The Effectiveness Test on the Predictive Power of Taiwan Stock and Futures Technical Indicator

• Main Authors: Ling-Ling Yeh, 葉玲鈴
• Other Authors: Chung-Jen Yang
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/7cf2uh

碩士 === 銘傳大學 === 財務金融學系碩士在職專班 === 95 === Both in theory and in field tests, the stock index derivatives generally demonstrate faster response than stock index itself. As such, this study started with scrutiny of the interaction between futures and stocks, and then it went further to test the predictive power of the stock market trend through combining the futures and stock technical indicators. First this study makes use of the measures of Unit Root, Johansen’s Co-integration method, Granger’s Causality Test and Impulse Response were applied to sort out their relationship. Then the futures and stocks technical indicators were taken, through a process of mock transactions to study the returns on futures and stock markets with KD, RSI, MACD and DIF indicators. Which meant the study focused on whether the price fluctuations on stock index derivatives could send out more messages than the stock index itself. The study sourced daily market-close data from July 21, 1998 to January 31, 2007 and its results are summarized as follows: 1. Through Johansen’s Co-integration method we find that the futures market has achieved a stable long-term balanced relationship with the stock market, showing that two major economical functions of the futures market – (1) the price discovery function and (2) the hedging function, have been well established; 2. Granger’s Causality Test shows that there is no conclusive leading indicator relationship existed between the futures market and the stock market; 3. VECM Model demonstrated a conclusion that adjustments are often conducted in the futures market to rectify long existing imbalance between the futures market and the stock market. Meanwhile, the adjustments taking place in the stock market is often less meaningful. 4. Mock transactions: (1) Under transactions using standard parameters, futures KD indicators gives out a better predictive power, in spite of a lower return compared to buy and hold strategy while no such predictive power can be found in futures RSI and DIF indicators and these two indicators also lead to lower returns than buy and hold strategy; (2) Under the optimal parameter mock transaction environment, the KD and RSI indicators of the stock market are superior to those of stock market in terms of predictive power. Also the KD indicator sends out better predictive results than KD indicator of stock market. On top pf that, futures KD indicator also renders a higher return than the buy and hold strategy, a proof of price discovery ability. Yet the DIF indicator comes out with no clear market prediction ability under the optimal parameter mock transaction environment and it also results in a lower return than the buy and hold strategy.