Being Smart About Strategic Beta
In February, Rob Arnott, founder of Research Affiliates, wrote “How Can 'Smart Beta' Go Horribly Wrong?” The article has since been widely cited--including by a story memorably headlined, “The Godfather of Smart Beta Says That Smart Beta Is Stupid”--as well as disputed. Most notably, AQR chief Cliff Asness penned an aggressive rebuttal.
The topic: To what extent should investors consider current market conditions when investing in “smart beta” funds? In a nutshell, Asness says only rarely, while Arnott advocates greater activity. (The two agree on this topic far more than they disagree, but the media--and perhaps also the participants--enjoy a scrap.)
This column expands on Arnott’s thoughts, as given in a recent telephone conversation.
To back up: the term "smart beta" describes formula-based methods for building portfolios that break the link with market capitalization, based on strategies like Arnott's Fundamental Index. For example, his firm’s Fundamental indexes select companies that have large economic footprints, such as high revenues, book values, and cash flows, regardless of their stock market capitalizations. Funds created to mimic those portfolios are indeed index funds--but they are decidedly not neutral, relative to the stock market.
That might make them “smart” betas. On the other hand, they could be stupid betas, or unlucky betas, or fortunate betas, or accidental betas, or data-mined betas, or even glorious betas. (To complicate matters further, the inventor of the term “stock beta,” William Sharpe, says that they are not betas at all. But never mind that.) We don’t know what the financial markets will bring. That is why Morningstar renamed smart betas as “strategic betas.” We don’t know if they are smart, but we do know for certain that they are strategies.
Now to the point--strategies move into and out of favor. For example, sometimes value stocks beat growth stocks over a several-year period. If value is defined by price/book ratios, then value stocks will continue to have lower price/book ratios than growth stocks. (It can be no other way.) However, the gap between the two camps will have shrunk. Value stocks will still be cheaper, yes--but not that much cheaper. The argument for owning them has weakened.
What Goes Up ...
Arnott believes this to be important information. “Mean reversion is the most powerful and worst understood force in financial markets,” he states. It is powerful in that big moves from historic norms usually are not permanent; usually, at some place and at some time, the direction switches. (Technically, one would say that most wibbles eventually wobble.) It is misunderstood because the timing of these reversions is so difficult to predict that it hampers the effectiveness of mean-reversion tactics.
Hampers, but not eliminates, says Arnott. As he points out, everyone agrees that valuation matters when buying and selling individual securities, such as stocks or bonds. Most also consider valuation for asset allocations, tilting toward asset classes that currently look attractive (perhaps because of price, or perhaps because of fundamental reasons), and shading away from those that are not. The same would seem to hold true for strategies. “Why should smart beta be any different?”
Indeed, one can argue, paying attention to valuation is particularly important when investing in strategies. After all, to ask the rhetorical question, why do people invest in strategic-beta funds? Answer: Because their back-tested numbers impress. More than almost anywhere else, strategic-beta purveyors sell what worked in the past. The paradox is that the better a strategy has worked in recent years, the more expensive it has become, and the worse it will probably fare in the future.
That paradox underlies Arnott’s suggestion that strategic beta could go horribly wrong. Recently popular strategies--for example, low-volatility investing--might be caught in a vicious circle that masquerades as virtuous. The strategy performs well; its track record improves; more investors notice; new monies arrive; the fund puts those monies to work; the assets it buys increase in value; the strategy performs well; its track record improves; and so forth. Pete and Repeat sit on the fence, warbling happily.
Their mistake. At some point, the pattern reverses and the strategy trails, leaving those new investors grievously disappointed. They redeem the fund that (seemingly) misled them; those redemptions force the fund to sell into a declining market; the securities' prices are pushed down further; the strategy’s track record is depressed ... and we’ve been here before. Once again, Pete and Repeat sit on the fence. But this time, their tune is mournful.
That is how strategic beta could go horribly wrong.
Arnott’s advice, based on how a strategy is priced, relative to its historic norm (this norm may be calculated in isolation or it may be derived by comparing with another yardstick):
1) Top Quintile--Do not buy new positions. “Put a Post-it note on the strategy and wait a year.” Probably hold existing positions.
2) Top Decile--Do not buy new positions. Consider trimming existing positions.
3) Top 5%--Do not buy new positions. Consider clearing out of existing positions.
4) Bottom Quintile--Consider buying a new position.
5) Bottom Decile--Consider buying more.
6) Bottom 5%--Consider “loading up” on that strategy.
For an example of what the bottom 5% might look like, this chart compares emerging-markets stock price/earnings ratios against those of U.S. stocks. For Arnott, emerging-markets value stocks are indeed what he regards as a “load up” purchase. Those securities now trade at barely over 7 times their cyclically adjusted average annual earnings, as opposed to their historic average of 15.
My personal tastes are less aggressive than Arnott’s; I am tempted to make modest adjustments at the 5% mark and not to move at all for valuations that are measured in deciles or quintiles. For his part, Asness considers even 5% to be insufficient. So, there is plenty of room for disagreement on the specifics.
The general principle stands, however. Strategic-beta investments should be made only after checking their relative valuations. Perhaps these relative values will frequently affect the investment decision, perhaps occasionally, or perhaps only on Friday the 13ths, when the moon is blue, and Elvis is in attendance. But they nonetheless matter.