对外经济贸易大学《金融时间序列模型》文献阅读5Forecasting Three-Month Treasury Bills Using ARIMA and GARCH Models.pdfVIP

Econ 930 (Term Paper) Forecasting Three-Month Treasury Bills Using ARIMA and GARCH Models Hee-Seok Park Department of Economics Kansas State University 1999 11.18 ABSTRACT The purpose of this paper compares the three different types of forecasting models that are the autoregressive integrated moving average (ARIMA), univariate generalized autoregressive conditional heteroskedasticity (U-GARCH), and multivariate generalized autoregressive conditional heteroskedasticity (M-GARCH) models. According to the results of the ex-post forecast, the forecasted values of ARIMA and U-GARCH models do not follow the real data’s pattern of the three-month treasury bills (TB3). However, the M-GARCH model traces the real data’s trend of TB3 Key Words: Treasury Bills, ARIMA, GARCH, Volatility, and Forecasting All errors in this paper are solely mine. If you have any questions about this paper, please feel free to contact me at hsp3704@. I. Introduction The purpose of this paper compares the three different types of forecasting models. One is the autoregressive integrated moving average (ARIMA) models, which are traditional time series forecast techniques. The others are the univariate generalized autoregressive conditional heteroskedasticity (U-GARCH) and multivariate GARCH (M- GARCH) models, which are widely used in the analysis of financial time series data. It is not an easy job to forecast economic data. Especially, the data such as stock prices, interest rates and foreign exchange rates usually exhibit high volatility. This paper focuses on the multivariate GARCH model than the ARIMA model since multivariate models reflect the real trend of the three-month treasury bills