The returns of commodities futures are negatively correlated with stocks and bonds. Commodity futures provide diversification at times when stocks are performing at their worst. Commodity features have a positive correlation with inflation.
These enticing characteristics are the conclusion of a study* by Gary Gorton, of the University of Pennsylvania’s finance department and the National Bureau of Economic Research (NBER), and K Geert Rouwenhorst of the Yale School of Management’s International Center for Finance.
The research is an excellent primer on commodity futures and comes recommended by the head of research at one major investment consultancy.
Gorton and Rouwenhorst take as the basis for their research an equally weighted commodities index, derived from the Commodities Research Bureau (CRB) Index, because it goes back to 1959 and because equal weighting means that everything from soy beans to lean hogs to copper gets included. Illiquid contracts, however, which have fallen into disuse through lack of popularity, do not feature.
The authors derive monthly prices by rolling on the day prior to expiration the value of one contract into the next nearest such contract. This is vital because institutional investors never want to be in a situation where a futures contract expires and an obligation to take delivery of hundreds or thousands of tonnes of wheat materialises. Commodities are not their business and the ‘roll’ between contracts is an important, if debated, source of additional return.
Gorton and Rouwenhorst’s index value is the total of each different category return divided by the number of categories.
Their findings show that between 1959 and 2004, commodity futures’ annualised returns were almost as great as those for stocks and far higher than those for bonds. Commodity futures exhibited a standard deviation slightly below that of US equities. Moreover, commodities had a positive skew in contrast to equities’ negative skew; in other words, more months of positive than negative returns.
The authors conclude that in addition to offering high returns, the historical risk of an investment in commodity futures has been relatively low – especially if evaluated in terms of its contribution to a portfolio of stocks and bonds.
However, it should be noted that in effect Gorton and Rouwenhorst have created an index for commodity futures, albeit based on historical figures.
Their artifice is important because Gorton and Rouwenhorst’s index is not investable. No one is likely to have achieved returns exactly as they have calculated them, which makes the paper an excellent primer but no more.
So if Gorton and Rouwenhorst throw light on the subject but theirs is not an investable index, where do interested investors turn for further guidance? The CRB Index itself is investable but just $1bn (€824m) of assets is estimated to be following it. Other index suppliers include Dow Jones/AIG, Goldman Sachs, Standard & Poor’s, Rogers and Deutsche Bank. Each index has its peculiarities – for example, we have already noted that the CRB is equally weighted in each category. GSCI determines weighting by world production, which results in a major skew to the energy sector. Dow Jones/AIG also determines by world production and futures liquidity. Deutsche only has six components – crude oil, heating oil, aluminium, gold, wheat and corn. Deutsche believes that these give reasonable representation of the entire commodities universe with enhanced liquidity.
In terms of volume, Goldman Sachs estimates that assets worth about $35bn track its index, versus $10bn for DJ/AIG-IC, $5bn each for Deutsche and Rogers, and $1bn for CRB.
So which should investors choose?
The ground is certainly moving as the number of indices proliferates, according to Benno Meier, former head of indexed commodity strategy at BGI. Standard & Poor’s and CRB have both recently refreshed their offering. Meier points out, however, that the newer entrants and smaller players in this market have to wrestle captive business away from the well-supported indices. DJ/AIG-IC for example, is the benchmark of choice for major pooled products from Pimco. With a supporter this size, the index looks favourable to other asset managers.
A paper from BGI explains that DJ/AIG does not permit less than 2% or more than 15% exposure to any commodity. No commodity plus its derivatives - eg crude oil, heating oil and unleaded gas - may constitute more than 25% of the index. Most significantly, DJ/AIG-IC does not allow one sector to weigh more than 33% of the total index.
Without these constraints, GSCI has a weighting to energy more than double the DJ/AIG-CI maximum and far more than the CRB, which is even further removed from weighting by world production.
In terms of risk and return, Meier believes that GSCI’s skew towards the energy sector makes it more volatile but with stronger diversification benefits than the DJ/AIG-CI. It is ultimately a matter for investors to decide how much volatility they want from their commodity investments. “It depends what the clients’ main drivers are,” says Meier. “Ideally, we would see more recommendations from consultants.”
But Simon Martin, head of research at Aon Consulting in the UK, reckons that pension funds might as well put money on the horses as commodities. And he is not the only consultant to express caution with commodity investing. Given such views, the choice of the index is irrelevant. Meier acknowledges that investors have a lot on their list and the index is not top of the agenda. He adds that there certainly is little in terms of cost to sway investors’ mind. There are so many novel attributes of commodities futures investing that the odd basis point here and there is not crucial.
But while some consultants may have made informed decisions not to recommend commodities, it is certainly true that more information is necessary for others.
Pension Fund Indicators is an esteemed and venerable UK publication from UBS Global Asset Management, and for successive years, as part of its discussion on their relative merits to investors, it has compared the spot returns from commodities and equities in graphic form. In this comparison, the former seem rather flat and consequently an unappealing investment.
However, Pension Fund Indicators does not offer a graph on commodity futures, even though these derivatives are the means used by institutional investors to gain exposure to diversification of commodities. Returns from futures have been higher than spot returns because the former include two additional elements. The first is the roll, which is necessary to avoid taking physical delivery of the underlying commodity, and the second is the collateral yield, which exists because most institutional investors cover their positions with cash.
Since 2000, the roll yield annualised has been 1.8%. The collateral yield has been under 3% as interest rates have slumped. Nevertheless, add these amounts onto spot prices and they are not insignificant.
But if pension funds are choosing futures, do they need the representation of an index? Why not play with individual commodities? Heather Shemilt, global head of commodity index marketing and head of business development for GSCI, believes that the index is the right way to start. “If an investor begins with a strategic allocation of say 5%, they may then underweight or overweight the index. From there, they may become more aggressive and typically alter sectors rather than individual commodity futures.” But buying an index is the first step.
*Facts and Fantasies about Commodity Futures, by K Geert Rouwenhorst and Gary B Gorton, Yale ICF working paper no. 04-20, June 2004