(Bloomberg) --Rob Arnott won’t let it go.
One year after battling Cliff Asness in public over whether smart beta ETFs are in a bubble, the Research Affiliates co-founder is doubling down on his warning about the hugely popular investment vehicles. He’s turning the firm’s website into a new tool for telling which funds are about to pop and has launched a line of indexes that rejigger holdings based on variables like valuation.
It’s a lot of effort in what remains a critical debate for investors in exchange-traded funds: whether price matters in collections of stocks tuned to traits like low volatility or high dividends. Arnott, a smart-beta pioneer who’s a less than neutral party because his shop specializes in cheap-stock ETFs, sees the issue as a looming catastrophe for buyers.
“Just like stocks and other asset classes can get cheap or rich, the same thing holds true for these strategies,” Arnott said. “I’d like in five years from now for people to automatically ask if the factor is trading rich.”
It’s an issue that’s dear to Arnott, who spent more than a year telling anyone who’d listen that the only reason most smart beta ETFs succeed is because people rushed into them and inflated their value. His thinking goes that with no structural or economic justification for the factor’s advance, it’s destined to revert to its mean and burn millions of investors in the process.
While an obscure point, a lot is riding on whether Arnott’s correct. Smart beta ETFs, which organize securities based on quantitative factors like volatility or cheapness, reached a record $500 billion under management this year. More than $17.5 billion ETF assets follow Research Affiliates’s RAFI smart beta indexes.
Arnott’s views are not universally accepted, with critics such as Asness, the head of AQR Capital Management, saying it’s unwise to pop in and out of almost any investment based on its valuation. But the question of whether you can successfully time smart beta ETFs is likely to get bigger as the ETFs proliferate.
Arnott says it’s regrettable that nobody’s heeding the research papers he started publishing on the topic last February. Investors accept the delusion that the past is key to predicting the future in smart beta, he says, leading to performance chasing where factors like low-volatility can become dangerously expensive. Indeed, they’re already down 12 percent since August, according to a Dow Jones U.S. market-neutral basket.
Research Affiliates launched a new suite of benchmarks in late January that purport to do a better job of predicting future returns with valuation inputs. If that’s not enough to re-up the conversation, the firm is also launching a public website on Thursday that will let users tinker with different factors to see which will deteriorate due to inflated valuations. It can also be adjusted to include variables like trading costs.
“The consultant community brings forward to their clients rosters of managers and strategies with great three, five, 10-year performance, and not one of them pairs that with whether the strategy is trading rich or cheap,” Arnott said. “Our website will put it on the front burner, it will make it so obvious that this is a very basic question.”
For example, low-volatility and even the Research Affiliates RAFI low-volatility index look historically expensive according to Arnott’s new tool, a signal that investors should reduce their weight. Meanwhile, quality and value look like good buys.
Arnott says he doesn’t advocate completely selling out of expensive strategies. Factors like low vol can be important portfolio building blocks and offer other benefits like diversification. Still, investors could juice their returns by adjusting their portfolios based on valuations, he said.
“It’s hard to beat a process in which you maintain steady tilts,” Arnott said. “The way you beat it is by making modest tilts, like fading what’s expensive.”
To contact the reporter on this story: Dani Burger in New York at email@example.com To contact the editors responsible for this story: Jeremy Herron at firstname.lastname@example.org Eric J. Weiner