Over the past decade, the “normal” price of oil increased from around $20 a barrel to above $80 a barrel. The price of oil increased at far greater than the inflation rate, and this signaled a fundamental change in the supply/demand equation for oil, which in turn signaled new opportunities for investors.

Some blamed increased speculation for the run-up in the price of oil. This was the premise behind Dan Dicker’s book Oil’s Endless Bid. A former oil trader, Dicker argued that the Commodities Futures Modernization Act enabled investment banks, energy hedge funds, managed future funds, and ETFs to dominate the oil market and turn it into a global Ponzi scheme. He downplayed the impact of fundamentals in driving oil prices to much higher levels.

A number of books and articles argued that the real cause of high oil prices was inadequate oil supplies, and that global oil production was on the cusp of a permanent decline. This decline is commonly referred to as “peak oil,” and the term entered the mainstream from about 2005 when oil prices really began to escalate. Many people made alarming predictions about the imminent threat of peak oil to civilization. Higher oil prices and fuel shortages were among the milder predictions put forward.

But what exactly is peak oil? Is it a real possibility within the next few years? And if so, does it present a threat or an opportunity for investors?

Peak Oil Primer

The scientific study of peak oil began in the 1950s, when Shell geophysicist M. King Hubbert reported on the evolution of production rates in oil and gas fields. In a 1956 paper, Hubbert suggested that oil production in a particular region would approximate a bell curve, increasing exponentially during the early stages of production before eventually slowing, reaching a peak when approximately half of a field had been extracted, and then going into terminal production decline.

Hubbert applied his reasoning to oil production for the Lower 48 US states and offshore areas. He estimated that the ultimate potential reserve of the Lower 48 US states and offshore areas was 150 billion barrels of oil. Based on that reserve estimate, the 6.6 million barrels per day (bpd) extraction rate in 1955, and the fact that 52.5 billion barrels of oil had been cumulatively produced in the US already, Hubbert estimated that oil production in the US would reach maximum production in 1965.

Hubbert further calculated that if the US oil reserve grew to 200 billion barrels, peak production would occur in 1970, a delay of five years from his base case. Hubbert’s estimate of 1970 US oil production based on the 200 billion barrel reserve case was 3 billion barrels, or 8.2 million bpd. Oil production in the US did in fact peak in 1970, albeit at 9.6 million bpd.

Peak oil advocates point to Hubbert’s 1970 prediction of a US peak as proof that his method has merit. However, most don’t realize that his prediction was based on a secondary case in which he assumed that US oil reserve grew to 200 billion barrels, and that Hubbert indicated skepticism that this was possible because it would require the US to find additional oil fields equivalent to “eight East Texas oil fields.” So Hubbert is widely credited with precisely calling the US peak in 1970 despite the fact that he was skeptical that the peak would take place that late.

Through 2011, cumulative US production stands at 205 billion barrels, with a remaining estimated reserve of 23 billion barrels. This is over 1 billion barrels higher than the proved reserves in 2000. Even correcting for Alaska’s cumulative 15 billion barrels of production and its remaining 3.5 billion barrel reserve that Hubbert didn’t consider, the total reserve of the Lower 48 was ultimately greater than Hubbert’s high estimate case (200 billion barrels) of Lower 48 oil reserves.

Hubbert’s critics argue that while his technique may have some utility, the methodology is simplistic and does not account for reserve growth, unconventional oil production, curtailed production (as we often see in OPEC countries) or geopolitical factors.

For example, in 1980 US proved reserves were 36.5 billion barrels. Thirty years later, in 2011, US proved reserves had only fallen to 30.9 billion barrels. Over the course of three decades, the US produced 103 billion barrels of oil, but US reserves only fell by 5.6 billion barrels. Those “eight East Texas oil fields” that Hubbert was skeptical of finding were actually found within fields that were already known.

These increases in US production while barely drawing down reserves are a result of continued improvements in oil extraction technology (e.g., hydraulic fracturing), improving recovery factors, and from new discoveries (particularly offshore). But it demonstrates that one must allow for reserves growth when looking at any country’s proved reserves. I think it’s a safe bet that even though US proved reserves are 30.9 billion barrels, a lot more remaining oil than that will ultimately be produced in the US.

Hubbert did address the issue of improved extraction techniques but argued that rather than significantly impacting the date of peak oil in the US, better extraction techniques would slow the rate of production decline—and thus skew the back half of the bell curve. And in fact, the average annual rate of decline in US oil production since the 1970 peak has been under 1 percent—much lower than the natural decline rate of most aging oil fields.

Hubbert also used his technique to estimate the global oil production peak. But as in the case of US reserves, Hubbert underestimated global oil reserves. Hubbert’s model was based on a presumed global reserve of 1.25 trillion barrels of oil, but the ultimate size of the reserve would grow far beyond Hubbert’s estimate. Cumulative global oil production through 2011 is just over 1 trillion barrels of oil, and remaining proved reserves are estimated to be over 1.6 trillion barrels.

Based on this underestimate of global oil reserves, Hubbert estimated that a global peak in oil production would occur around the year 2000 at an annual production rate of 34 million bpd. While it does not appear that global oil production has peaked, it is clear that it did not peak in 2000. The rate of global oil production (crude plus lease condensate) by 2011 was 74.1 million bpd—which represents a slight increase from the 73.6 million bpd rate achieved in 2005—and which many peak oilers confidently predicted was the ultimate global peak year.

Hubbert’s defenders point to his US oil peak prediction as proof of the utility of his methodology, and his critics point to the missed global peak prediction and gross underestimate of global oil production in 2000 as evidence of the shortcomings of his model. In fact, Hubbert’s method is not designed to predict reserve growth; a specific reserve size must be assumed. If the reserve size is accurately estimated, Hubbert’s model would be expected to give a fairly accurate estimate of an oil production peak. Thus, in the US, where the oil reserve estimate was fairly accurate, Hubbert’s model made a fairly accurate prediction on timing of the peak.

Peak Lite

While some insisted that oil prices were being driven higher by peak oil, others focused on the supply side of the equation. This argument—which was the position that I have taken—is that the rapid escalation of oil demand in developing countries stripped away most of the excess oil production capacity over the past decade.

Demand growth in China and India is a frequent topic in the news, but double-digit demand growth took place in every developing region of the world over the past decade. The net impact was that even though demand in the US and the EU fell between 2000 and 2010, global oil demand increased by nearly 11 million bpd over that time period. Because the oil markets are global, there was no relief from oil prices for consumers in the US or the EU.

Global production increased to meet the increased demand, but at a much higher price. This implies that the cost of production increased, and/or the spare production capacity has been lost.

I posed this question recently in an interview of former Shell President John Hofmeister, and he agreed: “All in all, when you do all the puts and takes, what most of the world does not take time to recognize is that the decline rate of existing fields, which equals 4 to 5 million bpd, means every year just to stay even we need 4 to 5 million new barrels per day of production every year. And if demand grows beyond that we need even more. That’s been the problem. We have not been able to keep up with demand growth and the decline rate simultaneously.”

This presents a scenario with effects similar to peak oil, but with an important difference. While peak oil focuses on a global decline in oil production, the end effect is that there would not be enough supply to meet demand at current prices. But the focus on the decline in oil production was misplaced, and as oil production crept ahead over the past few years, many ridiculed the predictions of prominent peak oilers.

Yet the price of oil remains much higher than it was a decade ago, so the peak oil camp did get that part right. What they got wrong was that some of the effects predicted by peak oil could still take place even as oil supplies expanded. I coined the term “Peak Lite” to describe this effect.

It is “peak,” because the symptoms will mostly manifest themselves as those of a true production peak: not enough supply to meet demand at current prices. In fact, we have already passed the point at which there is enough $25/bbl oil supply to meet everyone’s desires. But production can still grow in this scenario, which is why it is “lite.”

Because oil production grew over the past decade, many failed to recognize the significance that it was not growing fast enough to maintain the excess capacity cushion. Thus, many prognosticators failed to forecast the decade-long bull market in oil. The price predictions from peak oilers were actually more accurate than those of prominent industry experts like Daniel Yergin and of organizations like the Energy Information Administration–as each made numerous predictions that oil prices would remain low over the past decade.

Besides the actual timing of the peak, another factor many in the peak oil camp missed is the extent to which oil production is a function of oil prices.

As oil prices rise, oil resources become oil reserves and oil production increases. Thus, while conventional oil that is economical to produce at $25 a barrel may be past peak, unconventional oil production grew rapidly as oil prices headed to $100 a barrel. These increases in unconventional oil production have been large enough to allow global oil production to continue to expand through present day.

This expansion in oil production is likely to continue for at least a few more years. There is a small possibility of oversupply in the short term, and some weakness in the price of oil. However, this situation is self-correcting, as lower oil prices will result in some unconventional production being shut in.

The end result is that even though there will be some periods of weakness in the price of oil in the years ahead, prices are unlikely to return to historically low levels. This is good news for most oil companies, although those with higher exposure to unconventional oil production will likely underperform during periods of softening oil prices.

Which Companies? 

Which companies are likely to perform well in the world of peak oil or peak lite? Those companies that have access to significant oil reserves. While the costs to produce do generally go up as oil prices rise, profit margins also rise. Large, integrated oil companies perform very well when oil prices are rising, but companies whose primary focus is oil production will be the biggest beneficiaries.

Our favorite portfolio holdings among companies that focus primarily on oil production include EOG Resources (NYSE: EOG), Oasis Petroleum (NYSE: OAS) and Occidental Petroleum Corp (NYSE: OXY). All three are well-run companies that are highly leveraged to rising oil prices thanks to their stakes in high-quality oil assets.

Buy EOG Resources below 125, Oasis Petroleum below 38 and Occidental Petroleum below 100.

Conclusions

I think as far as peak oil goes, most of us can agree that global oil production will inevitably decline. The points of contention are the timing, the steepness of the decline, the impact on the global economy, and the ability of other energy sources to fill the supply gap. Some believe there will be a smooth transition to alternatives, and some people believe peak oil will be catastrophic.

What do I believe? My idea of peak oil has always been one of supply struggling to keep up with demand. If this is correct, then oil prices are unlikely to ever return to the lows of a decade ago except for perhaps brief periods of time. There is a greater chance the oil prices will spike upwards, as we have seen over the past few years. I think that we will probably increase global production for a few more years, and that unconventional oil supplies will play an ever more important role in limiting the impacts of peak oil.

I believe that the peak lite scenario has been validated, although it is possible that there will still be times in which supplies do grow faster than demand. I expect a slow squeeze on western economies as developing countries continue to raise their standards of living – keeping fairly constant upward pressure on oil prices. This scenario presents a risk for many sectors, but I continue to believe that oil companies will be in a good position to profit from the tight balance between oil supplies and demand for the foreseeable future.