| Summary: | 碩士 === 國立臺中科技大學 === 財務金融研究所碩士班 === 105 === This study refers to 2,325 pieces of data from January 1, 2008 until May 31, 2017, taking samples of 26 China concept stocks, selecting and using overbought/oversold of three institutional investors, proportion of day trading, balance of margin loan and stock loan as the sentiment indicators through a GARCH (1, 1) model to empirically demonstrate sentiment indicators on China concept stocks returns and risks. First and foremost, the empirical results of all stock returns show that the overbought/oversold of three institutional investors under the long-term condition have had a significant positive impact on returns of individual stocks, which indicates the changes of overbought/oversold sentiment indicators render the returns of individual stocks a positive change. Next, the long-term proportion of day trading has a positive impact on returns of individual stocks as well, which demonstrates the day trading increases the returns. In addition, the financing balance in the long term has a causal relationship with the stock prices; however, there is an uncertainty over the future direction of share price returns. Finally, the long-term increased balance of stock loan leads to a negative trend of future share prices. The empirical results of all stock risks are shown as follows: first of all, the higher (lower) the fluctuation of overbought and oversold of three institutional investors and sentiment indicators of day trading is, the greater (smaller) the risk change in individual stocks will be, of which the overbought/oversold of dealers is most obvious. Besides, the relationship between the balance of margin/stock loan and the risk of individual stocks is not fixed under the long-term influence. Moreover, the overbought/oversold of foreign investment, overbought of dealers and day trading proportion in the traditional industries (Taiwan Cement, Uni-President Enterprises Corporation and Nan Ya Plastics) have shown a significant positive impact. Finally, the empirical results were also presented that the variance was relatively sensitive to the proportion of day trading i.e. the higher the proportion of day trading is, the higher the risk of individual stocks will be.
|
|---|
