When it comes to smart beta funds, asset managers and index providers are straying far from the well-recognized factors that have decades of academic research supporting them, according to new research from Scientific Beta, a smart beta index provider that’s funded by EDHEC-Risk Institute.
Felix Goltz, research director at Scientific Beta and head of applied research at EDHEC-Risk Institute — an investment-focused, academic think tank — said the problem with managers and index providers finding factors that supposedly lead to an investment reward is that third parties have not replicated the results. This can lead to unintended exposures and a misunderstanding around the associated investment risks of those factors.
This is in contrast to factors, such as value and size, that have been tested over a long period of time, he said.
“The standard factors in the academic literature have been replicated over and over again,” said Goltz. “That’s why they’re still being used.”
Goltz said the question at the heart of the Scientific Beta paper, entitled “The Risks of Deviating from Academically-Validated Factors,” is reliability. (Scientific Beta, it must be noted, offers its own indexes that are based on factors found in the literature.)
“Providers are trying to create their own versions of factors, based on proprietary studies, and they haven’t been replicated by anyone else. How reliable are these so-called proprietary factors? And if you don’t have the same level of scrutiny applied to these proprietary factors, then you’re taking a higher level of risk,” he said. “Why not use what has been subjected to wide scrutiny?”
Factors have become popular because investors can assemble and choose specific exposures to include in their portfolios, according to the report.
“An often-cited analogy is to see factors as the ‘nutrients’ of investing. Just like information on the nutrients in food products is relevant to consumers, information on the factor exposures of investment products is relevant to investors,” the authors wrote.
But they warn that factors can’t be constructed arbitrarily, just as a food company can’t define protein or saturated fat in a proprietary way, ignoring medical research in the process.
[II Deep Dive: The Momentum Factor Is Real. Too Bad It Doesn't Work.]
That’s exactly what asset managers are doing, the authors argue, because they want to find factors that can lead to higher performance, said Goltz.
Managers can define factors in many different ways, for example, and then search for patterns in backtested information, reported Scientific Beta. From a statistical perspective researchers are bound to find something that works, said Goltz.
For example, he explained, academic research on the momentum factor looks at only a single variable, the return of a stock over the past 12 months (while omitting the last month, because of a short-term reversal effect). But commercial providers and asset managers often change the time period to six or three months or define returns relative to another stock in the sector or other region of the world, said Goltz.
The paper's authors argue that most of the major index providers have their own definitions of factors. Goltz said he didn’t have information on the amount of assets industry-wide that are tied to proprietary factors.
The authors emphasized that the number of known factors is quite limited and they included the well-known ones, such as profitability and investment. The industry, however, claims to have found far more that are also far more complex, they argue.
“Why does the industry have to use definitions that are much more complex than very sophisticated empirical work in financial economics?” said Goltz.
