The research comes with caveats. In an email, Sanger highlighted that it’s based on simulated results using backtesting. That’s a practice which some have suggested creates the “mirage” of smart-beta outperformance in the first place.
Meanwhile, there are reasons to be cautious about extrapolating the findings to the wider factor-investing universe. The rebalancing mechanism cannot explain most of the factors that are studied rigorously by academics and quant investors, according to Ashley Lester, global head of multi asset research and systematic investments at Schroder Investment.
“Most actual quants test factors by building ‘good’ and ‘bad’ portfolios of the factor, in such a way that both sides should benefit from the rebalancing mechanism,” he said.
“That way, if the ‘good’ side does better than the ‘bad,’ the result cannot be driven by rebalancing.”
It goes beyond testing. A true factor approach would involve taking both a long position on the stocks that fit most closely to the factor, and a short position in the least compliant. Smart-beta ETFs are usually long-only for simplicity and cost reasons, meaning they -- and the research -- don’t fully reflect the efficacy of factors.
Nonetheless, Olivier d’Assier, head of applied research for Asia-Pacific at Qontigo, said the paper’s findings confirm some of the market analytics company’s past studies. They have found strategies targeting a particular factor can often be overwhelmed by others if no effort is made to reduce their influence.
“The portfolios used in this study do not represent ‘pure’ factor portfolios but long-only ones designed to tilt the broad market portfolio towards a certain smart beta without neutralizing others,” he said. “It has also been our experience that these types of portfolio construction methodologies introduce a lot of noise in the performance attribution from incidental exposures to non-target factors.”
--With assistance from Yusuke Takeshita.
This article was provided by Bloomberg News.