Are Equal-Weighted Indexes Smart Beta? Do They Add Alpha?
I was recently asked my opinion on equal-weighted indexes relative to capitalization-weighted indexes. Specifically, I was asked whether they add alpha. As it’s a frequently asked question, and because there are lots of assets invested in equal-weighted funds, I thought I would write about the issue.
John McQuown, then at Wells Fargo, created one of the first index funds for the bank’s pension fund, in the early 1970s. In doing the research, he found that equal-weighting the stocks got higher returns.
However, he concluded at the time that if he tried to run the fund with equal weights, the transaction costs needed to maintain the equal weighting would quickly eat up any advantage. On the other hand, with a cap-weighted strategy, the market does the heavy lifting in terms of rebalancing.
Another important point is that there is nothing “magic” about equal-weighting an index rather than cap-weighting it. The net effect is that you wind up with an index tilted more toward small-cap and value companies. Thus, it shouldn’t be a surprise that an equal-weighted index outperforms a cap-weighted index.
But whether an equal-weighted index adds alpha is a very different question. To help answer it, I’ll examine the returns of three equal-weighted ETFs using the regression tool available at Portfolio Visualizer.
I’ll begin with the largest of the three, the Guggenheim S&P 500 Equal Weight ETF (RSP), with about $13.5 billion of assets as of June 9. For the period June 2003 through April 2017, the fund produced a Fama-French three-factor (beta, size and value) alpha of 0.46 basis points per month (0.06% per annum) with a t-stat of 0.07 (not even close to being statistically significant).
However, the fund did load 0.11 on size and 0.10 on value, indicating it has a tilt toward small-cap and value stocks. In addition, unsurprisingly given the higher weighting on small stocks, the fund had a market beta loading of 1.09. The r-squared figure was 97%, indicating the high explanatory power of the model.
In other words, there is nothing really unique about equal-weighting. And it isn’t “smart beta” either. It’s basically just providing exposure to the size and value factors.
While keeping the factor tilts in mind, for the one-year period ending June 9, 2017, RSP underperformed the SPDR Russell 3000 ETF (THRK) 15.0% versus 17.9%.
For the three-year period, it underperformed 8.4% versus 9.5%. For the five-year period, it outperformed 15.7% versus 15.2%. And for the 10-year period, it outperformed 7.7% versus 7.4%.
We can also examine the performance of the PowerShares Russell 1000 Equal Weight Portfolio (EQAL) ETF, which had roughly $308.3 million of assets as of June 9. Writing for ETF.com in March 2015, Clayton Fresk, a portfolio management analyst at Georgia-based Stadion Money Management, made the case for investing in this fund.
For the period January 2015 through April 2017, the fund produced a Fama-French three-factor alpha of -21.6 basis points per month (-2.59% per annum) with a t-stat of -1.29 (indicating it’s not statistically significant). However, the fund did load 0.10 on size and 0.04 on value, indicating it has a tilt toward small-cap stocks and a very small tilt toward value stocks. Also, the equal weighting led to a loading on market beta of 1.07.
Again, we see there is nothing really unique about equal-weighting. Moreover, it certainly hasn’t been “smart beta” in this case either. The r-squared figure was 95%, indicating the high explanatory power of the model. For the one-year period ending June 9, 2017, EQAL returned 12.3%, underperforming THRK, which returned 17.9%.
There’s one other equal-weighted fund we can look at, the iShares MSCI USA Equal Weighted ETF (EUSA), which had about $130.3 million of assets as of June 9. For the period June 2010 through April 2017, the fund produced a Fama-French three-factor alpha of 5.6 basis points per month (0.67% per annum) with a t-stat of 0.52 (indicating it’s not statistically significant).
Surprisingly, the fund had slightly negative loadings on both size (-0.02) and value (-0.06). Also, the equal weighting led to a loading on market beta of 0.95. The r-squared figure was 93%, indicating the high explanatory power of the model.
For the one-year period ending June 9, 2017, EUSA underperformed THRK 16.0% versus 17.9%. For the three-year period, it underperformed 8.8% versus 9.5%. And for the five-year period, it underperformed 14.6% versus 15.2%.
The bottom line is that there’s nothing especially “smart” about equal-weighting. It’s just another way to obtain exposure to common market factors.
This commentary originally appeared June 21 on ETF.com
By clicking on any of the links above, you acknowledge that they are solely for your convenience, and do not necessarily imply any affiliations, sponsorships, endorsements or representations whatsoever by us regarding third-party Web sites. We are not responsible for the content, availability or privacy policies of these sites, and shall not be responsible or liable for any information, opinions, advice, products or services available on or through them.
The opinions expressed by featured authors are their own and may not accurately reflect those of the BAM ALLIANCE. This article is for general information only and is not intended to serve as specific financial, accounting or tax advice.
© 2017, The BAM ALLIANCE