Designing a Model for Projection of Tehran Exchange Return Employing Autoregressive Moving Average (ARMA) and Autoregressive Moving Average with External Inputs (ARMAX) Models and Assessing the Performance Thereof

Capital markets’ return is influenced by multiple and various factors. These factors range from macro global factors to dependent variable historical behaviors. Each of many researchers has selected a segment or segments of this vast range of factors impacting capital market return in different countries and embarked on devising a model for projection of the respective capital market return. The present research has utilized self-explanatory and combined models for the purpose of modeling and projecting Tehran Exchange’s return using Autoregressive Moving Average (ARMA) and Autoregressive Moving Average with External Inputs (ARMAX) models. In this research, to utterly expound the model and to factor in maximum factors, following examination of return issues and the factors impacting return, the subject of projection and the common methodology thereof have been surveyed and different models of projecting capital market return have been examined in detail. Then, Classic Linear Regression Models, Autoregressive Moving Average (ARMA) and Autoregressive Moving Average with External Inputs (ARMAX) models were used to predict the return of Tehran Exchange. After estimating the above models, deploying 99-period data and confirming the power of expressing them via applying diagnostic tests, the return of Tehran Exchange for the next 4 periods was projected. These predictions by applying estimation models were compared with the real data and the optimum model was selected using Akaike Information Criterion (AIC), Schwartz-Bayesian Information Criterion, Hannan-Quinn and also MSE, MAE, MAPE criteria. The final result indicates superiority of ARMA over ARMAX