| Summary: | 碩士 === 國立臺北商業技術學院 === 財務金融研究所 === 97 === The purpose of this study is to investigate the relationship between investor sentiment, stock market returns and volatility in the Taiwan stock market. The sample period covers from January 1, 2005 to December 31, 2008. We select six proxies for investment sentiment indexes, including trading index(TRIN), turnover ratio, margin trading to short selling ratio, buy/sell ratio of institutional investors, put/call open interest ratio, put/call valume ratio. Using VAR-GARCH model and causality test, we fnd that stock returns have a bi-directional causality effect on the investment sentiment indexes, such as turnover ratio, buy/sell ratio of instituional investors and put/call open interest ratio. In addition, there is a unidirectional causality from the stock returns to margin trading ratio and put/call volume ratio. However, the trading index(TRIN) and stock returns do not Granger cause each other. Therefore, the results appear that stock returns seem to be the causes of investment sentiment indexes and investment sentiment indexes seem to be the cause of stock return volatility.
This study further discusses the effect of subprime mortgage mortgage crisis in 2007 on the relationship between investor sentiment, stock market returns and volatility. In order to investigate whether the relationship between investor sentiments, stock market returns and volatility change in response to the subprime mortgage crisis. We divide the whole sample period into two sub-samples: before-subprime crisis and after-subprime crisis. In this case, the results find that the Granger causality effects between investor sentiments, stock market returns and volatility becomes more significant. Therefore, it shows that the function of price discovery of the investment sentiment indexes increases following the subprime crisis.
Besides, this study extends the method of Wang, Keswani and Taylor(2006) to construct the predict model for testing the variation in investment sentiment index to the stock return volatility. We regarded subprime mortgage crisis in 2007 as the benchmark for the period of investigation. To examine whether the asymmetric information has impacts on the forecasting ability of variation in investment sentiment index to the stock return volatility or not, this study incorporate asymmetric information effect to observe the forecasting ability of variation in investment sentiment index to the stock return volatility. The empirical results find that the present volatility should be affected by the past volatility significantly. The explanatory power of investment sentiment index to the stock return volatility is better after-subprime crisis than before-subprime crisis. In addition, the asymmetric information increases the impact of variation in investment sentiment index to the stock return volatility and the explanatory power of the predict model. Morecover, investment sentiment indexes and its variation both have excess predict power to the stock return volatility. Finally, this study proposed the empirical value based on the above research.
|
|---|
