Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

5 Mutual Funds Poised To Pop Based on Leading Academic Models

Every mutual fund prospectus cautions that 'past performance is not an indication of future returns,' yet this is precisely the principle investors are forced to rely upon when using today's mutual fund rating systems. Even with the U.S. stock market’s appreciation in recent months, the gains have not eased a fundamental challenge facing mutual fund investors: on the heels of a year when every mutual fund category was clobbered, using mutual fund rating systems based on historical performance such as Morningstar offer little guidance on which mutual funds will earn superior returns in the future.


To date, mutual fund investors could hardly be blamed from relying solely on historical performance models because they were the only game in town. But an emerging approach to rating mutual funds has demonstrated superior returns and, of greater relevance, can for the first time be leveraged by everyday investors and financial advisors.


The approach is unique because it is not predicated on historical fund performance, but on proven, time-tested mutual fund models that are predictive in nature. The academic models are based on leading academic research using state-of-the-art mathematical and statistical techniques to identify unique variable to predict mutual fund performance. The professors who developed the models are from leading business schools and are widely recognized as the top researchers in the field.  The papers describing the models’ results have been presented at top financial conferences and have, or will be, published in the leading financial journals.


Unfortunately, this research did not exist in a form that could be used by investors.  Until now. Earlier this year, a mutual fund rating system launched that grades over 3,600 U.S. equity funds using a proprietary formula that factors results based on four leading academic models:


  • Forecasting Alphas, which generates a better alpha measure so investors can identify mutual funds most likely to deliver positive risk-adjusted returns in the future.


  • Judging Fund Managers, which finds fund managers who use similar techniques to top performing managers and are likely to deliver similar performance in the future.


  • Return Gap Model, which analyzes the difference between a fund’s actual return and the return it would have earned by following a buy and hold strategy.


  • Active Share Model, which measures the extent to which a mutual fund’s portfolio holdings differ from the portfolio’s benchmark index – and if the fund outperforms its benchmark.


The essence of the predictive academic mutual fund models is that they seek to find those less than 10% of the mutual fund managers who can consistently outperform the market. The models evaluate a manager’s performance on an absolute basis and relative to other managers.  The approach varies by model but the premise is the same – there are a few smart managers out there whose funds are better positioned and/or who see or react to events better than others and, thus, consistently beat the market.


The new, time-tested mutual fund rating system based on these four academic models currently identifies the following as the top five funds:


1)    Heartland Select Value Fund (MUTF:HRSVX)

2)    Yacktman Fund (MUTF:YACKX)

3)    Thornburg Value Fund I (MUTF:TVIFX)

4)    Thornburg Value Fund A (MUTF:TVAFX)

5)    Thornburg Value Fund R3 (MUTF:TVRFX)


There are four significant advantages to ranking mutual funds based on academic research relative to methods based on funds’ past performance:


  • The Unique Variable Or Measure Of Each Academic Model – Each academic model made available develops a unique fund measure using data that is not utilized and, in fact, often ignored by traditional mutual fund data providers.
  • The Predictive Nature of Each Model – The models often use portfolio holdings data, employ state of the art statistical techniques, and require extensive calculations that provide the models with much greater predictive power than traditional rating systems.
  • Each Model Tested Over Minimum of 20 Years – The four academic models have been developed through extensive research and thoroughly vetted by the academic community.  Each leading academic model for which funds can be ranked at MUTUALdecision studies specific aspects of fund performance over a minimum of 20 years, covering recessions and occurrences such as the 1987 market crash, 1997 financial crisis and 9/11 and demonstrated returns above the appropriate benchmark.
  • Uses Sharpe Analysis to Determine a Fund’s Investment Style – The fund investment style characteristics are determined by Sharpe style regression techniques, whereby fund returns are regressed against various style index returns, an accepted academic approach. 


Early evidence reinforces the models’ continued predictive abilities and value of this alternative mutual fund rating system.  The following table shows year-to-date returns for the top 25 funds in each of the four models upon which the new mutual fund rating system is based on; and compares the returns to the S&P 500.  The results are current through August 31, 2009:


                                                Actual Return               Over/Under S&P 500 Return
Return Gap                            14.9%                           +1.9%
Active Share                          31.7%                           +18.7%
Judging Fund Managers      20.3%                           +7.3%
Forecasting Alphas              28.8%                           +5.2%

The academic community has identified a better way to find the best performing mutual funds, enabling investors to earn superior returns.  These useable, predictive academic models are no longer confined to academic journals and conferences, but can finally be leveraged by investors and financial advisors to forecast future fund performance and earn superior returns.


Disclosure: No positions in any mutual funds referenced in this article