Analysis of Mutual Fund Performance

In her PhD thesis <link erim events _blank>The Analysis of Mutual Fund Performance: Evidence from U.S. Equity Mutual Funds, Diana Budiono finds that the performance of mutual funds persists. From a business perspective, it is relieving for funds investors that the existence of persistence comes from both the good-performing funds as well as poor-performing funds and that the strategy of selecting a fraction of past good-performing funds still returns profitable outcome during ex-post periods.

Budiono also shows that it is possible to select superior funds by using more than only past performance characteristics to select mutual funds. Importantly, this strategy also requires less turnover (lower cost), which is more appealing from the economic point of view. Moreover, she finds that style timing skills exist, and it is possible for funds investors to identify funds that have these skills, select them in the portfolio, and have higher ex-post performance of the portfolio.

In general, this dissertation focuses on the analyses of mutual fund performance. It studies primarily the determination of the performance, the timing skills of mutual funds, and how to predict mutual funds that will have the best performance by additionally using funds characteristics.

Diana Budiono has defended her thesis on February 4. Her promoters are <link people marno-verbeek _blank>Prof.dr. M.J.C.M. Verbeek, Professor of Finance, RSM and Dr.ir. M. Martens. Other members of the doctoral committee are Prof.dr. W.F.C. Verschoor, Prof.dr. J. Spronk and Prof.dr. J.R. ter Horst.

About Diana Budiono

Diana Patricia Budiono was born on April 27th 1981 in Surabaya, Indonesia. She studied her Bachelor Degree in Industrial Engineering at Petra Christian University Indonesia, and graduated cum laude in 2003. Soon after the graduation, she continued a Master Degree in Financial Engineering at the University of Twente, The Netherlands. In the final year of her Master studies, she joined ING Bank in Amsterdam as an intern. After receiving the MSc, she directly started her PhD in 2005 at the Department of Finance in Erasmus University, The Netherlands. She has presented her research at international conferences, such as in Imperial College London, EURO Working Group on Financial Modeling, and the Southern Finance Association. The article at the basis of Chapter 3 in her dissertation has been accepted for publication in the Journal of Financial Research. Next to her research, she supervises Master theses at Erasmus University. Her research interest includes portfolio management, performance and risk analysis, asset management, mutual funds, and asset pricing.

Abstract of the Dissertation

We study the mutual fund performance for about 45 years. There are several key points that can be learned from this dissertation. First, to study the persistence of mutual fund performance, it is important to consider time-varying exposures because when they are ignored, the persistence will be overestimated or underestimated. Second, the popular investment strategy in literature is to use only past performance to select mutual funds. We find that an investor can select superior funds by additionally using fund characteristics (fund turnover ratio and ability). Importantly, this strategy also requires less turnover, which is more appealing from the economic point of view. Third, the average alpha of mutual funds is an indication of whether it pays off to invest in actively managed funds. We show that a substantial part of the variation in the average alpha can be explained by the average expense ratio, the ratio between skilled and unskilled funds, and combining the average turnover ratio with the skill ratio and trading costs. The latter demonstrates that average turnover hurts the average funds performance due to there not being enough skilled funds. Fourth, selecting mutual funds on only alpha or a single style timing skill leads to overestimating the loading on the selected characteristic and a negative bias towards other characteristics. By estimating for each fund simultaneously alpha and style timing skills over its complete ex-ante available history based on daily returns we achieve two important results, namely the estimated alphas and style timing loadings of the top decile are estimated more accurately; and the ex-post performance of the top decile is superior to that of deciles selected on a subset of characteristics, using monthly data or a shorter estimation window.

More Information

Picture of the Defense
Full Text of the Dissertation