An In-Depth Look at the MFS Value Fund

During the most recent 10-year period, MEIAX posted a return 0.2 percentage points lower than DFUVX but did so with lower volatility. This resulted in a higher Sharpe ratio. MEIAX also outperformed the Russell 1000 Value Index on all fronts: higher return, lower volatility and a higher Sharpe ratio.

2010-2014

During the most recent five-year period, MEIAX once again produced lower returns (in this case 3.1 percentage points lower) than DFUVX. Again, its lower level of volatility resulted in a virtually identical Sharpe ratio. This time, however, MEIAX underperformed the Russell 1000 Value Index on all fronts. It produced a lower return, exhibited higher volatility and had a much lower Sharpe ratio.

I’ll now dive deeper into MEIAX’s performance to see if we can identify the sources of the fund’s returns. I’ll use the factor analysis tools provided by Portfolio Visualizer to analyze the fund’s performance.

Factor analysis

I begin by examining the fund’s performance for the full period from 1996 through 2014 using a three-factor (beta, size and value) analysis.

The three-factor analysis shows that the fund was a true large-value fund with statistically significant loadings on the size and value factors. The annual alpha was 1.44%, though it wasn’t statistically significant at the 5% level.

My next step is to examine the fund’s performance through the lens of a four-factor model, adding momentum as the fourth factor.

Here we see similar results. There is little change in the loadings on the three original factors and a small negative loading on momentum. That is no surprise for a value fund. All the loading factors are statistically significant. The annual alpha increased to almost 2% and is closer to being statistically significant at the 5% level.

We’ll now add the quality factor to discover whether its inclusion provides us with additional explanatory power. The research demonstrates high-quality stocks that are profitable, stable, growing and have a high payout ratio tend to outperform low-quality stocks with the opposite characteristics.

Adding the quality factor to our analysis not only improved the explanatory power of the model (the r-squared increased from 0.86 to 0.89), but the fund’s alpha went from nearly 2% to about -1%— although it wasn’t statistically significant in either case.

Again, all the loading factors were statistically significant. What this shows is that the fund’s success was driven not by individual stock selection. Instead, it was driven by investing in stocks that loaded on the quality factor. Of course, this doesn’t take anything away from the fund’s performance because the managers were investing in these stocks before the academics had identified the factor’s existence. But today there are passively managed funds with much lower costs that load on this same factor.

In other words, what was once alpha (for which managers could charge large fees because it’s a scarce resource) has become beta (a commodity which can be purchased at lower costs); stated slightly differently, alpha can be achieved not just through individual stock selection or market timing but also by loading on a factor that has yet to be "discovered" or identified.

This process of academics using research to convert alpha into beta is one of the four main themes my co-author, Andrew Berkin, and I discuss in our new book, The Incredible Shrinking Alpha. This evolution has led to the quest for alpha becoming more and more frustrating for active managers.

The three other themes covered in the book are the shrinking supply of victims that can be exploited, the increasing skill level of the competition and the dramatic growth of the supply of capital chasing the shrinking pool of available alpha.

But let’s get back to our analysis. There’s one more factor, the sixth, I can add: low beta.

The results are virtually identical to what we saw from the five-factor model with the fund’s annual alpha becoming slightly more negative.

Recent performance

I’ll do one final round of analysis to ascertain if it’s possible to gather additional insights. To determine if I can detect any trends, I’ll examine the fund’s performance for the most recent 15-, 10- and 5-year periods using the six-factor model. We’ll begin with the last 15 calendar years.

The analysis confirms again that the fund was a large-value fund with an emphasis on quality stocks. To reiterate, since value and momentum are negatively correlated, it’s not surprising to see a small negative loading on momentum. In addition, the fund had a small positive loading on low beta. That’s also not surprising since quality and low beta are somewhat related. All of the loadings were statistically significant. There was a small positive annual alpha of 0.19%. The r-squared is now higher at 0.93, meaning the model is doing a pretty good job of explaining returns.

We’ll now look at the last 10 calendar years.

While it remained a large-value fund that also loaded on quality (with statistically significant t-stats), the fund’s loadings on value and quality fell quite a bit. And while the annual alpha isn’t statistically significant, it’s now slightly negative. What’s more, the explanatory power of the model rose to 0.97.

My last look at the data is for the most recent five-year period.

The big change we see here is that the fund’s annual alpha was almost -2%, and it was approaching the level of statistical significance. Recall that during this period, MEIAX returned 14.1% per year, 3.1 percentage points a year less than the return of the similar DFA fund (although it also exhibited much lower volatility, resulting in very similar Sharpe ratios). We can see something less than statistical significance in some of the loading factors, though that’s not unexpected as the period is relatively short.

In summary, MEIAX certainly has an impressive track record over its full life. The analysis revealed that its outcomes were a result not of individual stock selection. Instead, they were due to loading heavily on the quality factor (the six-factor analysis produced a negative alpha for the full term). Of course, that doesn’t in any way detract from the performance of the fund because the quality factor has only recently come into the literature.  The fund managers were exploiting this factor before the academics “uncovered” it.

The picture changes, however, when we focus on returns. All of the higher returns relative to the comparable DFA fund occurred during the first four years of the MEIAX’s existence. Since then, its returns have been lower, though so has its volatility. In addition, while the fund’s alpha was slightly positive over the most recent 15-year period, it turned slightly negative over the last 10 years and became sharply negative during the last five years. That’s the period in which its returns were 3.1 percentage points lower than the comparable DFA fund. Although, once again, it did exhibit less volatility and had a similar Sharpe ratio.

Perhaps the fund’s assets under management have grown so large that they have become too big of a burden for its managers to overcome. Or perhaps the markets have become more efficient and the competition increasingly more skilled as the losers in the game of active management exit and convert to passive strategies. A trend of declining performance can be found when examining the performance of actively managed funds. That’s one reason why the SEC warns that past performance is not a predictor of future performance. It also explains the research on the performance of pension plans, which hire leading consulting firms to help them perform due diligence in choosing managers. This research shows that while they hire managers with very strong alphas, the alphas disappear after hiring.

Only time will tell if MEIAX can once again generate superior returns and justify its relatively high expense ratio of 0.88%. However, the evidence, as presented in The Incredible Shrinking Alpha, makes a compelling case that the odds are against that happening.

Larry Swedroe is director of research for the BAM Alliance, a community of more than 150 independent registered investment advisors throughout the country.

Read more articles by Larry Swedroe