Font Size: a A A

A Study On Combination Forecasting Of RMB Exchange Rate With The GARCH-GRNN Model

Posted on:2010-01-07Degree:MasterType:Thesis
Country:ChinaCandidate:L M HuangFull Text:PDF
GTID:2189360275982245Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
With the continuous advance of modern economic and financial globalization, the international capital flows frequently. So the exchange rate plays a more and more important role in the economy both at home and abroad. Recently, under the impact of the financial crisis, the global exchange rates fluctuate intensely, and the corresponding exchange rate risk is also increasing. Since the reform on China's exchange rate system on July 21, 2005, the government has relaxed the controls on the exchange rate step by step, which consequently leads to the unprecedented volatility of the RMB, and finally a more flexible exchange rate mechanism. Therefore, it is vital to forecast the exchange rate accurately so that such micro-entities as enterprises can avoid foreign exchange risks, the central bank can tighten financial supervision, as well as the government can make effective exchange rate policy.Firstly, the thesis thoroughly reviews the development of exchange rate forecasting theory and technology, and specially elaborates the application of generalized autoregressive conditional heteroskedasticity model and artificial neural network in exchange rate forecasting. Secondly, on the basis of summarizing and evaluating the latest combination forecasting methods of exchange rate, the thesis analyzes the strengths and weaknesses of linear and nonlinear combination models. At the same time, the thesis explores the development trend of the combination forecasting models of exchange rate. It is found that the direction of the further research is to propose nonlinear models combined with artificial intelligence technologies such as artificial neural networks, genetic algorithms, wavelet analysis, support vector machines and so on. At last, the thesis proposes a combination model, with the help of the GARCH model and GRNN technology, to predict RMB/USD and RMB/EUR, from which we can make a conclusion that the combination model does improve the forecast accuracy of the exchange rate and that the GARCH-GRNN combination models works better than the GARCH-BPNN combination models.
Keywords/Search Tags:RMB Exchange Rate, Exchange Rate Forecasting, Combination models, GARCH model, Generalized Regression Neural Network
PDF Full Text Request
Related items