A Study On Asymmetric Volatility Of Chinese Share Market

Since 1990s,the study of financial market volatility is always one of most important field of theoretic and exercise study. Many researcher studied financial market volatility based on different methods. These studies show that the characteristic of financial market volatility is important in investment, security valuation ,risk management and monetary policy, and will affect financial market, then global economy. Forecast of financial market volatility is also becoming one important variable of macro financial policy for most central banks. For instance, the Federal Reserve of US will take the volatility of equity market, bond market, money market and commodity market into decision of monetary policy. Bank of England also take volatility of key financial variables into monetary meetings.The history of Chinese financial market is not as long as foreign market, includes institutional arrangements of financial market, the development of financial product, is in the stage of development. However, with the deepen of social communist market economy, and the further development of financial system reform, financial market will play a more and more important role in Chinese financial system, the role of resource allocation, risk management, and price discovery will be the key function of financial market. But china faces more complex environment of economy and financial development, and more uncertainties. A series of theoretic and exercise problems should be solved when push forward Chinese financial market reform. So it is important to study financial market volatility for theoretic and exercise.Long term memory and asymmetrical volatility are the two basic characteristics of financial market, as the existed researches, this paper will focus on these two characteristics.The aims of this paper are as follows. Firstly, the long term characteristics of equity return serial and conditional heteroscedasticity serial. Secondly, the asymmetrical effect of volatility. Thirdly, comparative studies of asymmetrical effect among different stages.The logistical of the paper is as follows: firstly, literature reviews; then analysis of basic statistical characteristics; thirdly, study of asymmetrical volatility; and fourthly, comparative study of asymmetrical volatility among different stages. The main research methods: For the study of long term memory, the paper uses revised R/S, ARFIMA model and FIEGARCH model; in the study of asymmetric volatility, the paper uses TARCH model, EGARCH model, GJR model and APARCH model.Results and conclusions:The basic statistical characteristics and test analysis. The log-return series of shanghai and Shenzhen stock market are stationary series, and have peak and fat tail. Normality test tells us that the series are not obey normal distribution. The series are not independent and have some degree of series auto-correlation, which can be eliminated by ARMA (1, 1) model. The volatility of Shanghai stock market is larger than Shenzhen, both of them have ARCH effect.Study on long term memory effect of Chinese share market. To study long term effect of the return series, absolute return series and series of squared return, the paper uses revised R/S and ARFIMA(1,d,1) model. Under different distribution hypothesis, ARFIMA (1, d, 1)-FIGARCH (1, d, 1) model give us different conclusion. Provided that return series are normal distribution and the residual series are not heteroscedasticity, the return series of both Shanghai and Shenzhen stock market do not exist long term memory, but series of absolute return and squared return do. When take heteroscedasticity of return series into account, under hypothesis of student distribution, long term memory effect exists in return series and conditional heteroscedasticity, under hypothesis of normal distribution, long term memory effect only exists in conditional heteroscedasticity. Furthermore, the paper compares long term memory effect of different periods of Chinese share market, and concludes that in each periods, there exists different long term memory effect.Study on asymmetric effect of Chinese share market. After introduced some approaches about asymmetric effect test and empirical study, based on basic statistical characteristics, set up empirical models of volatility asymmetric effect. Among these models, we didn't consider memory characteristic in TARCH and EGARCH model, but considered in GJR and APARCH model. For Shanghai stock market, TARCH, EGARCH and APARCH models show asymmetric effect, but the direction in TARCH model is not same, GJR model doesn't show asymmetric effect. For Shenzhen stock market, TARCH model shows asymmetric effect under hypothesis of both normal and Student distributions, but doesn't exist under GED distribution hypothesis, EGARCH, GJR and APARCH models don't show us asymmetric effect. So, there exists asymmetrical effect in Chinese share market, which is the same as the conclusions of foreign markets.Comparative Study on asymmetric effect of different period of Chinese share market. According development of institutional, Chinese stock market can be divided into 3 stages, then compare the asymmetric effect of each stage. The study tells us that there exists significant difference among each stage. Firstly, before price volatility limit arrangement, asymmetrical effect didn't exist both in Shanghai and Shenzhen stock market. Secondly, there existed distinguished asymmetrical effect before the reform of non-tradable share, but did after the reform. Thirdly, before non-tradable share reform, the direction of symmetrical effect is the same, negative information will led more volatility than positive information. Fourthly, in bull market of prior non-tradable share reform, asymmetrical effect didn't exist, but after reform, which existed in both markets, positive information led more volatility. Fifthly, in bear market, asymmetrical effect existed in both Shanghai and Shenzhen market before and after non-tradable share reform. Sixthly, before non-tradable share reform, asymmetrical effect was decided by bear market, after non-tradable share reform, because the direction was completely adverse, so there didn't exist asymmetrical effect in the whole period.