Mark Carver, executive director, Americas index products at MSCI, says that from a practical standpoint, factors look at characteristics that help investors understand the behavior of an asset. Factor portfolios are generally thought of as a long/short portfolio, where the investor is long equities that have high exposure to the factor and short equities that have low exposure to the factor, and the factor return is the difference between those two.
Smart beta is usually long only and weights by company fundamentals.
Carver also doesn’t care for the term smart beta from a semantic standpoint since “that assumes there's an opposite to that which would be dumb, but we would strongly disagree with that idea. There’s just beta.”
Another way he sees factors as different from smart beta is when it comes to creating indexes. MSCI, which has studied factors for more than 40 years, tries to isolate specific characteristics, Carver says, and organizations like MSCI have a set of factors they think earn a reward over time, such as value, momentum, quality, size, etc.
“We're building indexes against those factors that have a lot of empirical evidence that they earn a reward over long periods of time, and that’s different than what you're doing in smart beta, which could be just trying to get a thematic idea that may or may not hold up to long-term rigorous evidence,” Carver said.
Some fund issuers mix fundamental weighting with a factor, blending the two concepts. When that happens, it can be hard to separate out what’s contributing to performance.
“I calculate there are 62 [ETFs] focused on the value factor. And how do I evaluate one versus the other, and which one do I figure out is right for my portfolio? That's the big question,” Venuto said. “Considering the limited resources out there, I don't know how people are answering it.”
And Bruno says factor investing that seeks to be market neutral by going long one factor and short another may not give as much return as the hype around it would lead you to expect. The absolute return from spread between the factor may only be 1-2%, while the actual market supplies the rest of the return.
“The predictability is difficult,” he noted. “If it were always positive, that spread would be fantastic enough: 2% on uncorrelated data returned to market would be the holy grail. The problem is we don’t know when value will outperform growth [for example].”
500+ Supporting Papers?
While factor investing may have a lot of history behind it, Arnott famously pointed out two years ago in his research paper “How Can ‘Smart Beta’ Go Horribly Wrong?” that there is a lot of data-massaging to prove factors work. He says there are more than 500 papers that were published by the end of 2016 showing factor investing worked. But he thinks those papers missed an important point.
“Not a single one of the 500 papers asked the question, did we win because we had a tail wind from rising valuation because our strategy was getting more expensive?” Arnott said. “Anything that's newly expensive, I guarantee you is likely to have wonderful past returns.”
“There's the selection bias that publishing factors means you're self-selecting to factors likely currently overpriced, that are likely to work much less well than in the past… ,” he added. “A stock can get very expensive, and if there's any mean reversion, it could tumble. The same holds true for strategies. The same holds true for factors.”
Arnott also has issues with the multifactor strategies that blend two or more factors and tout that doing so smooths out the ride for investors, which he says is “a gross oversimplification. I think if you use a multifactor approach, some of the factors have merit, some of them were discovered based on noise … but you don't know which is which, so using multiple factors increases the likelihood that you'll have one or two factors that actually do work to smooth the ride.”
He’s not against factors per se, but says that fundamental weighting to buy cheaper companies can give factors more oomph. “We know that factors when they're trading cheap have high odds of success and when they're trading rich have very low odds of success,” he pointed out. “Why not be dynamic about it? Why not emphasize the factors that are trading cheaper than usual right now?”
Arnott’s firm provides the indexes underlying PIMCO’s dynamic multifactor ETFs.