Five years ago, John Tierney, a columnist with The New York Times, and Matt Simmons, peak oil guru and founder of energy investment bank Simmons & Co., made a bet. Simmons argued that oil prices would be much higher in 2010. Tierney, a believer in human ingenuity and a follower of economist Julian Simon, took the other position. Simon, a so-called Cornucopian, argued that there would always be abundant supplies of energy and other natural resources and that the real price of commodities like oil would remain stable or decline over time.
The Cornucopian thesis assumes that i) demand is largely stable, ii) substitutes are readily available, iii) demand is relatively small compared to the total resource, and iv) the cost of extraction remains largely constant, or more precisely, that productivity improvements in technology are at least able to offset diminishing returns in resource extraction.
Let’s start with the size of demand versus the resource. Global demand for crude oil runs about 30 billion barrels per year on 1.2 trillion barrels of proven reserves, which includes about 300 billion barrels of oil sands and tars and 200 billion barrels of questionable upward revisions by OPEC in the late 1980’s. But taken at face value, the world currently has about 40 years of proved reserves, and over time, actual reserves will likely prove greater. However, 40 years is not a particularly long stretch in human history; it implies that all oil resources will be materially consumed in the lives of people younger than 20 years old today. For any given year, demand is relatively small compared to reserves; however, when speaking in decades, demand is material indeed. Ten years sees the consumption of a quarter of the proved resource. Thus, this assumption of the Cornucopians is likely untrue: demand, measured in decades, is material to the size of the resource.
Further, total oil consumption to date is just about equal to remaining proved reserves. Thus, at least on paper, about half of the total endowment has been consumed. But does ‘halfway through’ imply ‘peak’? For the Cornucopians, a higher price, greater application of capital, and technological innovation will bring forth additional supply. And it does – to an extent. However, none of these appears to materially reverse long-term decline. For example, US crude oil field production has fallen since reaching its peak in 1970, regardless of technological progress and price.
Certainly, high prices can stimulate investment to stem decline for a while. This happened in the five years after the oil shock of 1979 and has been occurring for the last couple of years. US liquids production is actually up nearly one million b/d over the last three years, but half of this growth is either deepwater Gulf of Mexico or the oil shales of North Dakota. The rest comprises biofuels and natural gas liquids, neither traditionally considered oil per se. However, even if we include these, US liquids production remains a far cry from levels achieved in 1970. Thus, neither price nor technological progress has been shown to defeat depletion over the longer run, and neither the high prices following the second oil shock of 1979 nor the bust in oil prices after 1985 appears to have made much difference in long term crude oil production declines.
Nor has the production of oil become materially cheaper. From 2000 until 2008, upstream capital costs increased 130 percent, and although they fell modestly during the recent, brutal recession, they have again turned up.
CERA’s Upstream Capital Cost Index (2000 = 100)
Nor has capital become more productive. From 1995 to 2004, $180 billion of upstream capital expenditure would deliver 1 million b/d of incremental oil, condensate and natural gas liquids production. From 2005, when the oil supply stalled, through year-end 2010, producing the same volume required $1.07 trillion, a six-fold increase – even after allowing for the $100 billion the Saudis invested in their spare, and currently idle, capacity. Thus, increasing the oil supply now requires either six times as much money, or technology six times more productive, than it did a decade ago. This suggests that neither a modest increase in capital spending nor modest technological improvement will create additional supply. There is not a bigger oil resource available for just slightly more investment or marginal technology improvement. In the case of oil, ‘halfway through’ does actually imply peak oil. Put another way, Hubbert’s production curve is accompanied by Hubbert’s cost curve. The cost of extraction does increase, and it is likely to increase at an increasing pace.
Demand also matters, and here Tierney is on more solid ground. In a market economy, there are no shortages. The price of a commodity simply rises until demand falls away. Functioning markets are always “well supplied” by definition, even if the commodity is unaffordable for most people. Right now, the Chinese economic growth is driving oil demand. We estimate that China by the end of 2015 will put 120-150 million new cars on the road, an amount equal to the entire vehicle fleets of Germany, France and the UK combined. That kind of growth will keep constant pressure on the oil supply.
If prices rise too high, people in the slow-growing countries, for example the US, will consume less. But this should not be confused with abundance. Constraint of consumption in one area does not necessarily prevent economic growth or the progress of society, but it does mean that society is worse off by at least one tangible metric.
The process of adjusting to high oil prices is traumatic as well. In 2005 dollars, oil prices were $75 at year-end 2010, just modestly more than the $65 at the time of the Tierney-Simmons bet. But bridging those two dates was the second-worst recession in modern history. The two periods are not comparable. The US economy has historically shed oil consumption when expenditure on crude oil exceeds 4 percent of GDP, equal to $86 a barrel currently. This shedding is ordinarily accomplished through recessions, as it was in 2008. Thus, oil prices did fall to near 2005 levels, but did so by reducing US oil consumption by 10 percent per capita, and that in turn was achieved by increasing unemployment by five percentage points. The US achieved low oil prices by evicting seven million people from their jobs. Is this Cornucopian abundance? Can Tierney claim a moral victory if it is achieved on the backs of the legions of unemployed?
When Matt Simmons took the bet, he took it in Matt Simmons style: hyperbolically. Had he argued that oil would be more expensive in 2010, he would have been right. But he argued that oil would reach $200 a barrel in 2005 dollars. In terms of oil markets, that was a possibility. But if macroeconomics was taken into consideration, it was clearly a non-starter. Both Simmons, and some analysts, like Goldman Sachs’s Arjun Murti, missed the key point that the global economy cannot consume oil at any price; it will reduce economic activity instead. In the end, Matt Simmons may have been wrong about oil prices, but he was, to all appearances, right about oil-supply trends.
As we look forward, we can again consider some bets in Matt Simmons’s spirit. Here are a couple for John Tierney:
- US oil consumption – including biofuels and NGL’s, but excluding natural gas, CTL and GTL – will be at least 10 percent lower by 2018 than it is today.
- We’ll have another oil shock within three years.
Given the Cornucopians’ endless faith in cheap oil, it should be an easy wager.
Steven R. Kopits heads the New York office of Douglas-Westwood, energy business consultants. The firm assists energy service providers with market research, strategy development and commercial due diligence. The author, a former Cornucopian, is solely responsible for the opinions expressed – and wagers placed – herein.
(Note: Commentaries do not necessarily represent the ASPO-USA position.)