Font Size: a A A

State Transition-based Time-varying Combination Of The Rmb Exchange Rate Prediction

Posted on:2011-12-17Degree:MasterType:Thesis
Country:ChinaCandidate:W ZhangFull Text:PDF
GTID:2189360332455920Subject:Quantitative Economics
Abstract/Summary:PDF Full Text Request
July 21,2005 the Chinese government to reform the RMB exchange rate regime to give up a single dollar-pegged exchange rate regime pegged to a basket of currencies in a managed floating exchange rate system. The Government relaxed controls on the exchange rate, exchange rate fluctuations become increasingly severe. Therefore, the characteristics of the current RMB exchange rate behavior and manifestations of in-depth study revealed its inherent mechanisms, and to improve the accuracy of forecasts of exchange rate fluctuations, there is conducive to the country thus making the exchange rate on macroeconomic policies, timely and appropriate exchange rate adjustments in order to maintain the stability of national economic development, carrying out economic activities and foreign exchange can be related to relevant measures taken in advance to reduce exchange rate risk.Said the international lending, capital markets, said the fundamentals such as purchasing power parity analysis model was largely based on strict assumptions, and often can predict the long-term level of the exchange rate movements, in the short to medium term level, technical analysis method can provide better forecasting results. Early studies of technical analysis of exchange rate fluctuations, methods include moving average, exponential smoothing and exponential weighted average of such models are generally assumed exchange rate return series to obey normal distribution, exchange rate volatility is defined as the change over time and independent, identically distributed constant. With the further development of financial theory, many scholars study found that the experience of the exchange rate return series of independent normal distribution with significantly different, more showed a clear "spike thick tail" and the skewness of the unconditional distribution of non-zero the sequence variance or volatility of a certain self-correlation, there is a time-varying volatility, aggregation and non-symmetry. While scholars have individual forecasting methods for continuous improvements, better characterization, simulation and prediction of exchange rate fluctuations, but because of the complexity of the exchange rate market, making the individual forecasting models that incorporate and reflect only fluctuations in the local information and can not be a good All frequent fluctuations in exchange rates. The combination forecasting method is based on the largest, based on the use of information, through the combination of several individual models, the assembly of these models contained in the information, so in most cases, through the combined forecasting model for considering issues than a single prediction model more systematic, more comprehensive, can effectively reduce the individual forecasting models subject to certain environmental factors, thereby enhancing prediction accuracy. According to the individual forecasting model combined in different ways, combining forecasting can be divided into variable weight combination forecasting and constant weight combination forecasting. Individual model predictions will change as the economy showed "good times and bad" sex, while the variable weight combination forecasting can be in different economic times, according to the model goodness of fit of individual changes, to achieve individual model weights changes, compared with the same weight combination forecasting more scientific, you can better fit the complex exchange rate system to achieve the purpose of improving the forecast results.In this regard, this paper transfer of state-based time-varying combination of the exchange rate forecasting method, for July 21,2005 to July 31,2009 the central parity of RMB against the U.S. dollar daily exchange rate data, through the time-series model, the same weight combination forecasting and variable weight combination forecasting model for modeling, forecasting that the method is better than individual forecasts and the same weight combined forecasting to improve the forecast accuracy of the exchange rate, exchange rate projections have a certain value.This paper first describes the research status and forecast exchange rates combined forecasting methods at home and abroad, and then described in detail a number of commonly used exchange rate forecasting model. Combined with exchange rate fluctuations, the specific characteristics of individual forecasting methods commonly used to evaluate the prediction model according to the scope of the individual selected two kinds of individual forecasting methods, and then individual forecasting model is applied to the transfer of a combination of state-based forecasting methods, and their China's exchange rate applied to the projections made good fitting and prediction。...
Keywords/Search Tags:Markov-Regime-Switching, the Prediction of the Exchange Rate, Combination Forecasting
PDF Full Text Request
Related items