In the early eighties, researchers noticed that small stocks had been outperforming large stocks – for no obvious reason – for a very long time. This size premium has attracted a lot of attention and a lot of criticism over the years, raising the possibility that it may have been an arbitrage opportunity that dried up when it became well known or even just a case of data mining. But AQR Capital Management managing principle Cliff Asness says that the effect still remains if you take quality into account.

“Small quality stocks outperform large quality stocks and small junk stocks outperform large junk stocks, but the standard size effect suffers from a size-quality composition effect,” he writes with his colleagues Andrea Frazzini, Ronen Israel, Tobias Moskowitz, and Lasse Pedersen in their paper Size Matters, If You Control Your Junk

SMB hedged v unhedged 0115 Cliff Asness

Cliff Asness: Controlling for junk answers common critiques of the size premiium

Cliff Asness starts by outlining the main critiques of there being a size premium. Some researchers have found that it’s not statistically significant after the 1980s (implying that it was a bit of arbitrage that no longer exits); that the excess returns mostly happen in January and could be a side effect of re-balancing and ‘institutional and liquidity frictions’; that the result isn’t robust when you look at international markets (looks like data mining); or that the effect is concentrated in micro-stocks and of little use to institutional investors because of low liquidity and the small size of the trades involved.

But when you control for quality (using profitability, profit growth, stability of earnings, and other measures) the size premium reasserts itself.

“Controlling for quality/junk, the size premium emerges as a much larger, more stable and more robust return premium,” he writes.

The distribution of quality stocks has shifted over time

For example, part of the reason the size premium started to disappear in the early 80s was simply that the number of small, high quality stocks started to decline, while the reverse was true for large caps. Whatever the reason for this trend, if large cap stocks start exhibiting more signs of being high quality businesses than in the past it’s reasonable that returns should improve.

size dist junk stocks 0115 Cliff Asness

size dist qual stocks 0115 Cliff Asness

But if you restrict yourself to comparing stocks with the same level of quality, what Cliff Asness calls the Quality minus Junk factor or QMJ, then the relationship holds: smaller stocks tend to outperform larger ones. It might still be true that the effect is data mined, and Asness has just found another lever to work with, but if not it could help indexers improve their returns.

See full PDF below.