Identifying and Modeling Tehran Stock Market Calendar Effects: Using ARCH and GARCH Models

This paper examines the calendar anomalies in daily return of the Tehran stock market. ARCH and GARCH models are employed to capture the wide range of different calendar anomalies exist in the literature. This study finds the evidence of strong Esfand and Mehr effects in the stock return. In addition, the results show that the stock market return has decreased with the lapses of time. After identifying and removing the calendar effects from daily return, BDS statistic is used to test the presence of any remaining non-linearity in the residuals before employing the GARCH models. The BDS test shows that there is a high probability of the dependency between residuals in spite of removing calendar anomalies. The results confirm that both the ARCH and GARCH models have considerable success in modeling dependencies. Finally, the importance of calendar effects in return forecasting is tested. The conclusion is that the inclusion of calendar effects improves the forecast accuracy. However, simple regression which includes calendar effects has better performance than the GARCH (1, 1) models.