Professor Geert Rouwenhorst of Yale University cowrote the paper that kicked off the commodity index investing boom. Recently, however, he stepped outside the realms of the academy to launch a new commodities indexing firm that he believes may have a better mousetrap for investors. He spoke recently with IndexUniverse.com Editor Matt Hougan about why he thinks most investors can do better than a passive, long-only commodity index.

IndexUniverse.com: You’ve been described as one of the fathers of modern commodity index investing, as your academic research made the foundational case for the benefits of commodity indexes. Now you’re launching a new firm that says it can deliver better returns than a pure passive index. How do those two ideas line up?

Prof. Geert Rouwenhorst, Yale University and SummerHaven Investment Asset Management (Rouwenhorst):

There are two main commodity studies that I have been involved with. The first was “Facts and Fantasies About Commodity Futures,” which I think is what you were referring to.

The impetus to write Facts and Fantasies was that [co-author] Gary Gorton and I were facing the question, “What are the reasons investors would put money in commodity index products?”

As you know, asset managers and institutional investors typically require some data prior to becoming investors. It dawned on us that nobody had systematically put together a long time series of commodity futures returns and studied their performance.

For Facts and Fantasies, we constructed an index covering as long a period as seemed feasible from the data, 1959-2004. We constructed the simplest possible index: an equally weighted portfolio of commodity futures contracts.

We found, somewhat surprisingly, that there was a return premium over collateral being earned by passive long-only investors in commodity futures. That return premium seemed to have a very low correlation with stocks and bonds.

We found that not only are commodities great diversifiers, but they historically have offered a strong source of returns. In addition, they have opposite inflation exposure compared to stocks and bonds, and are also good diversifiers of macroeconomic risk. They also earn their return in slightly different stages of the business cycle than traditional assets.

IU: It seems strange that no one had looked at that until 2004. Commodities are so widely accepted as an asset class today.

Rouwenhorst: That paper got a lot of attention in part because the question of whether there was a risk premium paid for commodity futures had been a debate since Keynes formulated theory of normal backwardation in the 1930. Keynes hypothesized that hedgers would pay speculators to buy price insurance. But there were not many studies that even tried to measure that risk premium. The problem is that commodities are volatile, so it is difficult to detect that risk premium in short time series or in individual commodities.

The power of our study was that it measured the risk premium at the portfolio level. It’s very much like trying to measure the equity premium in individual stocks, which is difficult, versus measuring it in the S&P 500 over long periods of time, which is easier.

IU: How did that get us to where you are today, launching a new “active index” that aims to outperform these classic benchmarks?

Rouwenhorst: In response to the paper, many asked us which commodities had higher risk premiums. The question we asked in response to that was, how do I think about the fundamentals of commodity futures?

The common answer is inventories. We start with the premise that the risk premium on commodity futures represents, from a Keynesian view, an insurance premium that the short side pays to long side to obtain insurance against price risk. From there, it would seem logical to think the premium would be larger when there is more risk to insure.

This is where inventories com in—when inventories are low, price volatility (risk) would be higher.

Inventories act like buffer stocks. When inventories are low, the ability to absorb supply shocks is limited. When inventories are high, there’s enough stuff around to meet unexpected shocks.

This is not something I made up: It’s called the theory of storage, and it goes all the way back to research from the 1940s. In follow-up research, Gary Gorton, Fumio Hayashi and I have taken that theory to the data to see if it’s indeed the case that, when we build portfolios of commodities with low inventories, the risk premium paid away by hedgers is larger. When inventories are higher, risk premiums are lower.

That’s very difficult to test based on individual commodities, so we used a strategy similar to the Facts and Fantasies paper, and formed portfolios of high- and low-inventory commodities. They have to be rebalanced on a monthly basis because inventories fluctuate over time.

What we found in the paper was that, if you construct low-inventory portfolios, the risk premium on that group of commodities is significantly higher than if you hold the portfolios when inventories are high. Some of that is attributable to the fact that some commodities are difficult to store, so they have a higher risk premium.

That was in part the underlying philosophy behind the new SummerHaven Dynamic Commodity Index. If I want to try to capture that premium in a systematic away, I should periodically observe fundamental signals about the underlying commodity markets and construct an index that captures the performance of the subset of the market that is fundamentally attractive, relative to the other half. That naturally becomes a changing subset, which is why we think of it as an active index.

Previous Page

1  |  2

Next Page