Swedroe: Fund Construction Matters

July 17, 2015

“When I hear ‘smart beta,’ it makes me sick.” So said economist and Nobel Prize-winner William Sharpe after he was asked at a conference last year for his thoughts on strategies of this type.

And while much, if not the vast majority, of what Wall Street terms “smart beta” makes me sick as well, one can make the mistake of throwing the proverbial baby out with the bathwater. Unfortunately, that’s what I believe Sharpe and many others are doing.

I’ve previously explained why most of what is called smart beta is really nothing more than a marketing gimmick. It’s simply the result of loading on factors (such as size, value, momentum and profitability/quality) other than the market beta. In other words, most smart-beta products aren’t delivering alpha, just beta on other factors.

However, as I’ve also previously pointed out, there are some weaknesses with pure indexing strategies that can be minimized, if not eliminated. And creating fund construction and implementation rules that do so could be described as smart beta.

‘Smarter’ Beta

One good example of how a fund can demonstrate “smarter” beta is in its choice of fund construction rules. The point is best illustrated through a regression analysis on the four leading small-cap indexes: the Russell 2000, the CRSP 6-10, the S&P 600 and the MSCI Small Cap 1750.

The following table shows results obtained from a four-factor (beta, size, value and momentum) regression for the period January 1994 through February 2015. The t-stats are in parentheses.

January 1994 - February 2015

Index Annual Alpha (%) Beta Size Value Momentum R2 (%) Annualized Return (%)
CRSP 6-10 0.09 1.0 1.0 0.1 0.0 98 11.3
  (0.2) (89.4) (56.1) (5.6) (-0.4)    
MSCI 1750 -0.05 1.0 0.7 0.3 -0.1 97 11.1
  (-0.1) (70.4) (29.8) (10.7) (-5.1)    
Russell 2000 -1.88 1.0 0.9 0.4 -0.2 96 9.1
  (-2.1) (57.8) (31.6) (11.3) (-8.3)    
S&P 600 -0.15 1.0 0.8 0.5 -0.2 92 10.8
  (-0.1) (43.2) (21.9) (9.9) (-5.6)    

We’ll begin our analysis by observing that all the r-squared figures are relatively high, meaning the model is doing a good job of explaining returns. In addition, all of the loading statistics are highly significant.

As you can see, all four indexes had identical exposure to market beta (1.0). During the period examined, the size premium was 2.4 percent, the value premium was 2.3 percent and the momentum premium was 5.4 percent.

The CRSP 6-10 index had the highest exposure to the size factor (which provided a relative boost to returns) but also the lowest exposure to the value factor. The negative impact of its lower exposure to value offset the benefit of its higher exposure to the size premium, and the index basically produced no alpha.

The MSCI 1750 index had a lower loading on size (0.7) and a slightly negative loading on momentum (-0.1), which were basically offset by a higher loading on value (0.3). This index too basically produced no alpha.

A similar story is told by the results for the S&P 600 index. The index’s higher value loading (0.5) was almost sufficient to offset its lower size loading (0.8) and its negative momentum loading (-0.2). The outcome was a slightly negative, but also statistically insignificant, annual alpha of -0.15 percent.

The Russell 2000’s Negative Alpha

The Russell 2000 data, however, tells a very different story. Compared to the S&P 600, it had the same loadings on beta and momentum. Thus, you might expect the Russell 2000’s higher loading of 0.1 on the size factor to offset its lower loading of 0.1 on the value factor.

But the Russell 2000 produced a statistically significant negative alpha of 1.88 percent. We see that negative alpha in the returns of the Russell 2000 versus the returns of the other three indexes.

Hopefully by now it’s clear that the choice of index a fund uses to establish its fund construction rules can make a dramatic difference in the returns received by investors.

Value And Momentum

There’s one more point we need to cover. You will note that the higher an index’s loading on value, the more negative its loading on momentum. This should be expected, because value and momentum are negatively correlated. A way that a fund can add value is by incorporating a screen preventing the purchase of a stock entering its buy range if that stock is exhibiting negative momentum.

The screen would delay the purchase of such stocks until the negative momentum ceased. Whether you would call this “smart beta” I’ll leave to you. However, the evidence suggests that imposing rules of this type can improve returns (and actually reduces fund turnover by delaying purchases that would otherwise be made).

This example demonstrates why it’s so important not only for investors to make their choice of funds based on the amount of exposure they desire to each of the factors that explain returns, but also to consider how a fund’s construction and implementation rules can impact them.

Fund Design Impacts Returns

In short, a fund’s construction and implementation rules can have significant impact on investor returns. In this case, I would say that use of the CRSP 6-10, MSCI 1750 and S&P 600 were smart beta, at least relative to the Russell 2000. Alternatively, we could call the Russell 2000 “dumb beta.”

Yet despite the evidence, as of the end of April, there was $43 billion in mutual funds and ETFs linked to the Russell 2000. That compares to $28 billion in mutual funds and ETFs linked to the S&P 600 index and $55 billion in funds linked to the CRSP 6-10. (Figures for funds based on the MSCI 1750 were not available.)

Investors in funds based on the Russell 2000 took very similar risks to investors in the other three small-cap index funds, yet earned much lower returns. Obviously, the lower returns were not the result of poor stock-picking, but poor construction rules.


Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.

Find your next ETF

Reset All