯rm which does not go bankrupt) error costs into the prediction models and their evaluation. Our results indicate that a lending model which accounts for the cost of errors and ¯rm size yields higher pro¯ts than a model relying only on precision. This also supports our hypothesis that the usefulness of prediction models cannot be fully assessed independently of the costs of both types of forecast errors.