Empirical Study Of Chinese Stock Market Based On Spectral Improved Measures

April 16,2010 China's CSI 300 stock index futures contracts has been officially launched.Today its turnover and turnover growth has a double. Can be said that the official launch of stock index futures stock market in China has changed the basicsituation of risk management, and improve China's stock market hedgingmechanism, so that financial institutions and the investors can use derivatives to prevent market risk and reduce unnecessary losses, but also Makes more complexthe composition of the portfolio, risk management needs improvement, riskmanagement needs to be improved. In addition, the face of the deepeninginternational financial markets, China and Western countries, more and more closelylinked financial markets, the impact of risk relative to the local community, the impact of foreign capital market also can not be underestimated, the use of effective riskmanagement tools Integrate multiple sources of risk factors and its effective detectionand prevention is particularly important.As more and more financial institutions to recognize the importance of risk management, resulting in a risk measurement methods are numerous, including VaR (Value at Risk) method because the risks can best express visually the size and the probability of risk occurring Two aspects, which most financial institutions around the world recognized and widely used in market risk, credit risk, operational risk, financialrisk areas. However, despite the VaR method has its natural advantages and operational, the method still has some serious flaws, which will lead to the riskmeasure in the actual process of the emergence of large deviation, which also gave birth to the VaR method Framework for the improved model, the improved methodare optimized to varying degrees, VaR risk management approach also extends the idea.So the stock market in China at this stage the actual circumstances, the performance of these empirical methods on how to improve, there are problems; what ways it can meet the needs of effective management at this stage, a better measure of financial market risk? On these issues, research of the actual level of risk stock index also has important theoretical and practical significance.In the empirical process, through China's Shanghai Composite index the actualvolatility of the study, found that China's stock market is more significant marketvolatility period and the market is relatively stable during the interval division, theinterval has a statistically significant impact on the actual level of risk measure; In addition, based on the actual experience of foreign and domestic realities, in order toavoid the assumed distribution of income caused by the estimation error sequenceand can get an estimate of the convenience methods, the paper that the historical simulation method can most effectively meet the needs of our risk managementpractice. This article will form a continuous spectrum of risk estimation models intodiscrete functions to calculate the form of historical simulation method to estimate the value, in order to get a wider range of applications and better empirical results.Under the above two considerations, the paper will be data on the Shanghai Composite Index divided by volatility intensity of two intervals, comparing the characteristics of the two interval data, and will not consider the range of estimates bythe improved method of calculation results with the results of the phase Comparison,trying to improve the method can prove the measure of the risk profile of the stockmarket reality.Follow the above analysis, the empirical estimates, we can get the following conclusions:1. China's stock market volatility there is obvious variation, leading to the traditional method to estimate real risk the stock market in China there is a serious shortage of state;2. Historical simulation method can better measure the potential risks of stock market conditions, and simple and easy to popularize, but the method is still slightly overestimated the actual level of risk;3. Using ES method for China's stock market risk measure, we found that the method is somewhat conservative estimates, there is a tendency to overestimate the risk of stock market risk, risk aversion for cautious investors, this method may be a goodchoice; 4. Use of spectral risk measure model to measure stock market risk in China, we found that the empirical results of this method are between the historical simulation method and between the ES methods, and the method compared with ES and historical simulation method can be more flexible empowerment for the extreme values, while meeting the requirements of coherent risk measure, this paper that the method can measure the actual risk level of the stock market.5. When there is greater volatility in the stock market in China or extreme losses, the spectral risk measure using a modified model of the spectrum than ordinary in terms of risk estimation models to better estimate the actual risk level of the stock market.This paper is a lot of reading on the basis of relevant literature, thinking through their repeated arguments and empirical exploration written. This innovation is mainly reflected in the following points:1. Acerbi (2002) and later most of the documents referred to and the use of spectral model continuous risk estimation models for the expression, but this attempt to use the model of discrete measure expression of China's Shanghai Composite Index on the risk profile, and discussed in detail idea of this process, but also through empirical analysis of the data confirmed the effect of this improvement is quite good;2. Evidence in the domestic risk management methods for the spectral risk measure less study, and its only method of estimating parameters, while the article uses the historical simulation method for non-parametric methods of the empirical research carried out;3. In the spectrum of risk measures by using the method of calculation of market risk, because China's stock market index data, empirical results are unsatisfactory, and would result in the deviation value at risk measure, based on this method In this paper, HW volatility sample data weighted correction processing, and through the three methods of empirical risk measure comparing the spectrum proved that the improved method can better reflect the risk of stock market risk characteristics of the moment.