Three Tiers of Commodity Price Drivers
It is true that the steady general rise of commodity prices of latter years seems to have crossed an inflection point last March, along the week that followed the rescue of Bear Sterns and during which a massive closing of long positions in commodity markets took place. However, record oil prices since then, as well as soaring prices of some agriculture commodities, have reminded us of how treacherous any sweeping generalizations about commodity prices tend to be. Chart 1 shows how weighted-average index-based figures hide a wide variation of performances.
Chart 1 – Commodity Prices
Sources: Goldmann Sachs, “Commodities: focus on copper”, 04/16/2008 (above) & IIF (below)
In our view, it is worth decomposing and analyzing the evolution of commodity prices as the outcome of three sets of drivers, as these will affect differently each commodity and with different time spans. Financial market dynamics constitute a common denominator to commodities as an asset class and thus tend to move their prices in a correlated way. On the other hand, physical market factors turn their matrix of correlations a very heterogeneous one (wheat and rice are highly correlated in that domain, wheat and orange juice less so, and wheat and oil only to the extent that the latter affects costs of the former). Thirdly, given the weight of inventories and public policies in several cases, one may see “panic” or “collective-action” types of factors at place in the pricing of some commodities, as it has been recently the experience with grains. To make the task of forecasting even harder, the weights of the three tiers of drivers will vary over time.
Such an exercise of decomposition may lead to the conclusion that, even as an unwinding of net long positions in financial markets and a physical demand contraction both take place along the current global downturn, not necessarily all commodity prices will undergo a drastic downward adjustment. Furthermore, that decomposition may be of help in understanding and reacting to the current crisis of rising food prices.
The Financial Tier
Jeffrey Frankel has made the case for a financial determination of commodity prices historically: “real interest rates are an important determinant of real commodity prices”. Interest rates affect the demand for storable commodities, by altering incentives for extraction in the present rather than in the future, desires to carry inventories, and movements by speculators between spot commodity contracts and treasury bills. The financial determination includes Dornbusch’s types of over-shooting adjustments, as real factors on the physical side are sluggish in relative terms in their response to price changes, and thus it implies a proneness to occasional “bubbles”, particularly as the range of commodities fully-fledged as financial assets widens over time.
On a shorter time horizon, the roles played by real interest rates, liquidity and search for yield – i.e. movements by speculators – in the bubble-like evolution of commodity prices since last year can be noticed in the increase of non-commercial net long positions, as well as in the negative correlation of treasury-bill yields and the CRB index (Chart 2). The phenomenon of price correction last March was sparked by a widespread closing of positions that was triggered by cash needs elsewhere as a follow-up to the Bear Sterns dramatic weekend rescue. However, as the path of prices afterward has demonstrated, the financial tier is just one of the determinants, even if a powerful one.
Chart 2 – Commodities as a Financial Asset Class
In the near future, we are likely to see periodic commodity price spurts and bursts accompanying the evolution of money and financial markets as a whole, as the financial crisis unfolds. Anyway, either gradually or not, the tendency toward general financial deleveraging will extract steam out of the financially fed commodity price boom of the recent past. Nonetheless, no commodity-price long bust can be taken for granted.
The Physical Tier
The interplay of supply and demand at physical markets of commodities is intrinsically distinct from the one at their common financial sphere, as the fungibility of money and the speed of financial adjustments give place to gradual shifts at both supply and demand sides. Even when profit arbitrage and the equalization of rates of return succeed ultimately to operate as a center of gravity, the physical nature of goods implies long-run different cost-and-price trajectories, as production conditions and productivities vary geographically. It is on top of such structural trends that the inevitable lags in the response to rates of return at the supply side, coupled with the time required for substitution effects at the demand side, generate the kind of mid-term boom-and-bust cycles typical of commodities. Nonetheless, one should not lose sight of the price ball in the longer term, as the underlying physical conditions of production – duly modified by technological changes – will impose themselves in the formation of prices. 
Has the inflection point possibly crossed in financial markets last month coincided with the beginning of a bust phase of the commodity cycle in course since 2001? Has the response to rising prices already reached that stage in which investments have outpaced demand – or are about to as the global financial crisis transmits itself to the real side of the US and European economies?
First of all, the structural drivers of the rising demand for most commodities – approached by us here and here – still seem to be in place, as witnesses for instance the resurgence of freight rates over the last month. It is still to be seen the extent to which the future abatement of demand in G3 areas will substantially predominate over the resilience of demand from other regions where a domestically-driven growth seems to have been sparked.
Furthermore, on the supply side there is no widespread sign of excessive inventories or of idle capacity of production. On the contrary, any news on natural or political events keep on exerting impacts on prices of correspondingly affected commodities – as the story about “peak oil” in Russia did last week, as well as threats of life-imprisonment for rice hoarders in some countries or releases of studies concluding that crop yields have already been affected by global warming.
The Goldman Sachs team on commodities, led by Jeffrey Currie, has been since 2002 consistently hammering on the idea of a “revenge of the old economy” after the boom and bust of the IT-built “new economy”. After decades of low returns in the oil, gas, metal and mining industries, on top of which capital was for a while completely sucked by the “new economy”, all spare production capacity in the “old economy” came to be exhausted. The same applies in the case of agriculture, as protection-cum-subsidies in developed markets depressed production perspectives for decades in a large part of the developing world. As the global economy started a new growth cycle, one in which the natural-resource intensity per GDP unit was elevated by the weight of China, India and other emerging markets, an enduring upward phase of demand and rates of return in the “old economy” has been in place.
Besides initially hitting bottlenecks on access to both inputs (labor, steel etc.) and technology and facing steeply rising marginal costs, the old economy saw its “revenge” evolve into a “revenge of the old ‘political’ economy” in the case of oil, gas, metal and mining industries, as sovereign entities increased taxes and other forms of capturing gains from the suddenly valued natural resources. Policy constraints on the free flow of capital, labor and technology have hampered the functioning of the textbook mechanism of mid-term full supply adjustment and the supply growth has been constrained, regardless of prices and rates of return. “Capital flows not to the most efficient commodity investment but rather to the most freely accessible one that is usually inefficient, extremely high cost/tax with poor rates of return. This in turn puts upward pressure on prices or, in some cases, prevents capital flowing at all, creating physical shortages” (Goldman Sachs, “The revenge of the old ‘political’ economy”, March 14, 2008, p.7).
On the agriculture side, one may also notice a similar “political-economy” obstruction as the maintenance of protection-cum-subsidies in developed economies stifles incentives toward transferring production to regions outside where the scarcity of land and other natural resources is much lower. The insistence on self-sufficiency in the production of biofuels is just a chapter of that distortion; albeit one that has led to such an intensified over-use of local natural resources as to spark an explosion of world prices of grains (see here).
It is our view that not only we are far short from a stage of capacity glut in most commodity cases, but also that the financial deleveraging and the probable forthcoming physical demand slowdown will not likely lead to a general collapse of commodity prices in the near future.
The “Panic” (“Collective-action”) Tier
An explosive interaction between physical and financial determinants of commodity prices may also be compounded by idiosyncratic self-fulfilling bouts of speculative behavior and self-defeating defensive behavior by public and private agents, as it has happened in the last few weeks with rice and other staples. The run for stocking rice and restrictions on exports exacerbated the problem for those who resorted to those moves in an attempt to defend themselves from expected price hikes. Certainly if a concerted collective action had been in place, instead of panic, rice prices would not have behaved as they have (Chart 3).
Chart 3 – Rice Prices
Beyond emergency measures and actions to mitigate the poor people suffering, a lasting and efficient reversal of the current basic-food price shock will thus require:
a) a concerted collective action called for by major public and multilateral stakeholders in the global economy;
b) with a credible commitment to an agenda addressing those impediments to agriculture supply adjustments present at the physical tier of commodity price drivers.