Measuring the performance of Taiwan equity mutual funds
Date Issued
2007
Date
2007
Author(s)
Chen, Chia-Fan
DOI
zh-TW
Abstract
In the present study a comprehensive assessment of equity mutual funds’ performance in Taiwan is presented. Using the sample of 163 funds for the period of 1997-2006, we investigate the stock-picking and market-timing ability of mutual fund managers and examine whether persistence in mutual fund performance exists in the Taiwan market. Employing the data from Taiwan Economic Journal, we subdivide all domestic equity funds into six subgroups: high-tech; general equity; medium and small cap; value; special situation; and China conceptual funds. We first measure fund managers’ stock-picking ability by utilizing one-factor CAPM and Carhart four-factor models, and subsequently Chang and Lewellen model is used to determine whether fund managers deliver good market timing. We also construct conditional four-factor model by introducing time-variation in betas, allowing managers to trade on publicly available information and employ dynamic strategies.
One-factor CAPM, Carhart four-factor and conditional four-factor models give the same result that in general fund managers do not possess good stock-picking ability. Looking into each subgroup, Carhart four-factor model and conditional four-factor model obtain the similar conclusion that managers of all subgroups do not have stock-picking ability and under-perform the market index. Besides, Wald test provides significant evidence for time-varying betas in conditional four-factor models. Consistent with prior findings from domestic mutual funds, there is no evidence of timing ability by the fund managers. Finally, we do not find the existence of persistence in mutual fund performance over either half-year or one-year horizons.
Subjects
共同基金
條件迴歸模型
選股能力
擇時能力
持續性
Mutual funds
Conditional model
Selectivity
Timing ability
Persistence
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-96-R94724079-1.pdf
Size
23.31 KB
Format
Adobe PDF
Checksum
(MD5):0d4350210b87dda3d2167c13b52e7ab7