- David Blitz
- To order reprints of this article, please contact David Rowe at d.rowe{at}pageantmedia.com or 646-891-2157.
TOPICS: Analysis of individual factors/risk premia, performance measurement
After the announcement of the Peter Bernstein Award, Ronald Kahn of BlackRock interviewed the winner, David Blitz to find out what initially prompted his research, his choice of the Fama-French approach, and the key takeaways his article provides. The full winning article, ‘Are Hedge Funds on the Other Side of the Low-Volatility Trade?’, can be found in Summer 2018 issue of The Journal of Alternative Investments.
Ronald Kahn: Your article investigates whether hedge funds overall favor high-volatility stocks over low-volatility ones—in spite of significant evidence that latter perform better. What prompted you to look at this?
David Blitz: This research was inspired by recurring questions from our clients. Robeco is a leading provider of low-volatility strategies, and our clients often voiced the concern that hedge funds might be arbitraging this anomaly away. This concern tended to be based on anecdotal evidence, e.g. they knew about one particular hedge fund which claimed to be targeting the low-volatility anomaly, amongst all the other things they were doing. So we decided to investigate this matter thoroughly. To be frank, we were quite surprised when we found that, collectively, hedge funds have a huge negative exposure to the low-volatility factor, i.e. that instead of arbitraging the anomaly away they actually appear to be on the other side of the trade.
RK: Is there a connection between positive exposure to the market (positive beta) and positive exposure to high-low beta/sigma? For example, you don’t see an exposure to high-low beta/sigma in market neutral funds which have low/zero exposure to beta.
DB: That is correct, the market neutral hedge fund category has zero exposure to the low-volatility factor. This makes sense, because if your stated objective is to be market neutral, it is difficult to take a position (either positive or negative) in low-volatility. Looking at all the different hedge fund categories we find that many have a strong negative exposure towards the low-volatility factor, while the remaining ones, such as the market neutral category and some others, have zero exposure. Crucially, not a single hedge fund category exhibits a positive exposure.
RK: Are there any industry effects related to your findings? For example, do hedge funds tend to be long higher volatility industries and short lower volatility industries?
DB: The low-volatility anomaly is present within but also across industries. These industry tilts are definitely present in the low-volatility factor that we used for this study, e.g. the low-volatility portfolio contains a lot of stocks from the utility industry, while the high-volatility portfolio contains a lot of tech stocks. Perhaps hedge funds are positioned against the low-volatility factor because they like high-volatility industries. My conjecture, however, is that you will get the same result if you would use an industry-neutral volatility factor. That would need to investigated further though.
RK: It seems that academics always use the Fama-French methodology for generating factor returns while practitioners use more precise factor portfolio approaches. Why did you choose to use the Fama-French approach?
DB: The nice thing about the Fama-French factors is that they are readily available, everyone is familiar with these factors, and everyone can easily replicate the analysis should they want to. Because the relation that we find is so strong (we are talking about t-statistics of 5 or even more here) I am confident that the results will hold up no matter what definition of the low-volatility factor one uses. I guess a practical alternative could be to take an index like MSCI Minimum Volatility, but then you run into the issue that MSCI does not provide a comparable high-volatility index.
RK: What do you believe the broader market implication is to your research? If an investor were to take away one point, what would that be?
DB: More philosophically I would say that no matter how plausible the argument sounds, always let the data speak, because, as we see so clearly illustrated in this case, it might tell a very different story. More practically, there is no need to be concerned that hedge funds are arbitraging the low-volatility anomaly away, and for a proper understanding of hedge fund performance it is actually crucial to recognize their sizable anti-low-volatility exposure.
- © 2018 Pageant Media Ltd