Do factor-based portfolios work?

March 19, 2012 | Last updated on March 19, 2012
1 min read

Factor-based portfolios that rely on Barra factors like momentum, value, growth and volatility don’t always cut it in Canadian markets, according to research by Pyramis Global Markets.

The study uses 20 years of data to show how some factors do dominate in the Canadian market but that they also underperform a good deal of the time. Joe Overdevest, portfolio manager with Pyramis Global Advisors, presented the research findings at an event in Toronto last week.

Overdevest noted that testing for Barra factors has been mainly limited to large markets like the U.S.—markets that are both bigger and more diversified than Canada, which is small and highly concentrated. Those unique characteristics make it more difficult to capture winners through factor-based approaches such as momentum, volatility or value.

Research also shows that Canadian portfolios based on specific Barra factors miss out on many of the top-performing stocks in Canada and end up capturing a lot of losers that negatively impact performance.

So what should plan sponsors do? The answer is good old-fashioned stock picking, said Overdevest. It might not be as exciting as factor models, but it can help capture more winners and avoid the losers in the Canadian market.

This article was originally published on