RMB Exchange Rate Forecasting Based On Taylor’s Rule

In this paper, we study the relationship between exchange rate and macro economy. By building Taylor’s Rule formula in open economy for China and America, and link them using the Uncovered Interest Rate Parity, we obtain the relationship between the change of exchange rate with macrocosmic data (production gap, inflation rate, real exchange rate and nominal interest rate). We choose post-reform monthly data between 2005.08 and 2013.12, and use last 3 years as forecast period. We use both fixed regression and rolling window regression to conduct prediction. In addition, we use the UIRP, Money Model and PPP model to compare the prediction results via the Mean Squared Prediction Error, TU, CW and DMW statistics. The results indicate that the ability to predict short-term fluctuation of Taylor’s Rule model is much better than UIRP, PPP and money model, with the minimum MSPE of all the models. The reason may be that Taylor’s Rule model take into consideration of the factors of monetary policy decisions and make up for the shortcomings of the traditional models.