The Pitfalls of Commodities and Indexing
Our look at commodities this week comes from the BusinessWeek article "Amber Waves of Pain." This article discussed some of the issues with investing in commodities and brought up some valid points. Let's look a little deeper at some of the issues raised.
One of the points of contention raised was how index (or other passively managed) funds can be exploited by traders aware of when index funds must trade to match the return of the index. (The commodity indexes roll one-fifth of their position on each of the fifth through the ninth business day of month.) In other words, index funds may incur additional costs by trying to eliminate all "tracking error" (producing returns different from their benchmark index).
This problem is well known and not limited to commodities -- it's also a problem with equity indexes. In fact, this is one of several reasons why Dimensional Fund Advisors doesn't have index funds, but creates its own benchmarks and ignores random tracking error.
Commodity funds can do the same thing. The PIMCO Commodity RealReturn Strategy Fund (PCRIX) tries to avoid such problems by not tying itself to a benchmark index. For example, it doesn't necessarily trade on the fifth through the ninth business day of the month and doesn't necessarily even use the nearest contract (as indexes do). Instead, it uses the cheapest month (has the lowest roll cost) to trade. It also mostly uses TIPS (instead of Treasury bills) as the collateral supporting futures contracts.
Finally, it's helpful if we further address the contango issue raised by BusinessWeek. From inception through June 2010, PCRIX returned 9.2 percent per year, 4 percent greater than the 5.2 percent return of the benchmark DJ-UBS Commodities Index. The difference is mostly a result of two things:
- The astute management of the rolls
- The returns on collateral investments versus the return of three-month Treasury bills used in the index
While my crystal ball remains cloudy, the following observations can be made. First, large persistent contango is unlikely to persist because it can eventually be arbitraged away. For example, if the spot price of oil is $75 and the next month contract is $80, an arbitrageur could buy oil in the spot market at $75, sell it forward and receive $80 at maturity. As long as the cost of financing this trade (including the cost of funds, storage and insurance) was less than $5, there would be a guaranteed profit. And, the very actions of the arbitrageurs would reduce the contango. (Their purchases would drive up the spot price and their sales would drive down the futures price.)
Second, investors earning lousy returns because of persistent large contango are likely to leave the asset class, thus reducing or even perhaps eliminating the problem. Investors have always chased the latest hot asset class (driving up valuations and driving down future returns) and abandoned it when returns prove unsustainable. So, while my crystal ball is always cloudy, my own personal guess is that the problem of large persistent contango is a temporary one.
An irony is that just prior to the appearance of the BusinessWeek article, the contango had been dramatically reduced. When I first read the article on July 28, the cost of the roll in the nearest contract (now September to October) had fallen from several dollars in the spring to about 30 cents (still not cheap, but much less expensive). Thus, while it's possible that the large persistent contango will continue (reducing the appeal of commodities as a diversifier of risk), it seems likely that commodities will still deserve consideration by investors as a diversifier of portfolio risk.