Comparing the Ability of Bayesian Networks and Adaboost for Predicting Financial Distress of Firms Listed on Tehran Stock Exchange (TSE)

Abstract

Financial distress and bankruptcy of companies may cause the resources to be wasted and the investment opportunities to be faded. Bankruptcy prediction by providing necessary warnings can make the companies aware of this problem. The aim of this study is to compare the ability of Bayesian networks and adaboost for predicting financial distress of firms listed on Tehran Stock Exchange (TSE). Two naive bayes models were developed based upon conditional correlation between variables and conditional likelihood. The accuracy in predicting bankruptcy of the first naive bayes model's performance that is based upon conditional correlation is 90% and the accuracy of the second naive bayes model is 93% and finally the accuracy of the adaboost that was built to compare with naive bayes models is 88%. Collectively the results show that it is possible to predict financial distress using Bayesian and Adaboost models. But, Bayesian networks are more capable to predict financial distress of companies listed on TSE compare to Adaboost. With respect to the variables in developed models in this research we find that firms with lower profitability and more long term liabilities and lower liquidity are more in Risk of financial distress. To reduce financial distress risk, firms should use more conservative methods which lead to decrease in debts and reduce their costs.