| The computer science and the Internet technology gains lots of financial products to thecontemporary financial market with continual improvements and innovations. To the investors who involvein the market， it is really important that how they could master the characteristics of the products and howthey use their knowledge to make the right decision about the financial products. This paper works on theCSI300index future. On the one hand, the future market guides the present market in some way and it canreflect the situation of the economy. On the other hand, the future has leverage which can stand for riskbetter. And we need to know more about the future contract, especially the feature of leverage. Hence, inthe term of investors, the most important one of the issues needed to be solved is balancing risks andpayoffs.It requires investors to have the knowledge of the risk theory and the risk model. This papertherefore is based on the basic definition of risks. Then research more about the risk theory and theestablishment of the risk model. In the investment, a risk doesn’t mean a danger equally. Instead it standsfor the uncertainty of an event that may occur in the future. The event could result in a loss or gain a profit.When it is necessary to make a judgment about a risk, it should rely on the eventual result of returns.Having realized of the meaning of risk，it will be useful to master the risk theory. Among a lot ofrisk theories, this paper chooses the VaR theory as the object in the research. Since the VaR theoryestablished in the ninety’s of the last century, it has been used in most financial institutions and written as aprinciple in the Basel Accord after developed for almost twenty years. It is obvious that the VaR theory isnot only a theory used in the academic circles, but also an important financial instrument in the real world.And it reflects its practical value on a method of quantitative analysis which can be compared. The originaldefinition of VaR is the largest loss of a financial product in some level of confidence. Here we can rely onthe method of frequency distribution or the sorting method. Also, we can get the value of VaR through thefeature number from the assumption of some distribution. And the former is the method of non parameticwhile the other one is the method of parametric. Nevertheless, it should be emphasized that the results ofthis quantitative analysis can be used in the particular suitable situation. And any exaggeration of the use ofthe method is not allowed, such as taking it for granted that only relying on one quota to avoid risks. In order to get more features of risks, I choose the yield rate as the object in the research and paymore attention to analyze features of the yield rate when it is violating. The reason why I choose the ARCHmodel as the main instrument to do the research is that this model aims at the volatility clustering. First ofall, ARCH model follows AR model. They both show that the today’s yield rate contains the historicalinformation and explain the phenomenon of the correlation with the historical values. What’s more, ARCHmodel has its own unique assumptions. It is against the classic homoskedasticity assumption. And itsatisfies the heteroskedasticity assumption. The variable variance is different from Brownian motion. Thevariance here is a conditional variance which is based on the historical information of errors. Because ofthe correlation between the variance and the historical errors, the variance could be in connection with thetime. Then, the volatility shows the phenomenon that some yield rates gather around with high volatilities.Later on，GARCH model which is based on ARCH model spreads the application horizon of ARCH model.Furthermore, it solved the complicate calculating issue of the high-order ARCH model. Therefore, if itcomes to the high order, it will make a better result with a less effort using GARCH (1,1).While doing the important empirical analysis, it needs to update the original data of CSI300index futures in order to get the yield rate data. Then, we start with the stationarity test and go forward untilwe set the GARCH (1,1) model ultimately. And take advantage of the model to prove that CSI300indexfutures have the feature of volatility clustering using the results of the stationarity test. Therefore, itcontains the historical information when its yield rate is violating. What’s more, the unusual phenomenonwill last for days instead of dismissing when it changes a lot. On the other hand, use the historicalsimulation method to calculate the VaR in order to fulfill the analysis of the risk of yield rates. Through theanalysis of the risk, we can find that the distribution of the CSI300index future has a long fat tail on the left.This distribution reveals that the probability of the great loss of the future product can be larger than that inthe theory. All in all, when we need to know about the risk of a future contract, we should try to dig andfind more about features of it. |
PDF Full Text Request