By Dave Nadig
Matt Hougan’s love affair with the upcoming managed futures ETF is just as naive as buying one stock and praying.
It’s easy to understand why you’re excited about the new SummerHaven/U.S. Commodity Funds ETF, Matt: It’s based on actual research, instead of just a fire-and-forget strategy of buying what’s in front of you. But I think it’s a mistake to fall in love too quickly. After all, we’ve seen these “intelligent” commodities strategies before, most particularly in those funds that don’t simply buy the front-month futures contract.
As an example, you need look no further than the PowerShares DB Crude Oil Long ETN (OLO). OLO tries to mitigate contango and profit from backwardation, using a flexible approach to decide what contracts to roll into when those being held expire. The results?
click to enlarge
Sometimes it works, and sometimes it doesn’t. Since the February bottom in oil, the more complex strategy has returned almost exactly 75 percent, the same return as the simple “buy the front-month” strategy used by the U.S. Oil Fund (USO). Both strategies have beaten the “average-out-a-year” strategy used by the U.S. 12-Month Oil Fund (USL). Of course, picking your time frame is everything—since the July highs of 2008, OLO has beaten USL by about 5 percent, “only” down 49 percent.
What’s at work here isn’t so much active management as good old-fashioned quant work, and yes, I draw a distinction there. There are structural components to the commodities markets that simply aren’t present in boring old equities—most particularly, the roll yield as a component of returns. That’s what managed futures are all about.
A great place to start in understanding these structural differences is here: “The Tactical and Strategic Value of Commodity Futures,” a paper by Claude Erb and Campbell Harvey from 2006. One of their key assertions is the most shocking, but in a way, the most obvious: “commodity indices are strategies.” In trying to make sense of commodity returns, they naturally looked at the big indexes: the GSCI, the DJ-UBS and the CRB. They found startlingly different risk and return characteristics for pretty much any time period, and attribute these differences to this core insight.
Commodity indexes are, by default, weighted and then rebalanced. Equity and even bond indexes are generally market-cap weighted and left to their own devices. Obviously this is an overgeneralization, but fundamental and equally weighted indexes outside commodities are considered to be the rogues, and nearly active management bets. But it’s the norm for commodities.
It shouldn’t be surprising, then, that commodity ETFs have sometimes curious returns. Funds like USO aren’t indexes, they’re proxy portfolios. Nobody who buys USO is buying “the oil market”—they are instead mimicking the returns of a single, very narrow strategy: “Buy front-month oil and roll it every month.” It’s no more sensible an ETF strategy than “Buy all the stocks that end in–'yme'” in the absence of academic finance backing up the approach.
The strategy you’re in love with Matt—the one where the ETF will simply hold those commodities in backwardation due to low inventories—will in fact backtest you some fine returns. But these returns are also present in dozens of other varieties when you dig into the historical futures market data. Erb and Harvey present the case for a simplistic long/short strategy where you just go long contracts with good 12-month momentum and short those with bad 12-month momentum:
Source: “The Tactical and Strategic Value of Commodity Futures,” Erb and Harvey
The message here is really quite simple, and one worth repeating: All indexes—not just commodity indexes—are in a way an active bet. Your choice of the S&P 500 as a proxy for U.S. equities as an asset class is a choice not to buy an equally weighted index, a fundamentally weighted index or an index of more than 500 stocks. In the case of equities, we’ve got mountains of empirical data to make you comfortable with that decision, because stocks are a fairly simple asset class.
With futures, however, there’s a whole lot more going on. That means your choice of proxy is much more important, and inherently much more active.