The 30-Day CHIBOR Forecast by Combining Regression and ARIMA Model
Selecting 30-day China inter-bank offered rates (CHIBOR) as variable being explained,and monthly industry added value index,the growth rate of money & quasi money supply,and the growth rate of corporate goods price index as explanatory variables,a 30-day CHIBOR forecast model only based on the regression method is first established.Next,by combining the regression model and autoregressive integrated moving average (ARIMA) model,second model forecasting 30-day CHIBOR rates is founded.Then,an out-of-sample forecast to 30-day CHIBOR rates are done by using the two models and a comparison to the forecast performance about two models is conducted.The results show that the second model accurately captures the change of CHIBOR rates in 30-day maturity while the first does not and the forecast errors of the second model are smaller than those of the first,which suggests that the choice to three explanatory variables are rational and a model based on the regression and ARIMA method is better than the other one only based on the regression method in forecasting 30-day CHIBOR.
CHIBOR Forecast Regression ARIMA
Linchang Pu Daoping Chen
School of Economics and Management Chongqing Normal University Chongqing,China
国际会议
青岛
英文
100-103
2012-07-20（万方平台首次上网日期，不代表论文的发表时间）