Sunday, April 30, 2017

"Some Cautionary Notes on Smart Beta "

We've been sitting on some links about the war of words going on between Research Affiliates' Rob Arnott and one of the quantfathers, AQR Capital's Cliff Asness. This is as good a place to start as any and we'll have more next week.
From Barron's ETF Focus, April 29:

For several years, Research Affiliates’s Rob Arnott has warned about the dangers of many smart-beta strategies. He may be protesting too much.
When a pioneer of a popular investing trend amplifies his warnings about that very trend, it’s worth paying attention—even if it seems self-serving. Over the past year, Research Affiliates chief Rob Arnott has fired cautionary flares about smart-beta. He first warned investors to be mindful of valuations; now he’s focused on momentum and value strategies.

Smart-beta exchange-traded funds try to meld active and passive management, exploiting factors such as value, low volatility, or momentum to outpace long-term returns from market-weighted strategies. Investors have embraced smart beta. The industry has responded with sundry strategies, with one-fifth of all ETF assets falling under this motley category. Assets in the U.S. totaled $501 billion at the end of the first quarter, up 34% from two years ago, according to research and consulting firm ETFGI. Almost three-quarters use a fundamental approach Arnott has helped popularize that systematically weights portfolios based on metrics like book value, cash flow, dividends, or sales.

In its latest paper, Research Affiliates says funds using value factors generated only 60% of the average annual premium in returns implied by the theoretical research that often draws investors to smart-beta strategies. Momentum investors fared even worse, reaping almost no edge in returns over a 25-year period. Research Affiliates’ head of equity research Vitali Kalesnik says high turnover and trading costs associated with momentum investing may account for part of the discrepancy between results and research. Apples-to-oranges comparisons also play a role. Much of the research is based on long-short portfolios. Yet, most funds are long only, missing out on returns from shorting.

The research casts a benign light on Research Affiliates’ own approach, finding that low beta, growth, and contrarian strategies stacked up better versus the literature. “We are trying to show that some smart beta—or the way investors approach these strategies—is not that smart,” Kalesnik says.

OTHERS HAVE QUESTIONED the efficacy of factors compared with 10 to 15 years ago. Momentum is one of the most challenged, says Morgan Stanley equity strategist Brian Hayes. One reason: greater use of momentum as more investors use quantitative models....MORE
Here's the Research Affiliates link page for both the original paper and the abridged version:

Alice in Factorland
Factor tilt strategies have generally produced less alpha in live portfolios compared to theoretical factor long–short paper portfolios and have largely been unsuccessful in replicating smart beta strategies. End-investors, consequently, often reap a much smaller return from factor exposure than they expect. The winning approach to factor investing is buying the losers: Past negative performance appears to be predictive of positive future returns. 

https://www.researchaffiliates.com/content/dam/ra/exhibit/RA-Visual_Insights-003-Market_Factor_Returns_Slippage.png