An Application of Smooth Support Vector Regression on Investment Performance Evaluation of China QDII Fund

• Main Authors: Heng-Hsu Chang, 張恆勖
• Other Authors: Shang-Wu Yu
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/85782114660805636503

碩士 === 國立臺灣科技大學 === 資訊管理系 === 99 === This research is concerned with the SSVR model applied to China’s QDII fund in 2010. We build up the investment strategies that the performance will be better than Shanghai Stock Index and SSE FUND INDEX. Finally, the investors could have some suggestions and make decisions from the result of this research. The empirical results are as follow: 1.The SSVR model predicts the Net Value of QDII fund. It can find that the averages of both MAPE and RMSE have reached good forecasting ability, and the Trend Accuracy is up to 80%. For the China’s QDII fund, issued less than four years, the SSVR model can be used to forecast Net Value. 2.We find that the performance of traditional investment strategies (Contrarian Strategy, Dollar Cost Averaging, Buy and Hold, Momentum Strategy) of China’s QDII fund in 2010 is as follows. The best one is “Contrarian Strategy” followed by, “Dollar Cost Averaging”, “Buy and Hold”, and “Momentum Strategy”. 3.Use SSVR model in China’s QDII fund to compare with the traditional investment strategies, Shanghai Stock Index and SSE FUND INDEX. We find that the performance of the trade strategy (SSVR-2) is better than traditional investment strategies, Shanghai Stock Index and SSE FUND INDEX.