Smart beta ETFs look tasty

Smart beta products have been hailed as “the next big thing” — so many times, in fact, that it’s become a cliché. The marketing muscle behind these products is equally enthusiastic — just google “smart beta ETFs” and look at the ads. But before you take a bite of these passive investment treats, shouldn’t you know what’s in them? How do you distinguish one smart beta exchange-traded fund from the other? And just how smart — and how beta — is that smart beta ETF you are considering?

In the days of simple equity portfolios, investors knew exactly what they were getting: maybe some Ford, some U.S. Steel, some GE. But the investment landscape has changed. Actively managed strategies — not to mention active managers — are being pushed aside by a burgeoning selection of passive alternatives.

The numbers don’t lie. The flow of money into actively managed funds has declined steadily, while flows into passive funds have ballooned. In fact, 2015 saw more than $207 billion flow out of actively managed funds, while inflows into passive funds soared for the second straight year by more than $410 billion. And half of all the ETFs launched in 2015 were smart beta ETFs.

So passive investing is hot, and smart beta products are the hottest of all beta offerings. But for all that heat, total returns in 2015 left most investors a bit cold, with actively managed returns falling 2.2% and passively managed returns falling even further, down 2.7%. That is not outperformance by anyone’s measure.

Given all the hype and hyperbole, what do we really need to know about smart beta?

First of all, smart beta is a quant strategy. The goal of smart beta is to capture returns from defined systematic factors, including style factors, such as momentum, value and liquidity, as well as industry factors, currency factors and so on. The factors represent the systematic return that investors are attempting to capture, and the goal is to classify and find portfolios that allow us to invest in those factors. Just to be explicitly clear: Quant strategies are not arbitrary. Unlike active portfolio management, where personal judgment, experience and any number of other intangibles might come into play, quant strategies are by definition transparent, rules-based, repeatable and scalable.

And here’s where we encounter one of our first problems. Many providers define smart beta strategies simply as index-based alternatives that are not traditionally market-capitalization weighted. The issue with this definition is that it doesn’t say what smart beta is, it says what smart beta is not. And in doing so, it creates a tremendous amount of wiggle room for classifying things as smart beta, when they are not.

Why does this matter? It matters because smart beta products that might look and sound identical are, in reality, often very different.

Axioma recently analyzed four smart beta ETFs from four major providers. All, according to the marketing material, focused on exposure to the dividend yield factor. The results of our analysis were unambiguous. While all the ETFs had dividend yields of about 3%, they all went about attempting to reach that yield in different ways. For example, the number of names in the ETFs ranged from a low of 76 to a high of 426. Sector exposures also varied substantially, with some invested heavily in one area, with others focused elsewhere.

Among all the tests we conducted, the most interesting results came from a performance attribution analysis, which showed that the returns from these funds were driven more by low-beta and low-volatility factors than by the high-dividend-yield factor — which actually detracted from that return! And yet, low-beta, low-volatility and high-dividend-yield ETFs are all marketed as very different products.

What this tells us is that these products are not smart beta products, but more akin to multifactor products, with imprecise exposures to their targeted factors. Again, the results of the performance attribution analysis demonstrate that the marketing story attached to the ETFs we analyzed — that they are high-dividend-yield ETFs — is simply not accurate and not rigorously executed. To reiterate, they did achieve the goal of high dividend yield, but they each did so in very different ways, which ultimately led to different performance.

The evidence is clear that current smart beta strategies do not offer pure factor exposures, at least not at this stage. To achieve that goal, a greater level of quant discipline must be brought to the process, particularly with respect to portfolio construction. Capturing risk premiums associated with systematic alpha is not just about identifying factors, but creating portfolios that capture them most effectively. Key considerations not only include which stocks to choose and how to weight them, but how to maximize exposures to targeted factors, while perhaps minimizing exposures to others — all while holding down trading costs associated with portfolio rebalancing.

But above all, just as we have benchmarks and risk-analysis tools to obtain insights into the performance of active managers, we must have a similar set of analytical tools that provides transparency to smart beta and other quant-based strategies. Tools that convey the “ingredients, nutritional value and potential risks of smart beta, so that investors know exactly what they are sinking their teeth into.

Sebastian Ceria is CEO and founder of Axioma Inc.