By Abraham Bailin
The past several years have seen the rise of proliferative interest in commodities as an asset class. Given the recent crash and the Fed's policy response, the cascade of long-only investment demand that poured into the space was born of several needs that commodities do a good job of satisfying.
Historically, commodities have shown to provide diversification away from the correlation of traditional asset classes. On that basis, commodities were looked to as an oasis of diversification following the downturn. As the Fed executed extremely loose monetary policy, hoping to prevent a deflationary spiral through the direct injection of capital to the economy, inflationary fears began to surface. The flames of demand were fanned, as commodities were looked to for their inflation-hedging benefits.
In select cases, investors were afforded access to the spot market by physically backed commodity offerings like SPDR Gold Shares (GLD) and iShares Silver Trust (SLV). "Spot" refers to the market for immediately deliverable commodities, but in most cases it isn't feasible to maintain a large physical commodity withholding for the purpose of backing an investment vehicle and delivering spot pricing. For the broad slate of commodities, investors looking to procure direct exposure are left with commodities-futures-based products.
Today, there are a variety of these products with varying strategies, though it wasn't always so. The very first of these products tracked only front-month futures of a single commodity, a strategy that resulted in very poor returns relative to the spot price over the long term. Here we look to explain the phenomenon that caused such poor tracking, and reconcile the undeserved distaste that has led front-month futures-tracking ETPs to be considered "broken" exchange-traded funds.
A Quick Primer on Futures
Commodities-futures contracts are derivative instruments that obligate parties on either side of the trade to the purchase or sale of a fixed quantity of a commodity at a fixed price and on a fixed date, sometime in the future. Because the actual purchase and sale of the commodity occurs in the future, prices are not set at spot. Rather they are based on the expected price of the commodity upon contract expiration.
Expectations of future commodity prices being driven by extraneous factors will naturally cause current futures-contract prices to deviate from spot. Prices of futures contracts can become more expensive at progressively distant expiration dates, known as contango, or less expensive, known as backwardation. Exchange-traded products merely look to deliver the price performance of commodities to their investors, and have no interest in taking physical delivery of the spot commodity.
To avoid physical delivery, held contracts must be sold before expiration. To maintain exposure, however, contracts further from expiration must be purchased. This process of "rolling" a futures position forward will create a negative yield or loss in contangoed markets and a positive yield or gain in backwardated markets.
The First Generation
The first generation of futures-based commodity ETPs merely held the nearest-to-expiration futures contract, rolling into the next nearest contract as expiration became imminent. Some of the more popular products using this strategy, like United States Oil (USO) and United States Natural Gas (UNG), substantially underperformed their spot commodities due to a persistent and outsized negative roll yield. For instance, UNG lost 12% of its value during the summer of 2007 while the spot commodity actually gained 1%.
Upon expiration, a due futures contract will be worth no more or less than the spot commodity. The reason for this is intuitive, as a buyer will not pay more for a commodity than he would need to in the spot market. Likewise, a seller will not sell a commodity for less than he could receive in the spot market. This sets the stage for a natural convergence of futures prices to the spot price over time. That convergence tends to be felt most heavily by the near-to-expire contract, as it has the least amount of time for extraneous factors to exert price pressure. Put another way, the prices of futures contracts should be more sensitive to spot-price fluctuations the closer they are to expiration.
Not only did the persistently negative roll yield tend to be largest in the nearest-to-expire contracts, but because first-generation futures-based ETPs roll on a monthly basis, they suffered contango-related losses up to 12 times throughout the year. Proceeding offerings used strategies that tempered the issue.
By participating in futures contracts that sat further from expiration, later ETPs like United States 12 Month Oil (USL) and PowerShares DB Oil (DBO) tempered the outsized negative roll yield that arose in front-month-only strategies. Products like USL, considered second-generation, simply ladder their assets across a number of futures contracts with varying expirations. Those like DBO, considered third-generation, use a dynamic strategy to participate in a contract further from expiration, looking to avoid outsized levels of contango. In either case, these products do not roll 100% of their net assets every month, so if a negative roll yield was realized, it was much milder than was seen by first generation products.
Note that a growth of $10,000 beginning six months ago would have returned negative 0.62% for Spot WTI Crude, negative 0.57% for DBO, and negative 7.62% for USO.
The downfall of first-generation products was the fact that persistent and outsized levels of contango caused very poor tracking relative to the spot commodity over the longer term. Though later offerings provide strategies that somewhat mitigate the problem over the long term, they have a problem of their own in the short term.
Holding contracts that are further from expiration will afford investors less sensitivity to the fluctuations of the day-to-day spot price. Because non-front-month contracts have more time to expiration, extraneous pricing pressures have a higher likelihood of making an impact. It is because settlement is imminent that a front-month futures strategy will deliver greater sensitivity to the price perturbations of the spot commodity in the short run.
On this basis, we can make the argument for front-month-strategy ETPs being very well suited to short-term speculative ends and the non-front-month strategies being best suited to longer-term investment periods. Comparing the previous six-month growth to a much shorter five-day growth of the same constituents, we can glean anecdotal evidence of that being the case.
The five-day growth returns negative 2.95% to spot, negative 0.46% to DBO, but a much closer negative 1.69% to USO. With a solid hypothesis in mind, we move to ensure that our intuitive conclusions are reliably supported by day-to-day data.
To illustrate this phenomenon, we identify the tracking error of front-month and non-front-month strategies relative to the spot commodity over progressively longer time frames.
To begin, we chose USO and DBO as our front-month and non-front-month strategies. Our spot prices were those of Cushing, Oklahoma West Texas Intermediate Crude Oil. Our daily price data set extended back to Jan. 4, 2007.
Using the daily price streams of our three samples, we generated rolling five-, 90-, and 180-day cumulative-return streams. We followed by calculating the daily difference between the cumulative-return streams of spot oil and each of the other vehicles, for each time frame. By calculating the standard deviation of each of those samples of daily deviation from spot returns, we were able to identify which strategy provided returns that most closely resembled the spot commodity for each of the investment time frames.
- For the 180-day investment period, the standard deviation of USO from spot was 18% higher than that of DBO.
- For the 90-day investment period, the standard deviation of USO from spot was only 13% higher than DBO.
- For the five-day investment period, the standard deviation of USO from spot was actually 22% lower than DBO.
Over the longer investment time frame, the ability of non-front-month futures strategies to minimize the pernicious effects of contango allows for a better proxy for the spot commodity. Over the shorter time frame, certainly a time frame brief enough to avoid a negative roll yield, the higher sensitivity to spot that the front-month futures provided made for better tracking.
We arrive at these conclusions by identifying that, as the investment time frame shortens from 180 to five days, the standard deviation of daily deviations from spot returns becomes smaller for the front-month futures strategy than it does for the alternative.
As a check, we can take a look at the correlations of USO and DBO to Spot WTI Crude over a number of other time frames.
The higher correlation of DBO to spot relative to USO, as we move from daily to quarterly correlation calculations, indicates that our initial results were on target and that a front-month futures strategy does have its benefits.
Many have been scared away from using front-month futures-tracking ETPs entirely. While we do not encourage highly speculative short-term investment techniques, investors should understand the each exchange-traded product is merely a tool. It is the responsibility of the user to implement each tool correctly. Front-month strategy ETPs are not broken. Rather, their strengths lie in providing short-term commodity tracking. Again, though we would not advise such a stance, those partial to short-term speculation on individual commodities will get more bang for their buck with first-generation commodity-futures ETPs than their more sophisticated alternatives.
Disclosure: Morningstar licenses its indexes to certain ETF and ETN providers, including Barclays Global Investors (BGI), First Trust, and ELEMENTS, for use in exchange-traded funds and notes. These ETFs and ETNs are not sponsored, issued, or sold by Morningstar. Morningstar does not make any representation regarding the advisability of investing in ETFs or ETNs that are based on Morningstar indexes.