Predicting Fraud in Financial Statements(Time Varying Parameter Dynamic Model Averaging)

Financial statement fraud has become a serious problem for market participants and policy makers. In fact, it threatens the reliability of capital markets, corporate executives and even the auditing profession. The purpose of this study is to use the approach of dynamic averaging models to predict fraud in financial statements.The present research is applied in terms of method. The research period is 1390 to 1399 and in estimating the model, the data of selected companies in Tehran Stock Exchange has been used.Using the systematic elimination approach, the research sample size of 125 companies was selected. To estimate the model, MATLAB 2021 software has been used.In this research, based on the dynamic averaging model, we predicted the fraud and accuracy of the estimation models. Based on the results of asset return variables; Return on equity; Operating profit margin; Asset turnover ratio and operating cash-to-sales ratio have a negative effect on fraud and other variables have a positive effect.