Comparison and Prediction of Nonlinearity of Exchange Rate Fluctuation Based on ANN Method
Comparison and Prediction of Nonlinearity of Exchange Rate Fluctuation Based on ANN Method
Author：Pang Xiaobo,Sun Yemeng,Wang Chen Journal：Jilin University Journal Social Science Date：2008(1)
Keywords Exchange rate; Nonlinear test; STAR model; ANN model
Abstract
In view of the fluctuation of the RMB / USD exchange rate since the reform of the exchange rate system, we use the EGARCH model, the STAR model and the ANN model to analyze the regular performance of the volatility. Based on the artificial neural network model, the nonlinear test of exchange rate fluctuation is realized. Compared with the ANN model, the fitting effect of EGARCH model is better than that of ANR model. By examining the fact that the exchange rate fluctuation is non-linear, China's exchange rate fluctuation already has a signal function that reflects the economic changes at home and abroad.
Attachment
