Modeling of Accounting and Non Accounting Items Affecting Shareholders, Wealth: Prediction and Validation

The Stock market is one of the markets from which investors try to earn interests. Stock returns are the most important measures for decision making of investors in this market. This study develops a model for predicting stock returns using financial statements, disclosure quality, corporate governance mechanisms, audit quality, stock market qualification and macroeconomic items. Data is collected from disclosures of firms listed in Tehran Stock Exchange during a period from 2004 to 2012. Based on this model, net income, price index for consumer goods and services, dividend, oil price index and gold coin price index are the most important variables effective on predicting stock returns. To determine the developed model’s accuracy and validity, stock returns are predicted for sample firms (86 firms) for years from 2013 to 2017 and then errors of predicted returns as compared to real returns for those periods are determined. Furthermore, other prior valid predicting models (capital assets pricing model (CAPM) , Fama and French three factors model and arbitrage pricing model) errors are calculated for that period. Comparing developed predicting model errors with other predicting models errors shows that this model has significantly fewer errors in predicting future stock returns than the prior models.