확률할인요소 및 가상펀드를 이용한 주식형 펀드의 성과 평가

In this paper, we evaluate performances of stock mutual funds of Korea using the stochastic discount factor model. We use the standard CAPM and the three-factor model of Fama and French(1993) as parametric models and the no arbitrage model of Hansen and Jagannathan(1991)as nonparametric model. To estimate the stochastic discount factor, industry portfolios and risk free asset are chosen as reference assets. We find that while about one third of the funds analyzed show positive excess returns, most of their magnitudes are not statistically significant. Thus domestic stock funds did not outperform the passive strategy using reference assets. A regression analysis shows that the relative performance is better as the size and the management fee of the fund is greater. To evaluate stock picking and market timing abilities of fund managers we construct artificial stock picking and market timing funds and use their realized returns as benchmarks. Their risk-adjusted excess returns under the stochastic discount factor model shows a positive relation with the information signal, which represents the investment ability of the artificial fund. This evidence implies that the stochastic discount factor model is appropriate for evaluation of fund performances. Finally, we find that most of the funds analyzed performed as well as the artificial fund whose information signal is zero. This result is consistent with the earlier result that excess returns on most of funds are statistically insignificant.