Summary: | 碩士 === 世新大學 === 財務金融學研究所(含碩專班) === 104 === This study investigates whether utilizing data on net open interest of foreign investors, of dealers, and of investment trust could earn extra positive rate of return on Taiwan Stock Index Futures and MSCI Taiwan Stock Index Futures investment. I use daily price data on Taiwan Stock Index Futures and MSCI Taiwan Stock Index Futures from July 2, 2007 to April 15, 2015. The ordinary least squares (OLS) regression is used, where I try many models with various combinations of the changes of daily/weekly net open interest of foreign investors, dealers, and investment trust. Also, various lag periods are tried. In total, I make and test twenty-six models and calculate their out-of-the-sample rate of returns.
The empirical conclusions are as follows: Twenty investment models can achieve more than 50% forecast correct rate.
In particular, for daily investment, the best model for Taiwan Stock Index Futures is to utilize the changes of daily net open interest of foreign investors, of dealers, and of investment trust. In my sample, this model earns an out-of-the-sample annual rate of return of 19.88%. For MSCI Taiwan Stock Index Futures, the best investment model is to utilize the change of weekly accumulative net open interest of foreign investors and the weekly accumulative return of MSCI Taiwan Stock Index Futures. It achieves an out-of-the-sample annual rate of return of 19.44%.
For weekly investment, the best model for Taiwan Stock Index Futures investment is to utilize the change of weekly accumulative net open interest of investment trust. It achieves an out-of-the-sample annual rate of return of 14.09%. For MSCI Taiwan Stock Index Futures, the best model utilizes the change of weekly accumulative net open interest of dealers, and achieves an out-of-the-sample annual rate of return of 12.93%.
|