Modeling and Comparison of the Distribution Models of Tehran Stock Exchange Index

Study of the Extreme behavior of the Stock Market and the correct pattern of the distribution of returns has an important role in the management of market risk. The current study, based on BMM approach, examines the distribution pattern of return in Tehran Stock Exchange in different time intervals. In order to choose the appropriate model in different time series, L-Moment ratio diagram was used. Then, by using the calculation of parameters of the selected models based on the approach of Maximum likelihood, the conformity and selection of appropriate model in each time interval was performed based on AD test. To model the distribution of patterns of total market returns, the total stock index data of 2004-2016 was used and the basis of the performed calculation has the log of daily return of Tehran stock exchange. The results indicated that among the examined model, GL model in annual time interval and in minimum series, and GEV model in other time intervals of minimum and maximum series had a better performance.