The Characteristic Analysis And Modeling Of Exchange Rate Volatility In China’s Foreign Exchange Market

The exchange rate is one of the important variables of the international financial market. To analysis the RMB exchange rate fluctuation character-istics, can not only have an important reference to banks and other financial institutions in terms of the foreign exchange risk management, but also help to build the exchange rate derivatives market including foreign exchange for-ward, foreign exchange option, foreign exchange swaps. Since the deepening of China’s exchange market reform, the market characteristics of exchange rate fluctuations become more obvious than ever. At this point, we should trans-form our focus from the choice of exchange rate regime to the discussion of the RMB exchange rate fluctuations, and so we can analysis of the RMB exchange rate issue in a more scientific way.Financial Time Series widespread fluctuation, since 1982 Engle proposed ARCH model to describe volatility, the establishment of volatility models came out one after the other, which mainly include two categories:GARCH Models and SV Models. But there is no clear distinction on the merits of the above two models in terms of the fitting for our exchange rate of return sequence. Therefore, through the selecting of the return of exchange rate data, we ana-lyzing the volatility characteristics, and establishing appropriate GARCH and SV models, in the last, by selecting a certain criterion, we compare these two models.The central bank made appropriate adjustments of China’s exchange rate management in July 2008 to June 2010, to cope with the international financial crisis. Since during that time the government mainly controls the fluctuations of the RMB against the US dollar, and the intervention will cover up the role of the market. So in order to better reflect the implementation of the second currency market after the reform of the exchange rate, we select the RMB exchange rate data against the euro and USD from July 1,2010-March 4, 2016 and the offshore RMB exchange rate data against USD from April 30, 2012-March 4,2016, to analyze the return of exchange rate fluctuations.Firstly, we tried to inspect the exchange rate volatility form its distribu-tion, volatility clustering, leverage volatility, long memory and the spillover effects. By analyzing we knew, China’s exchange rate of return distribution has obvious fat tail and volatility clustering. And from the timing sequence diagram of exchange rate yield fluctuation we can see, the volatility in the exchange rate market showed an increasing trend, meanwhile as our central bank announced the exchange rate adjustment on August 11,2015, an abnor-mal fluctuation change has taken place in the dollar against the RMB exchange rate, which is the RMB has depreciated for more than 3% in three continuous days. And this further demonstrates market-oriented exchange rate has deep-ened in another way. We checked the leverage effect by establishing EGARCH model and in this paper, we first test the ARCH effect of three exchange rate sequence and find the EURO against the RMB sequence dosen’t have it, so we can’t use GARCH model to fit the data, and in contrast, we choose to use SV model. In view of the US dollar against the RMB and the US dollar against offshore RMB both have ARCH effect, so we use EGARCH model directly. And the result indicate that the former has obvious leverage effect, but the latter dosen’t have leverage effect for it has greater response to the same de-gree of good news rather than bad news. For the long memory test, we use the corrected R/S analysis method. And by comparing the Hurst exponent, we come to the conclusion that all three sequence have long memory. In the same time, this paper draws the loglog graph to get their maximum memory length.Secondly, considering the long memory of exchange rate volatility, we established FIGARCH model and LMSV model to fit the data, and we chose the MAE and the comparison diagram about predicted and actual value as the criteria to evaluate the power of the two models. At last we find the LMSV model has better fitting accuracy. At the end of this article, this paper summarizes the characteristics of the RMB exchange rate market, and then establishes appropriate models, furthermore evaluates the models, and finally we show the shortcomings of this paper with putting forward the prospects of the future.The innovation of this paper is in the following two aspects:firstly is in the selection of data. This paper chose the US dollar against the RMB、the euro against the RMB and US dollar against offshore RMB as the research object. Through the comprehensive research of the former two data we can get the general features of China’s currency market yields. In addition, by the contrast analysis of US dollar against the RMB and the US dollar against the offshore RMB exchange rate data, we can concluded the differences of China’s currency market offshore and onshore, and by that way we can come up with new ideas for the reform of China’s exchange rate market.Secondly, the inno-vation stays in the model selection. This paper chose EG ARCH% FIGARCH and LMSV model as a research tool. By using the EGARCH and FIGARCH model to get comprehensive analysis of the exchange rate fluctuations and long memory. Meanwhile selecting the long memory GARCH model and long memory SV model respectively when modeling the data fitting. By comparing the fitting precision, we can make a comparison of the two kinds of model, thus providing assistance for future research.