عنوان مقاله [English]
In this Thesis we examine the dynamic relationship between stock returns and mutual fund flows in Tehran Stock Exchange (TSE) by using VAR model. Afterthat, we checked the impulse response function and forecast error variance decomposition of VAR by Cholesky and generalized function. We find that spillover shocks that is, Tehran Stock Exchange main index (TEDPIX) return shocks and mutual fund flow shocks together explain some percent of the total forecast error variance of stock returns and mutual fund flows. Base on above mentioned results we used Aritificial Neural Network (ANN) with different learning functions for exmaning the relationship between Tehran Stock Exchange main index (TEDPIX) return and mutual fund flow. For selecting the best learning function in ANN, mean square Error has been used. For statistical siginificance, T-statistical paired comparison test was used. We create a spillover index of shocks emanating from stock returns and mutual fund flows and tests whether it can actually predict Tehran returns. We find it does. Using the spillover index, we forecast TEDPIX returns. At the end because of endogeneity, persistency and heteroscedasticity of predicting regression, we used Feasible-Quasi Generalized Least Square (FQGLS) for examining the statistics significance of spillover index, which turns out to be statistically significant.
JEL: G17, G23
How to cite this paper: Taiebysani, E., & Fallahpour, S. (2017). Presenting a Model for Measuring Predictability Strength and the Relationship of Stock Index Return and Mutual Fund Flow. Quarterly Journal of Risk Modeling and Financial Engineering, 2(3), 297–319. (In Persian)