| Financial market volatility has always been the focus of the study object, since the1990s, many scholars have done a study of the volatility. Their findings show that the investment and monetary policy are subject to the impact of financial market volatility characteristics. Besides, volatility has a very big impact on the financial markets, even can have significant impact on global economy. Therefore, in the measure of financial markets and economic vulnerability, tend to the volatility of financial markets as an important indicator.China’s futures market has20years of history, but there is still not perfect. Futures market mechanism of the system still has a lot of unreasonable place, still need to continue to improve; Futures products development, compared with advanced foreign futures market, also has very big disparity. With the further advance of China’s financial system reform, and gradually deepening of China’s socialist market economic system, China’s futures market in the role played by more and more important in the financial system. Although futures market has the very big development potential in China, but face the test of futures market development in our country is very much also. First of all, the view from China’s specific national conditions, China’s social and economic uncertainties are more and more complex and changing environment; Secondly, from the development experience of the foreign futures market, the development of the futures market in many countries have faced a variety of variouskind of obstacles and difficulties, their development is not easy. Therefore, the development of China’s futures market, certainly will not be easy. In order to make our country’s futures market continue to in-depth development, a lot of practice and theory must be effectively solved, one very important aspect of China’s futures market is volatility asymmetry theory.The aims of this paper are as follows.Firstly,basic statistics of the yield.Mainly include stationary test, normal test, autocorrelation test, and autocorrelation correction, heteroscedasticity test. Secondly, the asymmetrical effect of volatility. That research on the positive impact of bad news on the futures prices, including EGARCH model and GJR-GARCH model. That is good and bad news on the futures prices, including EGARCH model and GJR-GARCH model.The logistical of the paper is as follows:firstly,overview on GARCH Model and the classic study of the leverage effect model.Secondly, leverage effect research at home and abroad. Thirdly,basic statistical analysis of China’s futures market yields.Finally, leverage effect of China’s futures market. This article in combing questions about financial market volatility, on the basis of research literature, summarized research has limitations and shortcomings, as the breakthrough point of the thesis.The research method of this paper:combining with the characteristics of the development of futures market in China,firstly analyzes the basic statistical characteristics of China’s futures market yields, Then based on the basic statistical characteristics, respectively in normal distribution, GED distribution and student-t distribution assumptions, using EGARCH model and GJR-GARCH model, studies our country futures market volatility of leverage.Results and conclusions:The basic statistical characteristics and test analysis. First calculate the copper and aluminum futures, wheat and soybean futures returns the basic statistics, and describes their basic statistical characteristics of yield,found that simple yield volatility significantly greater than that of logarithmic yield volatility. And from the basic statistics, simple rate of return of the maximum and the minimum, standard deviation, skewness and kurtosis were greater than logarithm yield of corresponding value, This also suggests that simply returns volatility than the number of returns volatility is larger, at the same time also shows their yields are not obey normal distribution, is very typical rush characteristics. At the same time, consider to issue more yields, logarithmic yield is better than simple income calculation easy. In addition, the logarithmic sequence yields and prices of the logarithmic difference, therefore, is to a certain extent, can eliminate price series nonstationarity and relevance. So, in the study of the futures market volatility of the characteristics, we usually use the logarithmic yield is, so this paper chose the logarithm yield as research data.Stationarity test part, we use ADF test method, through EViews7.0econometric analysis software to calculate, and concluded that copper and aluminum futures, wheat and soybean futures returns series is stationary series. From copper and aluminum futures, wheat and soybean futures yield sequence of basic statistical characteristics, partial degrees are not zero, kurtosis is greater than3, so copper and aluminum futures, wheat and soybean futures yield sequence has obvious peak and fat tail characteristics, suggesting that their yields sequence distribution may not obey normal distribution,normal distribution test also shows that the copper and aluminum futures, wheat and soybean futures yield sequence do not obey normal distribution. Through to the copper and aluminum futures and wheat and soybean futures to independence and correlation test, we found that their yields sequence is not independent, and there is a certain correlation. So this article use ARM A model to modify their yield sequence, eliminates the sequence correlation, and residual error sequence of each sequence. Heteroscedasticity testing section, list four different variance test method, and use the ARCH inspection method for copper and aluminum futures, wheat and soybean futures to test, test results show that they are there existed a certain degree of the sequence of yield of the ARCH effect.Study on the leverage effect of Chinese futures market, leverage effect in China. This section first to the volatility of leverage effect is defined and measured. This part firstly defines and measure the volatility of leverage effect. Many scholars found that yield and volatility is a negative correlation, negative earnings would make the conditional variances upward revisions, and positive earnings always causes a conditional variances downward revisions, this is known as asymmetric, also known as leverage effect. In order to measure the lever effect, define "bad" for unintended negative impact, the "good" is defined as a positive impact. Only need to pay attention to the fact, don’t need to distinguish between cause fluctuations in the price of sources of information, are the benefits of this definition. GARCH model is to analyze the expected return, so, we first established a copper and aluminum futures, wheat and soybean futures earnings model, then obtain the expected return from their revenue model. Because their yields are autocorrelation sequence, so we introduce the ARMA model of the sequence of linear filtering, obtain the residual yield, do so mainly for the purpose of removing yields predictable part of the sequence. In the study of this paper is using residual error to estimate. Then introduces the definition of leverage and leverage point measurement, combined with the basic features of futures market, volatility, leverage effect research model is established, and in normal distribution, GED distribution student-t distribution assumptions are studied. Through the study found that the copper futures on the distribution of the three conditions, EGARCH model and GJR-GARCH model estimation results show that it does not have leverage; Aluminum futures in the distribution of the three conditions, EGARCH model and GJR-GARCH model estimation results show that it has significant leverage effects; Wheat futures under the condition of normal distribution assumption, EGARCH model estimation results show that it has significant leverage, but in other distribution assumptions, EGARCH model estimation results show that it does not have leverage effect, under the condition of three kinds of distribution hypothesis, GJR-GARCH model estimation results show that it doesn’t have leverage; Soybean futures in the distribution of the three conditions, EGARCH model and GJR-GARCH model estimation results show that it does not have leverage effect. |
PDF Full Text Request