Shock modeling of influencing variables on stock return forecasting with the approach of BMA-BVR models

The purpose of the research is to predict stock returns using Bayesian averaging and BVAR. The current research is based on the applied research method and MATLAB 2021 and EVIEWS12 have been used to estimate the model. The time period of the research includes the years 2010 to 2019. First, 11 non-fragile variables out of 64 entered variables were identified with the Bayesian averaging model approach. Based on the results of the current ratio; ROE; P/E; oil revenue; The increasing coefficient of money in the whole period has a positive effect and inflation fluctuation variables; debt ratio; fluctuation of GDP growth; unofficial market exchange rate; Interest rate and systematic risk have a negative effect on yield in the whole period. Based on the results of variance analysis, the most explanatory of changes in stock returns is caused by the variable itself (20 percent), followed by interest rate variables (14 percent); Inflation volatility (13 percent) and debt ratio and systematic risk (10 percent) have the highest effect in explaining yield changes.