Has the factor magic disappeared? New research by index provider FTSE Russell suggests the answer to that question lies in the eye of the beholder.
The just-published FTSE Russell paper says recent reports of the death of factors could be premature.
Factor, or ‘smart beta’, investing targets a number of allegedly persistent characteristics buried in certain types of stocks lumped under labels such as size, momentum and value. Over the last few years investors have flocked to smart beta with about US$730 billion tracking passive factor strategies, FTSE Russell says.
“This popularity has led investors, academics and practitioners to question the extent to which factor premiums may have been arbitraged away and the future performance of such factor strategies,” the study says.
For example, a 2016 study by Rob Arnott, founder of smart beta investing house Research Affiliates, and others concludes “most factors are expensive relative to historic norms with the exception of the Value”.
“[Arnott et al] argue that since valuations are stretched in historical terms, future factor returns are likely to be lower as factor valuations revert to long-run levels,” the FTSE Russell paper says.
Also in 2016, Cliff Asness, head of quantitative shop AQR Capital Management, found “valuation levels have historically been a relatively poor guide to future factor performance” in a paper titled ‘The Siren Song of Factor Timing’.
However, the FTSE Russell study says Asness and Arnott used different valuation methods (price-to-earnings and price-to-book, respectively) to arrive at their findings.
“Despite the importance attached to valuation in general and its specific importance to factor investing, there is no consensus on the appropriate valuation metric or approach to follow to assess the valuation of portfolios as inter alia the discussion between Arnott et al and Asness highlights,” FTSE Russell says.
The report says different metrics “do not necessarily convey the same message on valuation”.
Both index and portfolio construction techniques also influence factor outcomes, according to the research. FTSE Russell found some variation between ‘standard’ and ‘pure’ factor indices (where the latter attempt to strip out any “off-factor” interference).
But a common factor portfolio construction method – dubbed selection and weighting (S&W) – did have a notable effect on valuations, the paper says. S&W techniques typically screen out, say, the top-scoring 50 per cent of factor-selected stocks that are then equal-weighted to create the final portfolio.
“The substantial active off-target factor exposures resulting from this [S&W] construction technique cause valuations to differ significantly from those of a pure Quality factor index,” the FTSE Russell report says. “Therefore, when assessing factor valuations, investors need to be mindful of the time varying influence of both on and off-target factor exposures on the relative valuation of factors, which arise from simple construction techniques.”
The NZ Superannuation Fund (NZS) handed two managers – AQR and Northern Trust – smart beta mandates (of about $1.7 billion in total) targeting value and low-volatility factors. Some local retail investors also access factor strategies via Dimensional Fund Advisors, including the Asset Class range offered in the Booster KiwiSaver scheme.