Usefulness of accounting information vs. market information in bankruptcy Prediction

Abstract

The usefulness of accounting information has always been source of concern for standard setters that view the main goal of financial reporting as the issue required for generating useful information to facilitate decision making process according to conceptual framework. Predicting bankruptcy is one of the main areas in which researchers have studied the usefulness of accounting information. However, to assess the usefulness of accounting information in current condition, it is necessary to compare it’s predicting ability in relation to other information resources such as market information. To this end, using genetic algorithm, models with different mixture of accounting and market variables are designed and tested. To improve the generalization of the results, cross-validation is used. The samples used in this study include the companies listed in Tehran Stock Exchange between 2001 to 2009. The fiundings show that accounting information is more useful in predicting bankruptcy than market information, and that market variables are not a good complement to accounting information in bankruptcy prediction.

Keywords

• key words: accounting information
• market information
• bankruptcy prediction
• genetic algorithm
• cross-validation