Commodities: Uncertain Futures

Commodity indexes are pulling in triple-digit returns. But the slight differences in performance offer clues to their potential volatility
Illustration by Olaf Hajek

Investors who have grown comfortable with stock index funds over the years are piling into funds that track commodity indexes—to the tune of an estimated $200 billion worldwide. To capitalize on that demand, a host of new and, their creators say, improved products that track commodity indexes are being launched. But whether long established or newly minted, the commodity index fund is a very different animal from its equity cousin. Investors who don't understand that may be in for a rude awakening when the commodity boom ends.

Perhaps no one cares right now that the PowerShares DB Oil (DBO) exchange-traded fund (ETF) is up 106% over the past year while the United States Oil Fund (USO) is up 115%. What's a nine-percentage-point difference when you have triple-digit gains? But it must strike attentive investors as odd that two funds with identical management fees that ostensibly invest in the exact same thing--light sweet crude oil--have such a performance gap. It's not as though they're buying different stocks. Oil is oil, right?

Well, not quite. These funds aren't investing in oil directly but in oil futures--contracts that are a bet on the price of oil in the future. And just how funds place and manage those bets can make all the difference in returns.

Here's why: Futures have two other things driving their performance besides the commodities they track. Since futures are bought largely on margin, funds must set aside liquid bonds or cash as collateral against potential losses. Interest earned on collateral is added to a fund's return. Collateral is often made up of short-term Treasury bills or cash, but some funds (Pimco is one) use slightly higher-yielding bonds, which can boost their returns.

Of far more consequence for returns is something called "roll yield." It's the positive or negative return a manager gets when he sells one futures contract or rolls an expiring contract into a new one. Futures are more like bonds than stocks in that they have maturity or expiration dates. When you buy an oil future you are not buying today's price of oil--what's known as the "spot price." You're buying what investors think the price of oil will be a month from now, two months, a year--whenever the contract expires. The difference between buying near-term futures or longer-dated ones can have a big effect on roll yield, and thus returns.


The older commodity indexes invest mostly in futures with the shortest maturity dates. The logic: They tend to be easier to trade and closely reflect current prices. "Oil shocks and other unexpected price movements tend to have the greatest impact on the shortest-dated futures that are closer to the spot price," says Eric Kolts, commodity indexes product manager at Standard & Poor's (MHP). And oil shocks are usually positive for oil investors, if not for consumers. The S&P GSCI Index invests mainly in one- or two-month futures.

But creators of what are being called "second-generation" commodity indexes say that buying just short-dated futures is short-sighted. "Would you only invest in three-month Treasury bills to get a diversified bond portfolio?" asks Kurt Nelson, head of U.S. commodity index marketing at UBS (UBS). "Probably not, but that, in effect is what the older commodity indexes are doing." This April, UBS launched a slew of exchange-traded notes (ETNs) called E-Tracs that track commodity indexes by buying futures that range from three month to three years. (ETNs are bonds whose performance is usually linked to a benchmark index.) UBS thinks a more diversified approach will increase returns while reducing volatility.

To determine how its new strategy would perform, UBS back-tested it, running computer simulations using historical market data.

Experienced investors often take back-tested performance claims with a grain of salt, since there's always leeway in how to run the numbers. S&P's Kolts is among the skeptics: "If I'm going to engineer a new index, I will test it back to make sure the roll yield schedule works best in my favor." Says Nelson: "We're capturing the whole market of future maturities, not cherry-picking individual ones."

Clearly, there will be significant differences between how E-Tracs and other commodity index funds and ETNs behave. According to the back test, which ran data from July 31, 1998, to Feb. 29, 2008, the UBS Bloomberg Constant Maturity Commodity Index delivered a 20.2% annualized return over the past 10 years, compared with 14.3% for the S&P GSCI index and 13.4% for the Dow Jones AIG Commodity Index Total Return (DJAIG), the two most popular commodity indexes. The UBS index did this with less volatility as well.


A significant part of the performance difference has to do with roll yield and two factors that go along with it--contango and backwardation. We'll use an example to explain those industry terms. Let's say you have a one-month oil futures contract expiring in August that trades for $100. Now, say you want to roll that $100 contract into one that expires in September, but it costs $110. You lose $10 when you purchase that September contract. When this happens--when the prices of later-dated futures are higher than shorter-dated ones--the contract is said to be "in contango." The reverse can be true as well, of course. The August contract could be $110 and the September, $100. Then you make $10 when you buy the September contract. That is called backwardation.

What leads to those two scenarios? Many things influence expectations for the price of oil. Among them may be the weight of $200 billion in commodity index funds that simultaneously bid on contracts every month. The competition can bid up prices and that may make it more costly for funds to implement strategies, effectively reducing returns. S&P's Kolts says the market is big enough to handle such inflows without affecting prices much. UBS's Nelson, however, says, "I think it's significant that when a lot of money started to flow into the DJAIG and GSCI indexes in 2005, that's when they started to underperform our index in a meaningful way."

UBS thinks it can outperform its peers by purchasing futures several months out--in effect, by avoiding the rush. It also rolls over contracts a little each day to smooth the transition to later-dated futures. Older indexes typically roll over their futures monthly, during a five-day window.

UBS isn't the only one using a new strategy with its commodity index fund. Deutsche Bank has developed ETFs and ETNs that use what it calls an "optimum yield" strategy. Rather than just roll a one-month contract automatically into the next month's contract, Deutsche Bank's fund analyzes a much wider array of contracts that cover the next 13 months, explains Kevin Rich, Deutsche Bank (DB)'s managing director of global markets investment products. Then UBS calculates the highest potential roll yield--sort of like choosing the highest bond yield. As with UBS's product, in back tests Deutsche's indexes beat its rivals handily.

But markets have a way of throwing a wrench into things. Take the ETF cited earlier, United States Oil. It invests in the shortest-dated futures, yet bested PowerShares DB Oil, which uses Deutsche Bank's optimum yield strategy. While in the back test the Deutsche Bank index almost doubled its peers' returns, a change in the shape of the futures curve in the past year threw its strategy off. Even so, no one with triple-digit gains probably cares--yet.

    Before it's here, it's on the Bloomberg Terminal.