Forecasting and testing spot interest rate volatility

• Main Authors: Chia-Yee Chen, 陳佳宜
• Other Authors: 欉清全
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/fby5ba

碩士 === 國立暨南國際大學 === 經濟學系 === 92 === This article presents several alternative models of the short-term interest rate volatility that must capture well-known empirical features, including mean-reverting and level effect, and therefore uses Chan, Karolyi, Longstaff, and Sanders（1992）’s level model as our baseline model. Furthermore, we will extend previous level model in three directions. First, we consider a version of short-term interest rate volatility being a function of the news arrival process, and use the GARCH model to capture this feature. Second, We introduce the possibility of an asymmetric volatility effect as modeled in Glosten, Jagannathan and Runkle (1993). Finally, we consider a new class of different economic regimes which are likely to appear to govern interest rate volatility, and use the Markov-Switching to account for this explicitly. The other objective of this article is to compare the out-of-sample forecasting performance of models of the spot interest rate volatility using Taiwan, America and Hong Kong short-term interest rates, 1986-2002. For a one-week horizon, the volatility forecasts evaluation shows that the best in-sample model does not necessarily have the highest forecasting power.