From the October 2007 issue of Wealth Manager Web • Subscribe!

Roll Reversal

Strategic-minded investors have discovered (some might say rediscovered) commodities as an asset class in the 21st century. But the recent surge of interest in raw materials doesn't make storing barrels of oil and bushels of corn any easier.

Buying and selling physical commodities is difficult and expensive, which is why financial investors tap commodities by proxy. Futures contracts simplify the headache of trading the real thing. But there's no free lunch: The derivatives introduce another layer of risk separate from the one associated with commodities proper.

Futures contracts come in a range of expiration dates and prices, which creates risk and opportunity. Indeed, even a passive strategy intent on long-term exposure to commodity futures requires relatively frequent trading. Carrying a long position in oil futures, for instance, obliges recurring purchases of longer-dated contracts as replacements for the expiring ones. The associated risk is that selling a contract that expires this month and buying one that expires next month or next year--so-called rolling--can produce a loss or profit. It all depends on the shape of the futures curve at the time of rolling.

A downward sloping curve is said to be in backwardation. That is, contracts expiring further out in time trade at lower prices than contracts expiring earlier. In that case, selling a March contract and rolling into a June produces a gain--a positive roll yield. In contrast, if the futures curve is in contango, or upward sloping, contracts expiring further out in time cost more and so rolling into longer-dated contracts loses money--negative roll yield (see chart on page XX).

The potential for losing money from negative roll yield adds to the fundamental risk tied to a commodity's outlook for its spot, or cash price. Consider that the spot, or near-term-contract price for crude oil has been in a bull market for five years. Nonetheless, it's been possible for long-only traders to lose money in oil futures in recent years because of contango's frequent presence.

Another complication is the fact that the standard investable commodity indices make little or no effort to minimize the fallout from the changing shape of futures curves. In fact, roll returns towered over spot returns as factors driving Goldman Sachs Commodity Index's (GSCI) excess returns (performance over T-bills) during the period from 1982 to 2004, according to a recent paper in Financial Analysts Journal ("The Strategic Value of Commodities," March/April 2006). The lesson: commodity investors ignore the roll return at their peril.

The caveat inspired Kochis Fitz to search for a better investable commodity index for its asset allocation strategies. After 16 months of research, the San Francisco wealth manager designed (in partnership with Goldman Sachs) the S&P GSCI Enhanced Commodity Total Return Strategy Index, which Goldman launched in July as an exchange traded note (NYSE: GSC). Although the new index underlying the ETN holds the same commodities in the same weights as per the standard GSCI, contract rolling is managed with an eye on exploiting opportunity--or at least, minimizing losses. At the same time, the enhanced rules are said to be designed and applied in a systematic, transparent manner that befits an index fund.

Economist Jason Thomas, chief investment officer at Kochis Fitz, says that its modified GSCI dramatically outperformed its standard counterpart by an annualized 900 basis points for the 10 years through July 31, 2007. It's anyone's guess if that back-tested premium will continue, although Kochis Fitz expects that the new benchmark will deliver a durable premium of some degree over the standard GSCI. The firm has put real money behind the bet on its modified commodity index to the tune of an initial investment of $70 million. Meanwhile, Kochis Fitz moved clients--who have an average portfolio of $5 million--out of another ETN that tracked the standard GSCI.

It's too soon to say if others will follow Kochis Fitz's example, but the opportunity is huge. The standard GSCI commands the lion's share of assets among investable commodity indices with $60 billion tracking the benchmark globally as of this past May, according to Standard & Poor's, which acquired the index from Goldman earlier this year.

For some details on Kochis Fitz's embrace of the enhanced GSCI, we recently talked with Thomas, who previously worked as chief economist for the fund-of-funds business at Wilshire Associates and in the private wealth management group at Goldman Sachs.

Why did you design an alternative to the standard GSCI?

I decided in early 2006 that the existing [commodity index] products weren't meeting our needs, which are, of course, the needs of our clients. Initially, we chose the Goldman Sachs Commodity Index. At the time--early 2006--the problem was that the oil market was in steep contango and we were losing money every month with the roll. We spoke to a number of different firms about various strategies. Commodity trading advisors and some hedge funds were making a lot of money from the situation, and I thought that we've got to find a way to incorporate some of those strategies into our portfolios.

We finally hit on a partnership with Goldman Sachs, which uses a number of interesting strategies for its commodities' trading that modify the commodity rolls. In effect, Goldman had a menu of various strategies. Ours was an interaction with the institutional group at Goldman--the firm's internal commodity trading desk where traders are intimately involved with the commodities markets. We worked with them on a number of strategies and hit on one that we thought was the best.

We were also adamant that the strategy needed to be in an exchange traded note [ETN] and not a mutual fund, ETF or a private-placement structured product.

Why an ETN?

ETNs allow the most flexibility for Goldman to manage its liabilities, which is providing the return of the underlying index. That, we think, results in a lower fee. Also, the strategy is much more tax-efficient [for investors] in an ETN rather than an ETF or mutual fund.

What exactly are investors buying with the new ETN?

They're buying the return of a basket of commodities that's the same as the standard GSCI in terms of composition and weights. The difference is the roll management.

There are three roll modifications: The first one, which adds the most value, takes advantage of seasonality. There's a growing cycle for corn, for instance. For cattle, there's a calving cycle. For heating oil, there's big demand in the winter, low demand in the summer. The standard GSCI pays no attention to the fact that commodities go through cycles. But seasonality factors cause predictable changes in the futures curve throughout the course of the year. If you roll every month, you make a lot of money at certain times of the year, and you expect to lose money at other times of the year. You hope that it nets out to a positive.

Can you give us an example of how managing seasonality can work to the new index's advantage?

Take heating oil, where you make a lot of money in the winter, and you lose a lot of money in the summer because of negative roll yield. Instead of doing that, our index rolls from January to the next January contract; we jump past the period in the year when we're losing money.

For other commodities, the timing of the roll varies, depending on the seasonal pattern, which exists for structural reasons. Commodities generally are difficult to store and transport, and so it's expensive to do so. Consider heating oil again, where all the demand comes in a short period of the year. It's a systematic trend, and for a good reason. An arbitrage that tries to exploit the trend would be to buy heating oil in the summer and store it. But that's difficult and expensive, and so the seasonal pattern is difficult to arbitrage away.

Seasonality is one modification. What's the second?

The so-called dynamic roll for oil, which added a lot of value in the backtest. The methodology for crude oil looks at the shape of the futures curve every month. A curve that's in backwardation has a positive yield, and so we earn money every time we roll. In those cases, there's no difference relative to the standard GSCI. But we roll less frequently when the oil futures curve is in contango--meaning that it's upward sloping so that contracts expiring further out are more expensive. That mitigates the negative roll yield. It's still negative if the curve is in contango, but it's less so.

So, the second modification applies only to crude oil, and the rules vary, depending on the shape of the oil futures curve?

Right, because there's no predictable, discernable seasonality to crude oil.

What's the third modification?

Rolling at different times of the month. The standard GSCI rolls all the contracts, starting on the fifth through the ninth of every month, into the next-nearest contract. Our index enhances the rule by rolling earlier--the first through the fifth day of the month. That means that we start selling before there's selling pressure and start buying before there's buying pressure.

In other words, you're trying to exploit opportunities created by pre-set buying and selling trades linked to the standard GSCI?

Right, although this modification is the most likely among the three to be arbitraged away. But that would first require that billions of dollars start rolling by our methodology.

Considering that the new index's rules vary depending on the state of the futures market at a given time, is it clear from the prospectus how your index is managed?

Yes. Now you're getting to the heart of my investment philosophy and the reason why I wanted to launch this ETN. We could have hired a commodity trading advisor [to manage the roll]. The reason we didn't is that I want commodities beta, and so I want something that's structural for the enhancements in the new index. I have a very skeptical view of people's ability to add value through active management. We have almost no active managers in our core client portfolios. I want something that makes sense to me as being durable for the long run.

Why should the three index modifications be durable?

Take natural gas, for example, which is one of the commodities we roll based on seasonality. We roll the contracts from January to the next January. I don't need someone in the market to make a mistake in order for us to outperform. Natural gas is used in the winter and it's expensive to transport and store, and so there's a structural reason why our methodology should outperform over time versus the standard GSCI's natural gas exposure.

Overall, the three enhancements may sound like minor tweaks, but there's a huge outperformance to be gained by making these tweaks.

There are other publicly traded commodity funds that manage the roll in various ways. Why didn't they suffice?

Yes, there are a few of them. But the problem with the other indices is that they don't give us the beta that we want. That is, they change the weightings and the components underlying the GSCI in a way that makes them less attractive to us for portfolio construction.

That raises the point that the GSCI is notable for its heavy energy/crude oil weighting relative to competing indices. Why is that bias attractive?

Our portfolios are heavily weighted toward equities. So, what we own on behalf of clients is a portfolio of companies whose primary business is taking raw materials and translating them into more valuable outputs. It's not an arbitrary decision to give oil the biggest weight in the GSCI. Energy is the biggest input for our companies and the global economy. Commodities are in our portfolio to hedge a risk that we have no control over--an uncompensated risk that's not adding any value to our [equity] holdings.

We can have academic debates about what captures commodities as an asset class. But we're not looking at it that way. Rather, we're asking, Where's our biggest risk as global equity investors? Having a little bit of commodities in the portfolio allows you to take more risk in areas where there's a sustainable and rational risk premium.

As it turns out, oil has been a fantastic return source for the past 20 years. But we give that assumption a haircut for the future, and so commodities for us are a low expected-return asset class relative to equity asset classes we own.

Why the low-return outlook for commodities?

It's a combination of things. It's a view based on our econometric analysis, but it's also recognition that there's a financial market component to commodity returns; there's a speculative component to the realized historical oil returns. There were particular crises that generated past returns for holders of financial instruments related to oil--returns that we shouldn't expect to get in the future. And the money that's recently been flowing into commodities will likely reduce the expected premium. Even if there's a Keynesian-type backwardation premium for taking the long side of the trade in oil, the more money that goes into the market, particularly long-term passive index money, the lower the premium should be. But commodities are still a fantastic addition to the portfolio and we want it to play the role that it needs to play, and the way to do that is primarily through oil.

James Picerno (jpicerno@highlinemedia.com) is senior writer at Wealth Manager.

Reprints Discuss this story
This is where the comments go.