Non-Liner Mean Reversion in Stock Prices

This thesis provides new evidence on the time-series predictability of stock market returns by applying a test of nonlinear mean reversion introduced by Bali et al., 2008, and also a nonlinear unit root test developed by Kapetanios et al., 2003. Using minimum of daily returns, the predictability of stock market index and decile size portfolios returns have been investigated. The results indicate a negative and significant relation between the monthly returns of the index and the minimum of daily returns during the past 10 to 16 months, demonstrating the existence of nonlinear mean reversion property in Tehran Stock Market. Moreover analyzing the decile portfolios nonlinear regression results shows that increasing the size of the portfolio raises the mean reversion speed, which supports several researcher's arguments. Furthermore, the nonlinear unit root test results on the stock market index compared to critical values, supports the existence of mean reversion in stock market and proves the robustness of the results.