Replicating top-performing mutual fund returns using ETFs : more bang for the buck.

• Main Authors: Kenneth, Sin., Lim, Wan Peng., Wong, Hong Jie.
• Other Authors: Charlie Charoenwong
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Business::Finance::Asset allocation
• Online Access: http://hdl.handle.net/10356/51391
• Institution: Nanyang Technological University

This study extends the existing literature on analysing the performance of open-end mutual funds through style analysis. Unlike previous studies which focus on market indexes, we leverage on investable market ETFs to replicate mutual fund returns and assess the skills of fund managers. Our replication methodology is based on the asset-class factor model developed by Sharpe (1992). By using a basket of 12 ETFs, from 2006 to 2012, we are able to replicate returns of mutual funds with different strategies. Our results show that the model performed reasonably well for mutual funds of certain investment strategies with some replicated portfolios actually outperforming actual mutual funds net of fees and expenses. Further, we investigate the robustness of the model with varying look-back period and rebalancing frequency. We find that a shorter look-back period with less frequent rebalancing does not compromise the effectiveness of the replication, but may in fact enhance it. In-sample R2 values increase from 0.82 to 0.88 when the look-back period is shortened from 3-years to 1-year. This augurs well for retail investors as it is possible for them to earn mutual fund returns without actual investment experience or expertise.