Jump Behaviors And Volatility Modelling

Lee and Mykland(2008) nonparametric jump test is utilized to study the jump behavior of China’s stock index future market. Weighted Standard Deviation Periodicity Factor is used to correct the test for ignoring the intraday deterministic diurnal volatility pattern. Jump behavior of stock index future market is researched thoroughly. Stock index future market is found more sensitive to information shock and produce more jump than the spot index. Comparing with the spot index before Stock index futures is list, the spot index after Jumps more frequently, but mainly jumps up and the frequency of jump down barely rise. No matter viewing from the statistical characteristic of5minutes yields (the standard deviation, extreme value, skewness and kurtosis of the yields), or from characteristic of the jumps (the mean, standard deviation, extreme value of the jump range), the volatility and extreme tail risks of the HS300index after the listing of the index futures are significantly lower. The extreme tail risk of short-term havoc is reduced, especially the slump situation.Anyway, the listing of the index futures make the HS300index moves more smoothly. Different from foreign research conclusion, the jump of China’s stock market has little connection with macro information released.According to the WSD Periodicity Factor, the volatility of stock index futures and spot index are found to have calendar effect. Influenced by information over the weekend, volatility is significantly larger at the opening of Monday morning. The intraday volatility of stock index futures is in a double "W" shape, and the spot index intraday volatility is roughly in inclined inverted "L" type.Through the Analysis of yields after the jumps, the jumps with different directions and different sizes are found to have different behavior characteristics after the jump. Significantly negative jump will present reverse effect, and show the momentum effect by the following day; while significantly positive jump will show reverse effect that day and the next day. And moderate jumps have the biggest influence on the volatility after the jumps.According to different influences of different jumps on the volatility, and the short memory features of volatility in China’s stock index future market, a new model of HAR-RJ-MR (heterogeneous market model) is built from the perspective of behavioral finance. Significant leverage effect is found in china’s stock index future market, meaning that the negative jump has significantly greater influence on the volatility after than the positive one. And scale effect is not significant, but the moderate jumps have the biggest influence on the volatility. Compared with the general HAR-RJ models, HAR-RJ-MR has a better explanation and a better goodness of fit.