Commodities imposed themselves in the investment arena in 2004, when emerging countries and specifically China started deeply changing commodities’ long run fundamentals.
An annual consumption growth of more than 6% on many depletable commodities (energy and base metals markets) fostered a major bull run on commodities. The fundamentals are indeed preoccupying: China alone will consume 9 million barrels per day in 2010 (i.e. 10.5% of global consumption), a rise of 70% in 8 years. If it keeps the same pace over the next 10 years, China alone will add 10 million barrels per day to global consumption. Assuming that the consumption decrease in OECD countries will compensate for the consumption rise in other emerging countries (given the growth prospects in India, it is a very conservative hypothesis), global consumption will be around 95 million barrels in 2020, a quantity we don’t know how to extract at current price levels. The long term prospects for depletable commodities prices are inescapably bullish.
Nevertheless, investors try to oversimplify macro evidence and forget some aspects of the picture when making investment decisions. To respond to massive investment demand, several high profile brokerage houses designed commodities indices to capture commodities returns in the long run. The process consists in buying commodities futures contracts and rolling to the following maturity contract before the previous expires so that there is no physical delivery. The marketing argument is highly simple:
- This type of indices generated returns in the past. It returned an average 8% a year from 1990 to 2008.
- This type of indices is strongly de-correlated to other assets returns as exhibited by the monthly correlations below:
- The long term prospects are buoyant.
What investors have not fully understood is that commodities futures returns are made of three distinct parts:
1) The spot return represents the return made when storing the commodity physically (exempt storage costs)
2) The futures rolling return represent the return achieved through the process of recursively selling the nearest contract and buying the following. Let us explain into greater detail the Crude Oil roll return as of 26 February 2010:
Crude Oil WTI April 2010 Price=79.45 USD
Crude Oil WTI May 2010 Price=79.81 USD
Assume that the investor takes a long position in the April 2010 contract on Feb 26. If the spot price is unchanged when the futures contract is rolled forward (e.g. 20% of the portfolio rolled every day between the 5th and the 9th business day of each month in the case of the Goldman Sachs Commodity Index), then the investor sells the April 2010 futures at 79.45 and buys the following contract at 79.81, taking a loss of 0.40 USD. If the curve remains unchanged a whole year, it translates into a loss of 4.80 USD per year, i.e. a 6% annual loss for a crude oil price of 79.45 USD. More generally, every time the forward curve is upward sloping or in contango (resp. downward sloping or in backwardation), the roll return is negative (resp. positive).
3) Last but not least, as futures positions require very small initial collateral, the investor’s capital is mostly invested in US T-bills so that there is a cash return component in the total return, which relates to short-term government rates.
The issue in this passive commodities strategy is that the roll return has represented a major component of commodities total returns (TR) as demonstrated by the chart below:
As a matter of fact, the GSCI Agriculture total return (representing around 20% of GSCI) trailed the GSCI Spot return by half from 1990 onwards because of the persistent contango and the systematic rolling loss. To put it another way, the GSCI total return was null since 2006 while the spot index returned 12% a year. It means that investing in this sub-index was similar to investing in an equity which would deliver a negative dividend of 12%. Who would be silly enough to do that?
Apparently many people, if we look at Riskelia’s estimates of the total inflows into Commodities Indices (chart below). These inflows have certainly aggravated the contango (see our previous article why commodity prices rise together with inventories?) and adversely impacted total return performance.
And performance has been lackluster except for precious and base metals which have experienced a moderate contango due to cheap and almost unlimited metal storage:
As popular as commodities may have been as investment assets, they cannot be called an asset class since much of the total return is linked to the shape of the forward curve. By definition, an asset class consists of a portfolio of homogeneous assets delivering a positive excess return above the risk-free rate in the long-run, corresponding to a “risk premium” or a reward for the risk associated to the so-called “buy-and-hold” strategy. Obviously this property is not satisfied for a wide range of renewable commodities, from agriculture to livestock markets. As regards depletable resources like metals and energy commodities, a price rise is certainly expected in the long-run as the result of increasing extraction costs and the necessity of leaving a part of these resources to the use of future generations (see Hoteling’s seminal paper The Economics of Exhaustible Resources). But, as illustrated by the recent booms and busts in the oil and metal prices, commodity investors will have to ride across cycles of increased magnitude and frequency, due to supply rigidities and demand’s ample variations in emerging countries.
Besides, commodity index investors tend to behave like herds, subject to irrational manias when commodities are on the rise and inclined to panic fire sales when commodities start declining (see the chart representing inflows to commodity indices above). This type of attitude inevitably creates systemic risk in commodity markets. As excessive price volatility hits large segments of the industry, threatens the livelihoods of billions of poor farmers or consumers in the developing world and endangers the political and social stability of vulnerable commodity-producing or importing countries, a fierce regulatory response is to be expected.
When investors fully realize the harm inflicted by the contango and the reality of these regulatory threats, they will necessarily look for alternatives to Commodities indices.
Two strategies make sense to gain exposure to commodities prices: active futures commodities management (potentially long or short according to the contango, prices dynamics and macro fundamentals) or investment in the stocks of metals and energy producing companies, which follow commodities price moves in the long run.