Thanks to new research, factor-based investing is more intelligent than ever, but some “smart beta” products are still dumb.

That was the conclusion of panelists at FPA’s BE Baltimore conference last week, who said that multifactor investing is still in its infancy as academic research advances and asset managers launch hundreds of new funds to keep up.

“We’re seeing a growth in factor investing not only because data and tech have given us tools to bring scaled exposures, but also because the DOL is creating a demand for fee compression within the industry,” said Holly Framsted, a strategist with New York-based BlackRock’s iShares smart beta team, at the conference’s Factor Investing 2.0 discussion. “That’s why there are a lot of products hitting the markets.”

BlackRock has been involved with factor-based investing since the 1970s, when it launched an early computer model-based strategy that re-ranked the S&P 500 based on dividend yield.

At the time, BlackRock’s yield-based index was considered active management, said Framsted. More recently, the firm has provided access to smart beta factors through its suite of iShares ETFs.

Now, smart beta products are a $250 billion market in the U.S.—but Framsted says they have room to grow.

“The fastest-growing asset class has been minimum volatility, but there’s certainly been a lot of growth in factor-based investment across the board,” Framsted said. “If you look at flows, flows into minimum-volatility strategies make up four basis points of the underlying market capitalization of the securities they’re holding. We’re still a long way from seeing saturation in any of these factors. At this point, it’s still a tiny fraction of the investable universe.”

Yet scores of new “factor-based,” “strategic beta” and “smart beta” products have been released in 2016 as other major asset managers play catch-up to peers like BlackRock and specialized smart beta shops like Dimensional Fund Advisors. Most recently, Fidelity’s launch of its suite of six smart beta ETFs happened to coincide with the conference.

The responsibility for navigating the crowded smart beta waters falls squarely on an advisor’s shoulders, explained Jay Jacobs, director of research for New York-based Global X Funds.

“It doesn’t help to have so many names and hundreds of different products in the ETF space,” Jacobs said. “Many of them represent an oversimplification of factor-based investing, but they all sound the same. If you take into consideration how they’re accessing factors, you start to see huge differences within these portfolios. It puts the onus of doing the homework on the advisor or investment professional.”

Global X started down the smart beta path in 2011, when it launched SDIV, its Superdividend ETF, that attempts to offer exposure to the highest-yielding equities in the world. Since then, Global X has launched a suite of multifactor ETFs designed in partnership with academic researchers at Europe’s EDHEC business school.

The panelists agreed that there were four to six factors for equities, some concentrated around producing pricing anomalies that lead to greater total returns, like value, size, quality and dividend growth, and some focused on greater risk-adjusted returns, like low-volatility.

Recent research that suggests a larger universe of smart beta factors is either rediscovering already identified factors or is too recent to be confirmed, according to the panelists.

“We think about a factor having a long data history that allows us to see if this anomaly exists historically,” said Abdur Nimeri, senior investment strategist with Northern Trust in Chicago. “We start with that as our basic premise, then we start to think about whether the factor impacts shareholder value, and why.”

Northern Trust approached factor-based investing from the direction of dividends as it considered how to overlay a quality screen onto its suite of dividend products. The firm now offers multifactor products in addition to its suites of core allocation and thematic funds.

Nimeri said that having a long data history separates "confirmed" factors like value and size from others that could emerge from data mining that limits the time period or area of markets in which factor research is conducted.

“We can all say that we believe in the value premium, but that doesn’t necessarily preclude it from being data mined,” Jacobs said. “A lot of what we’ve seen is redefining factors since they've already been proven by academia. Fama and French introduced the three factor model that includes both value and size, and htey used the price-to-book definition of value, and since hten you've actually seen dozens of different defintions of what value is. Think about all the different valuation metrics you can use... it introduces a bunch of different ways to measure value and this is where we think data mining starts to come into the equation, because suddenly you're opening up the question of is this really persistant in the way the academics have judged it,.”

Jacobs said that advisors should be careful to select smart beta products matching factors with measurements that academic researchers used.

The panelists identified a sixth factor, momentum, that neither fit the definition of a return-enhancing factor nor a risk-mitigating factor.

“We believe momentum is a stand-alone factor,” said Framsted. “Where does it fit in the Morningstar nine-box? Most of the time it seems to be large-growth oriented, but there’s some variability in the type of exposure you get.”

The panelists agreed that timing factors would be too time consuming and difficult for most advisors and investors, thus they all advocated diversified multifactor approaches to smart beta when a factor appears to underperform.

For example, during the 2008-2009 financial crisis, the advantages of smart beta factors were neutralized, said Nimeri, and value subsequently underperformed because many investors were risk averse for an extended period of time.

“Factors tend to be counter-cyclical,” Nimeri said. “You can have long data and short data factors, and in some cases they will appear as if the risk premium has been wiped out, but when you look over a longer cycle you’ll see a persistent premium that you can access over a period of time.”

Advisors should also keep in mind that smart beta products come with caveats, said Jacobs.

“If you look at smart beta from a process perspective, these are passive products,” Jacobs said. “There’s a methodology around all of our approaches. The experience, though, is more like an active fund. You’ll see higher turnover than a traditional index. Smart beta funds tend to have 30 to 40 percent turnover, and you’ll see more tracking error.”

Thus, Jacobs said it is important for advisors to have discriminating tastes when it comes to smart beta, and not to approach new factors, measurements and products with caution.

“I would have a healthy does of skepticism with any new factor research promising returns,” Jacobs said. “Factor investing is designed to be a long-term investing strategy