In a recent discussion about value funds, triggered by my July 2018 article about how to properly view expense ratios, an investor suggested the Invesco S&P SmallCap 600 Pure Value ETF (RZV) was a superior choice to the Bridgeway Omni Small-Cap Value Fund (BOSVX), the fund my firm recommends.
The investor’s rationale was that RZV not only has a lower expense ratio (0.35% versus 0.60%), but also that, using the regression tool available at Portfolio Visualizer and an analysis using the three Fama-French factors, it had somewhat higher loadings on the market beta factor (1.18 versus 1.04), the SMB (size) factor (1.08 versus 0.90) and the HML (value) factor (0.71 versus 0.68) over the period September 2011 (BOSVX’s inception) through May 2018. Before jumping to conclusions, however, we need to explore a few issues.
As discussed in my book, “Your Complete Guide to Factor-Based Investing,” co-authored with Andrew Berkin, Bridgeway’s director of research, one of my five criteria for allocating assets to a factor is that the factor should be robust to various definitions. The value factor met that criterion (as well as the other criteria that we established; namely, persistence, pervasiveness, implementability and intuitive explanations). A strong value premium existed regardless of whether one determined value by price-to-book, price-to-cash flow, price-to-earnings or a variety of other measures, like price-to-enterprise value, dividends and sales.
Multiple Value Metrics
One of the reasons my firm’s Investment Policy committee recommends BOSVX is that the fund does not rely solely on the HML factor to determine value stocks eligible for purchase. Instead, it uses multiple value metrics. Reasons to consider multiple metrics include:
- Different measures work better in some industries than others.
- While value metrics are highly correlated, they are not perfectly so, providing a diversification benefit.
- There are long periods in which one metric works better than another. For example, in the U.S. from 2008 through 2017, while price-to-book (P/B) ratio produced a negative premium of 0.8% and price-to-cash flow (P/CF) ratio produced a negative premium of 1.0%, price-to-earnings (P/E) ratio produced a positive premium of 1.2%.
We can look at the long-term evidence on different value metrics. The following Fama-French data show the value premium over the period 1952 through 2017:
- P/B: 4.8%
- P/E: 6.4%
- P/CF: 4.7%
- P/D: 2.0%
Note how the price-to-dividend (P/D) metric provided the weakest value premium of these four measures.
For those interested, I recommend an interesting paper by O’Shaughnessy Asset Management, which suggest reasons that P/B has lost some explanatory power.
With these facts in mind, when we look at regression results showing various funds’ loadings on a value factor based solely on P/B, funds that use only P/B, or more heavily weight P/B, tend to have higher loadings on the value factor. On the other hand, funds that use multiple metrics may actually have more exposure to value stocks, but not when measured solely by the P/B metric.
We can see this at work when comparing the portfolios of RZV and BOSVX. Using data from Morningstar, the following table shows their value metrics as of March 2018:
As you can see, while RZV has a lower P/B ratio than BOSVX, their P/E ratios are virtually identical (although BOSVX’s is slightly lower). In addition, BOSVX has a much lower P/CF ratio. Looking at the total picture, it would be hard to draw the conclusion that RZV has more exposure to the value factor.
I would also add that, at least currently, Morningstar reports that even though RZV shows a higher loading on SMB, BOSVX’s average market cap is slightly lower than RZV’s ($835 million versus $854 million). Of course, these figures can change over time.
There’s another important point we need to cover. Because value and momentum historically have been negatively correlated, funds with high value loadings typically have large negative exposure to the momentum factor, which historically has been a drag on returns. Funds such as Bridgeway’s BOSVX address this problem by screening out negative-momentum stocks from their eligible buy lists and by more patiently trading positive-momentum stocks when they exit the eligible universe. Doing so reduces the negative loading on momentum, though it also reduces the positive loading on value.
Keeping that in mind, when considering a value fund, investors should also consider its loading on momentum. To do that, we need to expand our regression to a four-factor analysis.
Using Portfolio Visualizer’s regression tool, we see the following loadings:
The four-factor analysis reveals a very different picture than the one we started with. While RZV still has higher loadings on market beta and size, we now see that BOSVX has a higher loading on value and a much more favorable loading on momentum.
Another important point to consider is the funds’ alpha, which can be affected by trading strategies and fund construction rules. The regression analysis also shows that BOSVX was able to fully recover its 0.60 expense ratio and produced a very slight positive alpha of 0.02%. On the other hand, RZV produced a negative alpha of 0.59%, slightly more than its 0.35% expense ratio. Note that this is often typical, as funds not only have expense ratios, but trading costs as well.
Finally, we can look at the funds’ returns. From September 2011 through June 2018, despite its higher expense ratio, BOSVX returned 14.5% and outperformed RZV, which returned 13.5%. BOSVX also exhibited less volatility, 16.5% versus 19.6%, over that period. The result was that BOSVX had a Sharpe ratio of 0.89 versus RZV’s Sharpe ratio of 0.73.
Before concluding, there’s one other important point we need to cover. A mutual fund that is not a pure index fund has the advantage of being able to patiently trade, using algorithmic programs. An index fund or ETF does not have that advantage. Patient trading can reduce trading costs, especially in less liquid small-cap stocks. Forced trading can lead to market impact costs. That advantage could help explain the difference in alphas we saw.
The bottom line is that, before you draw any conclusion about a fund based on its factor loadings, it’s important to have a deep understanding of the measures and to make sure you are looking at the full picture. (To repeat, in the interest of full disclosure, my firm, Buckingham Strategic Wealth, recommends Bridgeway funds in constructing client portfolios.)
Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.