Why the "discredited" peak oil model is still the best guide to investing in oil, copper, water, and other commodities

02/07/2012 8:30 am EST

Focus: STOCKS

Jim Jubak

Founder and Editor, JubakPicks.com

Now that oil is a long way from the $145 per barrel peak it hit in July 2008 and nobody on Wall Street is predicting, as Goldman Sachs did in 2008, that oil is headed to $250 a barrel, we’re not hearing much about peak oil anymore.

The peak oil model, initially developed by oil geologist King Hubbert and which accurately predicted a peak in U.S. oil production between 1965 and 1970, says that the production from an oil field grows exponentially over time, then peaks, and finally declines. The model has been applied to individual oil fields, national oil industries, and global oil production. Back in 2008, the fiercest proponents of peak oil as a global model were predicting that the world would start running out of oil sometime around 2020.

Now that the world is awash in oil, the only people talking about peak oil are its opponents, who are dancing on what they depict as the grave of what they call a “theory” that was never worth the graph paper it was plotted on.

Well, I still think that the peak oil model is the most useful description of what we see happening in the oil industry today—even if West Texas Intermediate, the U.S. benchmark, closed at a twitch under $100 a barrel on Friday, February 3. (Brent crude, the European benchmark closed at $114.58.)

And, I’d go on to say that the peak oil model is the best way to understand what’s happening to the prices of other commodities, especially copper.  (Full disclosure: I predicted that oil would go to $180 a barrel shortly before it began its collapse from the $145 a barrel high in 2008. And full, full disclosure: The only one predicting $250 a barrel oil right now is Iran, which is threatening that prices will reach that level if developed economies impose tougher sanctions on the Iranian economy in an attempt to slow or stop that country’s development of a nuclear bomb.)

And I think it’s even useful for thinking about how to invest in commodities such zs iron ore that, currently, don’t fit the peak oil model at all.

Let me explain why I still find so much value in this “discredited” theory.

The most damage to the peak oil model resulted from the over enthusiasm of its friends during the commodities boom that topped out in 2008. A view that I’ve called “hard peak oil” held that Hubbert’s model had predicted that world oil reserves were about to go into decline, that oil production was about to plunge, and that the world was about to run out of oil.

Those were all extensions—unjustified in my view—of Hubbert’s model. Hubbert’s formulation talked only about production rates and wasn’t a prediction of the measured levels of global oil reserves.  Hubbert’s model used a relatively narrow definition of oil, not surprising in an era when the conventional oil production of Texas, California, and Louisiana dominated the U.S. industry.

When oil companies continued to find oil and global reserves and estimates of global reserves continued to climb, peak oil theory took a ding. When the global oil industry discovered huge unconventional sources of oil in the Canadian oil sands and the tight shale formations of first the United States, and then Argentina, and China, and Europe, reviving production in mature oil producing countries such as the United States, the theory looked loopy.

But to see how useful a peak oil model can be to an investor look at the latest quarterly results from the big international oil companies.

Let’s start with Royal Dutch Shell (RDS).

Production volumes at Shell fell 5% year over year in the fourth quarter. Full-year production was down 3% from 2010. Shell told shareholders that it would reverse that downward trend and increase production in the low single digits in 2012.

What interests me is how much money Shell will invest in its attempt to reverse declining production. Shell will increase its total capital investment to $32 billion to $33 billion in 2012 from $31.5 billion in 2011. The actual increase in the capital budget for oil exploration, development, and production will go to $24 billion in 2012 from $20 billon in 2011. That’s a 20% increase.

And what will Shell and its investors get for those bucks? If recent history is any guide, not as much as they used to get.  Shell’s return on average capital employed (which is one of the most accurate measures of the profitability of an oil company’s capital investments) in 2011 fell to 15.9%.  A few years ago, when oil prices were much lower, return on average capital employed checked in above 20%.

Shell has had troubled increasing production in recent years, but the drop in return on average capital employed is an oil industry wide problem. For example, Chevron (CVX), one of the international majors that has been most successful at adding reserves in recent years, showed a return on average capital employed 20% lower in 2011 than in 2008. ExxonMobil (XOM), which is historically more profitable than its peers among the international majors, averaged a return on average capital employed of more than 27% from 2006 through 2010. In 2010 the company’s return on average capital employed fell to what was still an industry-leading 22%.  (ExxonMobil’s big acquisition of XTO Energy in June 2010 makes it tough to compare figures for 2011 with previous years.)

These trends are just about what you’d expect from the peak oil model. As reservoirs mature, oil produced from them gets more expensive as companies have to invest more in methods to extract oil. As fields and national reserves, mature companies can continue to add new oil discoveries but the cost of each new discovery is likely to rise because the oil is coming from less hospitable geologies.

And I’d add this corollary to Hubbert’s original model: As oil prices rise, oil companies invest in unconventional oil reserves to increase their reserves, but producing oil from these unconventional sources—whether it’s from the oil sands of Alberta or the tight shales of Eagle Ford or the deep ocean pre-salt formations off Brazil—is more expensive than producing conventional oil. The world can certainly continue to expand its reserves of oil—but only by increasing its investment in exploration and development. (Another corollary might be that since unconventional sources don’t have nearly the production history of conventional sources, it is much harder to generate reliable estimates of reserve size, recoverability ratios, and reserve life span. The natural gas industry in the United States is coping with exactly these problems in development of gas from tight shale formations.)

A very clear investment strategy results from this peak oil view of the industry. You should invest in the companies that sell the stuff that oil companies are buying with those ever increasing capital budgets—that means companies like Schlumberger (SLB), National Oilwell Varco (NOV), and Ensco (ESV). And it means buying smaller oil companies with promising areas of exploration and some proven production. Buying and then expanding already established production is one way the majors can try to beat the rising cost of finding and producing oil. I’d look at Tullow Oil (TUWOY in New York and TLW.LN in London) and Talisman Energy (TLM) internationally, as well as U.S. producers such as Pioneer Natural Resources (PXD) and Oasis Petroleum (OAS.)

But there’s no reason you should limit your peak oil thinking to oil. Right now the copper industry is probably the best example of peak oil economics among global commodities. And with the huge $88 billion merger of Glencore (GLEN.LN in London) and Xstrata (XSRAY in New York and XTA in London) galvanizing the acquisitions in the mining sector in the same way that the 2001 combination of Billiton and BHP to form BHP Billiton (BHP) did, I think investors are looking at a scenario where just about every producer is potentially in play.

The supply side story for copper is what I’d call a typical peak-oil situation. The world isn’t running out of copper—if all the copper in the earth’s crust were recoverable, the resulting supply would be just about inexhaustible. But the world is running out of the best grades of copper ores in relatively easily accessible geographies ruled by reasonably stable governments. In 1906 (a period when production from the United States was on its way to dominating the global industry) the average copper ore graded 2.5% for copper. By 1935 the average ore was down to 1.89%. In 2009 copper ore in the United States graded at just 0.43% copper. The U.S. Geological Survey reported a current reserve base of 1.6 billion metric tons of copper in the United States as of 2005—but the survey considerable only 950 million tons of that economically recoverable.

One result of this situation is that copper producers are facing their own version of the oil industry’s declining return on average capital employed—only in the copper industry it seems to be embodied both in rising capital spending budgets and in more and more projects facing major delays in reaching full production. One result of that has been significant declines in 2011 copper production versus 2010 levels with production year-to-year falling 23% at Anglo American (AAUKY in New York or AAL.LN in London), 25% at Xstrata, 67% at BHP Billiton, and 73% at Freeport McMoRan Copper and Gold (FCX) as a result of a long strike at its Grasberg mine.

Because the copper industry is already so concentrated (and already so dominated by diversified miners), I’d look to relatively big (and diversified) companies for potential acquisition candidates. (Smaller producers aren’t going to move the needle for a company like BHP Billiton.) The name that keeps coming up over and over again is Anglo American.

The company is big—but at a market capitalization of $60 billion it’s not so big that BHP Billiton or a combined Glencore/Xstrata couldn’t do the deal. At the moment it’s relatively cheap at 6.4 times earnings. Besides copper, Anglo American produces coal and about 40% of the world’s platinum.

The peak oil model also suggests taking a look at silver (long-term) and water through desalination plays such as Singapore’s Keppel (KEPLY in New York and KEP.SP in Singapore.)

But a peak oil model doesn’t apply to every commodity. For example, I don’t think it tells investors much about potash fertilizer producers. But even in some of those instances it has useful things to tell investors. Right now, for example, looking at the relative ease with which new iron ore mines are coming on line, a peak oil model would suggest that this isn’t the best of times for buying shares in iron ore miners.

So no peak oil models don’t explain the trajectory of every commodity but then, hey, physics is still looking for a unified grand theory too.

Full disclosure: I don’t own shares of any of the companies mentioned in this post in my personal portfolio. The mutual fund I manage, Jubak Global Equity Fund http://jubakfund.com/ , may or may not now own positions in any stock mentioned in this post. The fund owned shares of Ensco, Freeport McMoRan, Keppel, National Oilwell Varco, Oasis Petroleum, Pioneer Natural Resources, Schlumberger, Talisman Energy, and Tullow Oil as of the end of September. For a full list of the stocks in the fund as of the end of September see the fund’s portfolio at http://jubakfund.com/about-the-fund/holdings/

Related Articles on STOCKS