LONDON: Despite widespread declines in commodity prices in the first half of May, which pushed most commodity indices and hedge funds into the red for the year, most analysts and investors are optimistic the setback is temporary, and prices remain on an uptrend in the medium term.
But what if the consensus is wrong? What if commentators are overly influenced by the memory of the massive run up in commodity prices which ended in 2008, ignoring longer trends showing supply adapts to huge increases in demand in the medium and long-run, and technology can transform the cost structure of production?
In predicting the future, most forecasters are over-influenced by the recent past. It is a form of the “anchoring” bias explored by the pioneers of behavioral psychology and economics Amos Tversky and Daniel Kahneman.
Anchoring is one reason investors under-estimate volatility following a long period of calm (a phenomenon noted by economist Hyman Minsky), and why businesses and consumers are more likely to react to a rise in the price of oil that sets a new peak rather than one that merely reverses a recent decline (noted by California University Professor James Hamilton in his theory about oil shocks).
The expectation that the future will resemble the recent past is the main reason most commodity analysts missed the start of the huge rise in oil and other commodity prices between 2002 and 2008.
But cognitive bias operates in both directions. Haunted by under-estimates for oil prices last decade, and recalling the vertiginous rise in prices during 2007 and 2008, many analysts now confidently predict the era of cheap oil is over. But there is no reason to think this forecast will be any more accurate than their earlier projections.
In all cases, the bullish outlook for energy, food and metal prices rests on some combination of rising demand from developing nations, restrictions on supply, and escalating production costs. For most analysts and investors the commodity “super-cycle” is not over, just pausing, before prices resume their upward march.
Each of the arguments for rising commodity prices in future (demand, supply, costs) turns out to be based on a fallacy and a limited reading of past experience.
The demand argument is the most easily disposed of. Following a long tradition of scarcity theorists, beginning with Thomas Malthus in his “Essay on the Principle of Population” (1798) and economist William Stanley Jevons’ “The Coal Question” (1866) through to the Club of Rome’s “Limits to Growth” (1972), commodity bulls insist growing consumption will result in shortages of food, energy and metals and push prices higher.
But there is not a shred of evidence to support this hypothesis. Global consumption of energy, cereals, meat and steel has risen by several orders of magnitude since the eighteenth and nineteenth centuries, but their real price (measured in hours of work) has fallen sharply for most of the world’s population.
Malthus heirs’ have been warning for decades resources are running out, so far without success. The burden is therefore on the scarcity theorists to show why this time is different. So far there have been no compelling answers.
The other side of the Malthusian argument is that supply will struggle to keep up with growing demand. But again there is no evidence this is (generally) true. In many cases, supply is fixed in the short term but increasingly responsive to prices in the medium and long-term, as capital and labor are free to move into and out from the industry.
The definition of short, medium and long term vary depending on the characteristics of each industry (particularly capital intensity and competitive structure).
For agriculture, output is fixed by growing seasons, so the short term lasts 1-3 years, after planting and supply are more flexible. For capital-intensive industries like oil, iron ore and coal, the short-run is much longer, and the lead times for developing new production are far greater. But even for capital-intensive businesses, production starts to become more flexible after 3-5 years as a result of de-bottlenecking, and fully flexible after 5-10 years as a result of green field expansions.
The commodity super-cycle is now at least eight years old and maybe ten, depending on when the start of the cycle is dated. Too many analysts assume the oil and other industries remain stuck in the semi-fixed short run.
In fact, the oil business and other commodity producers are now well into the medium term if not the long run. Much greater responsiveness to prices is evident, for example in the development of relatively high cost oils like Bakken in the US and Brazil’s offshore presalt.
More sophisticated Malthusian models do not assume the supply of raw materials will run out but instead will become harder to produce and be more expensive.
Jevons did not believe Britain’s coal reserves would be exhausted. But he did argue the easily mined coal would be used up, forcing producers to go deeper and mine lower quality coals, adding dramatically to the cost and price of fuel. Like modern Malthusians, Jevons did not believe the era of coal was over, only the era of cheap coal.
Unfortunately, Jevons failed to see that much of coal’s role, particularly as a transport fuel, would be taken over by petroleum. He also failed to account for the tremendous improvements in mining technology that would allow miners to delve deeper at lower costs than he ever imagined. Some of those improvements were disruptive technologies, however most were simply continuous process improvements.
Modern Malthusians and peak oilers adhere to the same argument. Oil supplies may not be running out but it is getting progressively harder and more expensive to wring oil from the ground. Fracking and deepwater production are more expensive than conventional onshore wells in Saudi Arabia or the continental US. The marginal cost of supply is therefore rising.
In addition to cash costs, the “social costs” of oil production are also rising, as key producers need higher prices to balance their budgets and meet growing demand for social spending.
For the bulls, these upward pressures have pushed the marginal cost of finding and producing oil to $100 per barrel or more, effectively putting a long-term floor under oil prices at this level, and could escalate further in future, driving the floor price even higher.
But the argument is weak. Finding and development costs are largely a product of technology. Fracking costs have already begun to fall as companies become more efficient in pressure pumping. Improvements in seismic surveying should gradually cut the costs of even deepwater exploration.
And the argument about the “social cost” of production puts the cart before the horse. Oil and commodity market conditions tend to drive political and budgetary realities for producing states, rather than the other way around.
As a result, it is far from clear that the marginal cost of oil is really $100 as many in the industry now proclaim. Tight oil supplies and heavy oil sands are available for less than this, as are substitutes based on (plentiful) gas and coal to liquids technologies, or the increased use of LNG as a transport fuel.
None of these is available immediately, but all could be before the end of the decade. More prosaically, cost improvements in exploration and fracking are already lowering costs significantly. It is a mistake to project recent cost trends into the far future (5 years or more forward).
The consensus, which sees costs and prices only moving one way, is almost certainly wrong, based on previous experience. Investors and the industry would be better served by asking why previous forecasts went awry, and how current ones could be sensitive to even small changes in the assumptions.
— John Kemp is a Reuters market analyst. The views expressed are his own.
Anchoring and commodity price forecasts
