Dynamic Analysis Of RMB Exchange Rate Based On Markovion Regime Switching Model

In recent years, the RMB exchange rate fluctuates frequently and changescomplicated. Especially the U.S. subprime mortgage crisis and the European debt crisis,make the exchange rate fluctuates drastically. However, exchange rate movementsgenerate great impact on the domestic economy, therefore, the research of exchange ratevolatility has a great significance on ensuring the smoothly development of China’seconomy and accurately grasping the pulse of China’s economic.This paper selects the January1,2007to December2011,31RMB against the U.S.dollar, euro, yen and pound exchange rate data for the study. This article adopts MS-ARmodel and MS-ARCH model&MS-GARCH model which are established on the firstand on the second moment of Markovion Regime switching model respectively, andmakes contrast with the general time series AR model, ARCH and GARCH models. Atthe same time, supposing the MS-AR model is the normal distributed or the t distributed.Study the dynamic characteristics of the RMB against the U.S. dollar, euro, yen andpound exchange rate. The empirical results show that:(1) Time series models with statetransition have a better description of the fitting and volatility effect than the general timeseries models. The MS-AR model is more flexible and accurate than the AR model; Thenormal ARCH model omits the revealing of low-volatility state of exchange rate, whilethe MS-ARCH model can reveale the fluctuation of exchange rate more comprehensive;The MS-GARCH model can reduce the duration of the national currency exchange ratefluctuation than GARCH model.(2) The fluctuation of exchange rate is in contrast to theduration. The euro exchange rate has a most drastically volatility and a shortest duration,while the U.S. dollar exchange rate has a longest duration and a weakest volatility. Theresults show that, affected by the debt crisis in Europe, the euro exchange has a mostdrastically fluctuaion, sectors and individual investors should deal with the euro positionsefficently avoiding to the risk fluctuation of exchange rate and avoid exchange rate riskeffectivly.(3) The RMB against the U.S. dollar, British pound, Japanese yen exchangerate data have a more obvious characteristics of thick tail than the euro exchange rate data, so the residual distribution of the RMB against the U.S. dollar, British pound, theyen exchange rate data is fit for a t distributed, while the residual distribution of the RMBagainst the euro exchange rate data is adequate for the normal distributed.