Economic Variables and the Predictability of Stock Volatility in Taiwan

• Main Author: 吳湘韻
• Other Authors: 趙世偉
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/75317132076612030031

碩士 === 國立政治大學 === 金融研究所 === 103 === Aggregate stock volatility is changing anytime. The main topic of this thesis tests whether it is possible to improve the stock market volatility forecasts by adding the macroeconomics variables into the linear model. The paper uses the linear predictable model to construct the in-sample and out-of-sample analysis forecast model to test whether the model considering macroeconomics variables outperforms the benchmark model. It is difficult to find the empirical evidence that forecasts on conditioning an macroeconomic variable outperform than the volatility lagged information. Only the financial variables which are Taiwan stock monthly turnover rate and credit variables have improved the model forecast. The most successful approaches involve simple combination of individual forecasts.