Introducing an Early Warning System for High Volatility in Tehran Stock Exchange: Markov Switching GARCH Approach

The goal of this paper is to introduce a new model to predict the high volatility of Tehran Stock Exchange. For do it, a Markov switching GARCH models was modeled. With Estimating this model, the transition probability matrix of two states of high and low volatility, was calculated. Using this matrix, we can forecast the probability of market fluctuations in the each period ahead and we can obtain a suitable model for forecasting high volatility. According to the model selection criteria consist of AIC and BIC, the Markov regime switching GARCH model with GED distribution is the best model for forecasting volatility in Tehran Stock Exchange. Based on this model, in this paper, an Early Warning System has been introduced in Tehran Stock Exchange. This model can be used for policy makers to prevent the occurrence of high volatility and to increase the security of investors in Tehran Stock Exchange.