An Empirical Analysis On Change Points Of The RMB Exchange Rate Sequence Based On Bayes Methods

Exchange rate is one important tool for a country to adjust its economic balance for inside and outside,its changes will have a direct impact on the country’s economic growth. The decision depends on the exchange rate movements in the exchange rate of cognitive. To describe the characteristics of the exchange rate level and mastery of dynamics will contribute to build predictive models of exchange rate determination. And exchange rate fluctuation characteristics of time series decides interaction relationship between exchange rates and macroeconomic variables of the rest. With the development of China’s exchange management system reform, the volatility of the exchange market is increasingly apparent. The impact of international events on China’s exchange market is increasingly apparent, and increase the risk of foreign exchange market. In order to guard against exchange risks and to respond flexibly to the impact of international events, law of the RMB exchange rate volatility has become essential in today’s research. According to the RMB exchange rate fluctuations smaller sequence from stage to stage is characterized by large fluctuations. In this paper, the time series of RMB exchange rate gain sequence diagnostic studies, trying to find out the reasons for the number and location of the point mutation, and analyze the reason for the RMB exchange rate fluctuation trend sequence.The main idea of this paper is to introduce Bayes inference, by constructing the likelihood function, while taking advantage of prior information and sample information, and using Gibbs sampling to determine the mutation point. According diagnosed with a date variable number of points to point the date by finding similar events, and to analyze the reasons for the change point appearance. Firstly, assumes that the distribution of the random variable parameters to be studied family; secondly, based on past experience and other relevant information to determine its prior distribution; thirdly, uses WinBUGS to achieve approximate Gibbs sampling, get the edge density test parameters to be studied again; finally, gets back in the posterior distribution,carried out the study variables of statistical inference.Based on the above principle analysis, this paper analyze a set of time series of RMB exchange rate change point from January 1994 to February 2014 monthly. Firstly, determines the number of change points; secondly, divides the RMB exchange rate into different periods based upon detection of point mutations; thirdly, adds dummy variables and analyze characteristics of the RMB exchange rate.The results of this study showed that RMB exchange rate series of mutations in the presence of two points, and the distribution is located in August 2005 and July 2008. After the reform, China’s exchange market is affected greatly increased international events, the impact on China’s foreign exchange market events that can not be ignored. Therefore, investors can not only take interests in the country’s reform and changes in the economic environment, should be more concerned about the impact of foreign economic events. When the exchange rate of the RMB exchange rate reform regulation has a significant impact during the foreign exchange management system, decision-makers should be more careful and should avoid fluctuations in foreign exchange markets.