Sunday, April 22, 2012

Robeco Low Volatility Conference

Last week I was at a conference sponsored by one of the world’s leading low volatility fund managers, Robeco, over in Rotterdam. Low volatility investing was really pioneered there by Pim van Vliet, who was influenced by papers by Bob Haugen among others in graduate school (eg, Commonality of Stock Returns, 1996 JFE). That is, Pim believes, as I do, that higher risk generates lower-than-average returns, and this creates a great investing opportunity because obviously you can get lower risk and higher return by investing on risk; there is a dominant strategy implied by the empirical data if you believe in the normative implications of modern finance. This intuition guided him in building one of the world’s first institutional low volatility focused investment strategies. They have about $2.2B euros allocated to low volatility investing.


 Now, Pim is an anti-Taleb. That is, Nassim Taleb suggests that most economists are unaware of uncertainty, fat tails, overconfidence, or that theory is different than reality, all paradigm changers. That these insights all have long academic threads is important to acknowledge, however, because it implies their lack of obvious application is due to something important that would be useful to know. For example, fat tails do not alter any of the main insights from the gaussian assumption when calibrating risk premiums, everything works in the same way, just a little more or less. Thus, they can usually be ignored, because they are isomorphic to assuming people have greater risk aversion, or that gaussian volatility is higher than its measured volatility by some fraction. Suggesting this phenomenon is the key to why mistakes were made is very misleading, and leads to inefficient or wasteful correctives.

 In contrast, van Vliet is eager to highlight he is not saying something that is new so much as important and true. In Haugen’s case, his writings did a nice job of demonstrating raw volatility being inversely correlated with returns, which is nice because all you need is for the relation to be flat for low volatility investing to be superior. Subsequent to his discovering this phenomenon, he found my work, and thinks it’s fabulous that I am championing this idea, and so invites me to speak on this subject to his clients. I guess he’s trying to convince people that if other people found this independently, it’s really there. A pioneer, but not a singular one.

 By acknowledging others so vigorously he's showing a lot of intellectual modesty, but not so much that he does not think he is right and the conventional wisdom is wrong on something very important. Thus, he has managed a portfolio that combines the singular low volatility focus with complimentary factors such as quality and momentum, innovations that are straightforward, but it's hard to riff on a theme unless you have confidence that comes from knowing you aren’t hiding anything from others or yourself. It isn’t obvious how and when to combine good things, as any good chef knows, and they have something better than your simple SPLV ETF. A good example is how Huij, van Vliet , de Groot, and Zhou (2012) discuss how distress risk can compliment volatility metrics, as they are really different manifestations of the same thing.

 Now, Bob Haugen’s presentation there was very lively, and he stated he gave 60 talks on low volatility last year, mostly overseas.  I can see why because he is very provocative and not dry. One of his points was that he discovered low volatility anomaly back in a 1975 paper that even Haugen did not remark upon for several decades. If you read this papers you see that he indeed did say that higher risk did not generate a return premium, but his explanation centered on ‘market inefficiency.’ What does it mean that markets are inefficient? Well, all that means is that markets are predictable in at least one but perhaps several ways, and so it encourages one to adopt any of those dopey get rich quick strategies that crowd the investments section at the bookstore. Just as Haugen went on to highlight a variety of predictable patterns unrelated to volatility over the next 35 years, most readers of these papers would not take-away the idea that low volatility investing is superior to the indices.

 I would say Haugen discovered low volatility investing in the same way Vikings discovered America. He found it before most, but he didn’t know what it meant when he found it, and only with hindsight figured out it directly underlies something very valuable (ie, low volatility investing).

 There’s about $12 Trillion in the US equity market, and currently about $10-$20B worldwide is directed at low volatility. The discovery of low volatility investing as an attractive idea is a better Sharpe ratio improvement than the move from active to index funds because you can reduce volatility by a third and increase the returns by a couple percent, so I think this trend will have legs, and low volatility will grow 10 to 100 fold over the next ten years.

8 comments:

cig said...

20b on low vol sounds like an underestimate to me, how did you came up with that number?

You also get people who do low vol by accident, that is use some criteria which is a proxy for low vol even if the intention is something else, and stumble thus on the same advantage. They matter from a capacity viewpoint.

Anonymous said...

Eric,

Of the Low Vol ETFs, which one(s) are you most impressed with?

Eric Falkenstein said...

SPLV!

Anonymous said...

Any color on the distress measure Robeco are using? Would be interested to see if they are leaning towards here http://papers.ssrn.com/sol3/papers.cfm?abstract_id=770805 or here http://www.business.uts.edu.au/qfrc/research/research_papers/rp112.pdf?

James said...

Sorry, I'm not convinced. This "low volatility ETF" won't be if a black swan event occurs -- no ETF can excuse itself from Taleb's principles by merely holding stocks that exhibit low volatilities during quiet times. In a severe crash, all stocks are sold indiscriminately, including Kellogg, Procter Gamble, Heinz, etc. They really should label this ETF for what it is, a Low Beta ETF, and don't imply that loss of principal is somehow less likely during a Black Swan.

Eric Falkenstein said...

Well, if you think the equity risk premium is less than zero, all equities are lousy investments. The key, many think the equity premium i greater than zero, in which case low vol dominates.

JWO said...

I think this trend will have legs, and low volatility will grow 10 to 100 fold over the next ten years.

If this is correct will the premium disappear?

Anonymous said...

Why S&P index over the MSCI?