On the origins of factor investing

On the origins of factor investing

26-03-2018 | インタビュー

Where does factor investing come from? How does it work? Stephen Schaefer is Professor of Finance at the London Business School and a co-author of the influential 2009 report on the Norwegian Government Pension Fund and factor investing. In our Great Minds series – a set of interviews with renowned academics and investment experts - we spoke with him about how factor investing came about and went on to become a popular investment approach. We also talked about the major issues researchers should focus on going forward.

Speed read

  • Factor investing is rooted in the anomalies of the CAPM
  • Investors should be aware of their factor exposures
  • More research is needed to fully understand factors

You have worked with some of the most prominent academics in the field of finance of the past few decades. Through your work for the Norwegian Oil Fund you have a significant impact on the adoption of the concept of factor investing. How did this concept emerge? What exactly were academics looking for at the time?
“The term factor investing may be relatively new, but the ideas that underpin it have been around in different forms for quite a while now. Firms such as Dimension Fund Advisors (DFA) have been making use of concepts such as the premium on small firms for several decades. Taken individually, this and other findings such as the value premium were initially seen as small steps in the academic field of finance, and we certainly did not imagine they would become so influential on investment practice. Actually, it is only over the last ten years or so, that these concepts have caught on in a substantial way.”

“Let me tell you an anecdote that will illustrate what the situation was like 30 or 40 years ago. One of the very first anomalies to emerge in the literature was the small-firm effect, which was first discovered by Rolf Banz back in the late 1970s. I happened to be visiting the University of Chicago at the time and attended the seminar where Rolf first presented his results.”

“There are two striking things in this story. The first is that, initially, Rolf was not looking for a potential small-size effect in returns at all. He was trying to do something quite different and happened to rank firms in a way that nobody had done before: by market capitalization. This is how he discovered the extraordinary fact that the returns of small-capitalization stocks in his data were on average much higher than those on large-capitalization stocks and by an amount that was easily as large as the equity market premium.”

“The second remarkable thing, which at the time seemed totally reasonable, was the reaction of the audience. It was a very distinguished group, but their initial response was that Rolf had made a programing error, that he should go away and correct this technical mistake. Of course, there was no error and people finally took his conclusions on board.”

Individual factors are now increasingly considered as part of a broader family’

What have been the major changes in the way factor investing is perceived over the past few years?
“What has changed over the past decade is that individual factors are now increasingly considered as part of a broader family. Instead of being enthusiastic about, say, investing in small firms, or investing in value stocks, investors are getting used to the idea of somehow exploiting these investment strategies as a family and looking at the interaction between them. This is actually quite an important element and this is what is actually new.”

“In the early days it was natural to think about anomalies in the context of the Capital Asset Pricing Model (CAPM), in which many people at that time still had a lot of faith. Then, in an important paper, future Nobel prize winner Eugene Fama and co-author Ken French made things much clearer by showing that average returns on a stock were strongly related to its size (market capitalization) and its book-to-market ratio, in other words to its characteristics, and not just to beta.”

“Then, in a slightly later paper, they showed that average stock returns were also related to their sensitivity to portfolios constructed on the basis of size and book-to-market. This was a decisive step from the perspective of asset pricing theory and gave a much richer description of the pattern of return premia across stocks. And this second step was also extremely useful from an investment point of view.”

How do you see the success of factor investing after your report on the Norwegian pension fund was published in 2009? How has this report changed the way people perceive investment, in your view?
“Well, I am not sure you want to put this in the interview, but let me tell you another anecdote. Approximately five years after the publication of our report on the Norwegian pension fund, a major US investment bank invited me to talk about factor investing at one of their events. I thought: ‘Sure!’ And then they showed me some kind of booklet they had been using to pitch factor investing to potential clients. Well, on page two of this document was an extensive summary of our work on the Norwegian Oil Fund. So it does appear that, somehow, our work has been genuinely influential, although I did not know exactly to what extent for several years.”

“One crucial conclusion we came to was that a significant fraction of the Oil Fund’s outperformance, most of it really, was simply due to exposure to the very standard factors we have been talking about. But since these exposures were not built into the benchmark that the fund was using, it was viewed as ‘outperformance’. Divergence in factor exposure between the portfolio and the benchmark has important consequences for the way investors should view the performance of their asset managers nowadays.”

“Before the crisis and before the large negative returns relative to their benchmark in 2008, I think people viewed the Oil fund as an actively managed fund that was producing small excess returns relative to the benchmark with relatively low risk. Indeed, I think that’s the impression you get from reading the Fund’s own reports prior to the global financial crisis. The recommendation in our report – which they did not actually follow at the time – was: ‘if you want exposure to a given factor, that’s something you should decide: it should be in the benchmark.’ In other words, these exposures should not happen by accident, they should be the result of explicit decision by the fund sponsor.”

“There are two reasons for this. The first is that different factor exposures are accompanied by different types of risk, and the fund sponsor should decide whether or not these risks are acceptable. The second reason is that, if the fund decides to take on such risks, nowadays there are relatively cheap ways to get different types of factor exposure. Therefore, a fund should only pay an active management fee to someone who can actually outperform a benchmark that actually takes into account exposure to these factors.”

Until we find some satisfactory explanation, there will always be a concern that factors may not be permanent

What are the most important aspects or questions academics should focus their research on?

“Let me point out two important areas where I think further research is much needed. As we have already discussed, there is now a very long list of factors in the academic literature; a sort of ‘zoo’ with a lot of different species in it. I suspect that many these species are related in some way, and if we were able to get a better grasp on the relationship between them, that would help quite a lot.”

“For example, momentum and value are generally negatively correlated, and this is in itself very surprising, because both factors apparently have a positive risk premium. Finding an uncontroversial explanation for this phenomenon would be very useful in helping to understand the interrelationships between the different factors.”

“The second area of research I would mention is the quest for a genuinely convincing theory of why these factor premiums exist. Until we find some satisfactory explanation for what is going on, I think there will always be a concern that factors may not be permanent and could, therefore, disappear at some point.”

This article is an excerpt of a longer text published in our Robeco Quarterly magazine.

Read the full article here.


当資料は情報提供を目的として、Robeco Institutional Asset Management B.V.が作成した英文資料、もしくはその英文資料をロベコ・ジャパン株式会社が翻訳したものです。資料中の個別の金融商品の売買の勧誘や推奨等を目的とするものではありません。記載された情報は十分信頼できるものであると考えておりますが、その正確性、完全性を保証するものではありません。意見や見通しはあくまで作成日における弊社の判断に基づくものであり、今後予告なしに変更されることがあります。運用状況、市場動向、意見等は、過去の一時点あるいは過去の一定期間についてのものであり、過去の実績は将来の運用成果を保証または示唆するものではありません。また、記載された投資方針・戦略等は全ての投資家の皆様に適合するとは限りません。当資料は法律、税務、会計面での助言の提供を意図するものではありません。




商号等: ロベコ・ジャパン株式会社  金融商品取引業者 関東財務局長(金商)第2780号

加入協会: 一般社団法人 日本投資顧問業協会