Canadian Investment Review

Three blunders of factor investing

Written by Yaelle Gang on Tuesday, March 19th, 2019 at 8:28 am

Business Analysis and Data Technology as a Concept © Kheng Ho Toh /123RF Stock PhotosWhile factor investing can be a powerful tool, it may not be living up to its promise because of three common blunders, according to a forthcoming paper in the Journal of Portfolio Management.

“Investors are typically led to develop their factor return expectations with little more than an extrapolation of the factor’s past paper-portfolio returns,” noted the paper, which was co-authored by Robert D. Arnott and Vitali Kalesnik of Research Affiliates, Campbell Harvey of Duke University and Juhani Linnainmaa of the University of Southern California. “But past is not prologue. We argue that both the expected returns and the risks are rarely fully understood by investors, whether retail or institutional.”

The first blunder is that investors have exaggerated expectations about how a factor will perform because of data mining and basing expectations on an overfit historical sample, says Harvey.

“Try 100 things, 99 of them don’t work and the 100th works — well, that’s no surprise,” he says. “It doesn’t really work, it just happens to fit in the historical sample. When you take it out of the sample, it’s not going to work.” When investors don’t figure this out it will lead to disappointment, he adds.

The second blunder is that investors’ risk management tools are naïve and factors can experience bigger downside shocks than would be expected. As well, investors don’t do a very good job of risk management and view these factors as near normally distributed, says Harvey. “They’ve got a view of the tail behaviour of factors that’s unrealistic. And how that plays out, as we show in the paper, is that none of these factors are anywhere close to what you would expect with a so-called normal distribution. The sort of negative realizations that we’ve seen are, if they were normal, you might see them once every 10 million years.”

The third blunder is that investors perceive their portfolio as diversified if it has several factors, which isn’t the case because this diversification can disappear in certain economic conditions when factor returns are much more correlated, the paper said.

People may assume their risk is diversified when they invest in multiple factors instead of just one, but the diversification benefit may disappear exactly when you need it, says Linnainmaa. “But the problem in the data is that when one of those factors runs into trouble, typically all of the other factors run into trouble at the same time as well.”

While these blunders exist, the paper acknowledges that factor investing can still be powerful, though it also cautions investors to understand the risk. “It doesn’t say that you should be abandoning factor investing. It’s saying that your expectations should be more moderate,” says Linnainmaa.

Read the full paper here:

Copyright 2020. Canadian Investment Review. All Rights Reserved.