A New Approach to Predicting Fraud in Financial Statements (Compare models and simulations traditional and modern)

Financial statements fraud is increasingly a serious problem for businesses, governments and investors have become. In fact, the issue of the reliability of capital markets, bosses and even threaten the auditing profession. Auditors in particular face their apparent inability to detect large-scale fraud, and therefore various methods have been proposed to identify this problem. The aim of this study was to compare different models of traditional and modern simulation and predict fraud in the financial statements. Study the practical approach. Research period 1390 to 1398 and the model of the selected companies in the Tehran Stock Exchange has been used. In this study, based on three traditional approach, genetic algorithms and methods to predict non-linear Markov switching models have estimated fraud and accuracy.  Based on the prevailing regime in Tehran Stock Exchange regulations fraud is high; According to the results of highest accuracy in predicting fraud regime change and genetic algorithms and logistic highest and the accuracy of the model, respectively.