SocGen Quants Warn ‘Size Bias’ Is Killing Your Factor Returns

(Bloomberg) -- Value and quality funds have suffered this year, but don’t blame higher rates and deteriorating economic fundamentals. There’s another culprit hiding in plain sight.

It’s “unconscious size biases” baked into indexes tracking factors like quality and value.

This costly structural quirk has shaved close to 5 percent in performance from the gauges this year, Societe Generale SA quantitative analysts led by Sandrine Ungari wrote in a research report dated Dec. 13.

Benchmarks tracking so-called factors, or characteristics believed to help certain stocks outperform over time, have traditionally held shares in equal proportion, regardless of their market capitalization. While that’s been a profitable strategy in the past, broad sell-offs this year have exposed a breakdown in longstanding correlations which once helped these quantitative approaches outperform.

“In most implementations of equity risk premia, value strategies have a size bias, because of an equal weight weighting scheme, a choice driven by historical performance,” SocGen’s report said.

But from the spring onwards, small cap stocks have started to underperform large caps, punishing indexes that were constructed this way, according to the analysts.

‘Pure Factor’

Their finding underscores the stress that’s hit many quantitative investing styles in 2018. Higher rates have hurt quality shares, poor economic fundamentals have clobbered value, while a lack of clear directions has baffled trend followers, the analysts write.

Traditional factor strategies offered by quant shops are market-neutral, helping to buffer investing styles from unintended forces. But because of the costs associated with taking short positions or using derivatives to build a pure portfolio, most indexed factor products are long-only.

One way to avoid the size bias is to adopt a “pure factor” approach that weights selected stocks in a way that maximizes exposure to the factor, according to the SocGen quants.

“The performance profile of a pure factor approach, mixing quality and value, looks much more attractive over the long run,” the analysts said.

--With assistance from Cecile Gutscher and Dani Burger.

To contact the reporter on this story: Gregor Stuart Hunter in Hong Kong at ghunter21@bloomberg.net

To contact the editors responsible for this story: Christopher Anstey at canstey@bloomberg.net, Yakob Peterseil, Sid Verma

©2018 Bloomberg L.P.