Swedroe: Small Value Vs Midcap Value

Sometimes your choice of index is as important as which value-tilts you favor.

My column from July 14 on the persistence of the small-value premium resulted in some interesting discussions on the subject. I thought it would be informative to share one of them.

One reader pointed out that in the past 20 years, midcap value stocks had outperformed small-value stocks. For the period from July 1994 to June 2014, the Russell 2000 Value Index returned 11.04 percent while the Russell Midcap Value Index returned 12.47 percent. Thus, for that 20-year period, it looks like midcap value stocks outperformed small-cap value stocks by 1.43 percentage points per year.

Before taking a deeper look at the returns data, it’s important to note that, as with any factor, there’s a non-zero chance small value will provide a negative premium. This is true no matter what the factor or how long the horizon. Thus, while it would be an unexpected outcome, we shouldn’t be totally surprised if midcap value stocks did outperform small-value stocks over a 20-year period, or even a longer one. That’s one reason investors should diversify investments across factors and not concentrate all their eggs in a one-factor basket.

I was a bit surprised at the performance gap, so I thought it worthwhile to look at the evidence in more detail. For another point of comparison, I reviewed the returns for the Dimensional Fund Advisors Small-Cap Value Fund (DFSVX). (Full disclosure: My firm, Buckingham, recommends Dimensional funds in constructing client portfolios.)

I chose that fund because, since inception, it has done an exceptionally good job of tracking the benchmark Fama-French Small Value Index (ex-utilities). From April 1993 through May 2014, DFSVX returned 12.73 percent per year, virtually matching the performance of the Fama-French index and underperforming its benchmark by just 0.08 percentage points a year, despite an expense ratio of 0.52 percent.

While the Russell 2000 Value Index returned just 11.04 percent per year for the 20-year period in question—underperforming the Russell Midcap Value Index by 1.43 percentage points per year, as discussed earlier—DFSVX actually outperformed the Russell Midcap Value Index during this period by 0.5 percentage points per year, returning 12.97 percent. In other words, it isn’t that midcap value stocks outperformed small-value stocks over the period in question.

Rather, the underperformance resulted from the Russell 2000 Value Index being a poor choice of index to replicate. Among the problems with it are issues related to the index’s rebalancing on reconstitution dates. The transparency of the adjustments made to the index at reconstitution allows active managers to exploit those forced changes by trading ahead of them. But that’s not the only problem.

The authors of the 2008 study “Long-Term Impact of Russell 2000 Index Rebalancing,” Jie Cai and Todd Houge, compared the returns of the Russell 2000 Index with a buy-and-hold strategy. Instead of rebalancing, they took the previous year’s Russell 2000 and kept it the same. The buy-and-hold Russell 2000 outperformed the rebalanced index by 2.22 percentage points per year.

The authors found the superior performance was explained by the short-term momentum of stocks that leave the index on the high end and then continue to outperform as well as stocks that are deleted on the low end and tend to have reversals after the rebalance. In addition, IPOs in the Russell 2000 lag in performance.

A four-factor analysis—beta, size, value and momentum—of the Russell 2000 Value Index for the 20-year period ending May 2014 reveals that the index had an alpha of -1.32 percent per year. While the t-stat of the negative alpha was just -1.5, that isn’t significant at the 95 percent confidence level. However, it certainly is economically significant, and the confidence level is still pretty high—83 percent.

An important lesson to take from this tale is that the choice of benchmark index for a fund to replicate can be just as critical as the asset allocation decision itself. There’s also a second lesson here: You really shouldn’t be surprised if the “unexpected” happens.

For the 10 years ending June 2014, midcap value stocks actually did outperform small-cap value stocks. The Russell Midcap Value Index returned 10.66 percent a year while DFSVX returned 9.32 percent a year. While we should expect small-value stocks to outperform midcap value stocks, it’s certainly not guaranteed. And that’s why it’s prudent to diversify across factors.

Larry Swedroe is the director of research for the BAM Alliance, a community of more than 130 independent registered investment advisors throughout the country.