Factor Investing Wins in Emerging Markets
Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.
This article originally appeared on ETF.COM here.
As the director of research for Buckingham Strategic Wealth and The BAM Alliance, I am often asked to comment on articles or papers that make a case for active management being the winning strategy. Recently, I was asked to comment on a February 2018 article from Advisory Research titled “Emerging Markets: Active vs Passive.” At the end of 2017, the firm managed more than $7 billion in assets.
Not surprisingly, given the article’s position, Advisory Research presents evidence from an eVestment database of more than 300 institutional actively managed emerging market (EM) funds that shows a majority of active managers outperformed their EM benchmarks, and did so by a wide margin (on average 1.57%). The article thus concluded active management is the winning strategy in EM.
Before you jump to that same conclusion, let’s examine some other evidence. I’ll begin with a look at the latest S&P Dow Jones Indices SPIVA scorecard, from year-end 2017.
SPIVA data shows that, for the last 5-, 10- and 15-year periods, 78%, 85% and 95%, respectively, of actively managed, publicly available EM mutual funds underperformed their benchmarks. That’s a stark contrast to the figures presented by Advisory Research.
In addition, over the 15-year horizon ending 2017, on an equal-weighted (asset-weighted) basis, active EM funds underperformed by 2.6 percentage points (1.5 percentage points).
It is possible that some of the difference in the results could come from the choice of benchmark index. S&P Dow Jones Indices uses its own index in the SPIVA scorecard, not the MSCI index used by Advisory Research. While I don’t believe it’s an issue in this case, often, proponents of active management will choose the easiest-to-beat benchmark (such as the Russell 2000 instead of the S&P SmallCap 600 Index or the CRSP 6-10 Index for small stocks) to make their argument.
An explanation for the difference in the results from the two databases might be that active institutional funds have lower expense ratios than publicly available active mutual funds. But given the magnitude of the underperformance recorded in the SPIVA data that most likely cannot be the answer. A more plausible explanation is that we don’t really have an apples-to-apples comparison in the eVestment data, at least as presented.
Advisory Research compared active EM funds to the MSCI Emerging Markets Index, which is similar to the S&P 500 Index in that it consists of large-cap companies in its respective market. It’s possible the outperformance shown by institutional EM funds was due to their managers loading more heavily on small and value stocks than the large-cap index used in the analysis.
Factors work in emerging markets
In their 2016 study, “Emerging Market Portfolio Strategies, Investment Performance, Transaction Cost and Liquidity Risk,” Roberto Violi and Enrico Camerini found that factor-based investment strategies historically have worked in EM as well as in developed markets. We can see that by examining the returns of three passively managed funds from Dimensional Fund Advisors (DFA). (Full disclosure: My firm, Buckingham Strategic Wealth, recommends DFA funds in constructing client portfolios.)
The longest period for which data is available for all three funds begins May 1998. From May 1998 through February 2018, the MSCI Emerging Markets Index (again, a large-cap index) returned 8.1% and, according to data from DFA, the comparable DFA Emerging Markets Portfolio (DFEMX) returned 8.8%. The DFA Emerging Markets Value Portfolio (DFEVX) returned 11.2% and the DFA Emerging Markets Small Cap Portfolio (DEMSX) returned 11.8%. Both DFA’s small-cap and value funds outperformed the benchmark MSCI index by more than 3 percentage points. That’s more than twice the margin of average outperformance reported in the eVestment data shown by Advisory Research.
Thus, it is certainly possible the results reported were not risk-adjusted for factor exposures. In other words, the alpha reported from the analysis of the eVestment data might just have been nothing more than beta, or loading on common factors. If this were, in fact, the case, and the active institutional EM funds’ returns had been properly compared with more similar indexes, perhaps the alpha would have disappeared entirely – or turned negative.
There’s one more bit of evidence we can examine. While Morningstar’s data contains survivorship bias, which makes active management look like it performed better than it actually did, we can examine its recent 15-year rankings to see how the three passively managed funds from DFA stacked up.
Data is as of March 21, 2018. Morningstar ranked DFEMX in the 33rd percentile, DFEVX in the eighth percentile and DEMSX in the third percentile (with the first percentile being the best). Even with survivorship bias in the data (and at 15 years it can be quite large), each of the three funds outperformed the vast majority of their comparable actively managed counterparts. Highlighting the impact of survivorship bias is that the year-end 2017 SPIVA report found just 59% of active EM funds survived the prior 15-year period.
If I adjust the Morningstar rankings on the three DFA funds to account for survivorship bias of that magnitude, their rankings improve to the 18th, fourth and second percentiles. That’s on a pretax basis.
Because the greatest cost of active management is often taxes, an even more compelling case emerges that, for taxable investors, in EMs, active management is a loser’s game.
The performance of the DFA funds suggests EMs aren’t as ripe for the exploitation of mispricing (that is, they are not as inefficient) as some active managers would have you believe. However, there’s another explanation for active managers’ poor performance.
As Violi and Camerini explain in their aforementioned paper: “The cost of trading is structurally higher in EM. In particular, comparing transaction costs in emerging markets to more developed markets, it turns out that total costs are 95% larger relative to all other markets and over double those observed in the US … . Much of this difference lies in implicit trading costs, where EM costs are over 1.5 times those of more developed markets, but explicit costs also are a factor, being 70% higher in the EM space.”
The authors found that, while the total costs of a passive strategy are about 50-60 basis points, the costs of active strategies are, on average, almost triple that, at about 1.7%. That’s a high hurdle to overcome in the effort to achieve alpha.
Whenever you hear claims of active managers as a group outperforming in any asset class, you should be highly skeptical. I’d suggest making sure any data presented is free of survivorship bias and appropriately adjusted for risk (exposure to common factors), so you actually have an apples-to-apples comparison.
The additional data I have presented make a compelling case that, at least for investors who do not have access to institutional EM funds, active management is a loser’s game. It should also make you skeptical that active management is the winner’s game even for institutional EM managers, because the data used to reach that conclusion may not have been adjusted for risk.
What’s more, for institutional active EM managers to have outperformed by such wide margins, some other group of investors must have underperformed by similarly wide margins. Because emerging markets are not markets in which individual U.S. investors are likely to own individual stocks, it must be that retail active funds are the ones being exploited. And that just doesn’t seem like a logical explanation.
Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.