|Author's Email Address
||This thesis had been viewed 5332 times. Download 0 times.|
|Type of Document
||Enhanced Index Fund Performance Analysis using the Classification and Regression Tree Factor Timing Model|
|Date of Defense
Enhanced Index Fund
||There is much literature in the field of investment using the Classification and Regression Trees (CART) model to select stocks and construct a portfolio built on a foreign market. The purpose of this study is to examine whether the factor timing model is applicable to construct enhanced index funds in the Taiwan market.|
In this study, we use the method proposed by Miller et al. (2015), which uses the information coefficient as an indicator for judging the effectiveness of factors. Using the CART to establish a factor timing model to forecast effective factors, the effective factors are formed into a composite factor using the z-score of effective factors. This research uses the composite factor to select outperformers and underperformers in the stock pool to construct an enhanced index fund.
The empirical results show that the use of this sample to compare the factor timing model indicates that the performance of a long-short strategy with factor timing is better than without factor timing in the Taiwan market, and when an enhanced index fund has a good performance, the information ratio of the enhanced index fund is 0.95.
||Chia-Fen Tsai - chair|
Yi-Hsi Lee - co-chair
Yih Jeng - advisor
Indicate in-campus at 5 year and off-campus access at 5 year.|
|Date of Submission