Sunday, December 06, 2009

Reserves are bunk

Henry Ford is famous for having once said, "History is more or less bunk." He was, in fact, attacking tradition in an age of rapid technological and social change. Almost a century later we have a less ambitious observation which may not achieve the broad visceral appeal of Ford's statement, but one which may turn out to have a good deal of importance, to wit: Oil and natural gas reserve numbers are more or less bunk.

Let me introduce you to B. J. Doyle, vice president of operations for a small Houston-based oil and natural gas exploration company. Doyle's views on the oil and gas business have been on display for more than a year now at The Oil Drum, a site famous for its technical prowess and breadth of coverage when it comes to energy-related issues. On the site Doyle goes by the moniker Rockman, and through his frequent comments he has been trying to educate readers about the realities of the oil and gas business.

Now, he didn't actually say that oil and natural gas reserve numbers are more or less bunk. Nevertheless, that is a fair summary of what he told me when I spoke with him recently. To understand why an insider would cast aspersions on this sacred metric of the oil and gas industry, you need to know two things. First, Doyle doesn't have to please shareholders. The company he works for is privately held. Second, reserve numbers are meaningless unless they are indexed to a price.

Doyle began his explanation with a seemingly astounding statement: "One of the things we're least interested in is the amount of oil and gas that we are going to produce." How can this possibly be true? It turns out that the oil and gas industry uses a method common to nearly every modern business enterprise to evaluate its investments, namely, net present value analysis or NPV.

The concept is actually simple. If you have the choice of receiving $1,000 now or $1,000 three years from now, naturally you'd take the $1,000 today. That's because of what is called the time value of money. If you can invest the $1,000 today, say, in a bank CD, you can at least earn some interest in the next three years. Also, if you were foolish enough to wait for your money, inflation might undermine the purchasing power of that $1,000. The inflation calculator at the U. S. Bureau of Labor Statistics shows that it would take $1,072 in 2009 to equal the purchasing power of $1,000 received in 2006.

Every business knows that there are several ways in which it can invest its capital. So, business owners take the amount of the initial investment in, say, a new factory or a new oil well, and subtract that amount from the present value of what they forecast will be the future cash flows from that investment. If the amount is positive, then the project will be profitable and should be considered. If the amount is negative, the project should be abandoned. Of course, there are many factors when considering an investment, but a project that appears to be unprofitable will certainly not be considered.

Net present value analysis, however, doesn't describe the real world perfectly. This flows from the obvious truth that no one can actually know the future. One has to estimate the expected future cash flows. This is no easy feat when dealing with the uncertainties of yet-to-be drilled underground reservoirs, the challenges of operating producing wells, and the vagaries of the oil and natural gas markets. Then, one needs to apply a so-called discount rate. This process assumes that future cash flows received years down the road must be "discounted" to reflect the time value of money as described above. Doyle explains that discount rates applied in the oil and gas industry often range from 10 to 15 percent per year. He admits it's an arbitrary number, but it's arbitrary in every industry except perhaps as it reflects the presumed risks involved in the venture and the cost of capital (such as interest on loans).

When you work out what this implies for cash flow generated from a well several years after production begins, it becomes clear why the ultimate amount of oil and gas recovered from a well has little relevance to the decision to drill it. Let's do an example to see why. If you invest $3 million to drill a well (not an unusual amount these days) and expect to get cash flow of $1 million per year from the well for 10 years, on the face of it that sounds as if you are reaping more than three times your investment. But when you discount the cash flows appropriately, for example at 12 percent per year, you get an NPV of $5,650,223. That's $2,650,223 more than you are investing, so it's still a positive number even after discounting. And, it's 1.88 times the initial investment, a ratio that will become meaningful below. But the NPV of the $1,000,000 in annual cash flow in years 8, 9, 10 are as follows: $359,634; $316,478; and $278,500. (Remember: Each successive year's cash flow is discounted another 12 percent in this case until you get to the final year.) If the well keeps producing in year 20, the NPV of the cash flow in that year falls to just $77,562. If it is a very long-lived well, the NPV of the cash flow from year 40 is negligible, $6,015. All this serves to illustrate that the further any year's cash flow is from the present, the less valuable it will be to the company and therefore the less bearing it will have on a decision to drill a well.

As it turns out, few companies would even bother drilling such a prospect. Doyle says that right now his company won't even look at a prospect unless, based on seismic data and other information, it reasonably expects that the completed well will produce an NPV six or more times that of the initial investment. When there is keen competition for prospects, companies will drop their expectations down to three to four times the NPV.

This is where things get interesting. Doyle has seen some public companies drop their goal down to one. That's right. They will drill prospects that they believe have no reasonable chance of doing anything other than breaking even. Why will they do this? To boost stated reserves, a number by which Wall Street judges the value of oil and gas companies. They won't, however, make any true profit on these wells. But they will become what Wall Street calls an "asset play." They will be valued on their assets, in this case stated reserves, rather than on their profitability. This strategy has proven especially tempting to those engaged in the hunt for shale gas since drilling success rates are very high. This is a risky strategy, however, that leaves little margin for error. Prices lower than those forecast by such an analysis could quickly bankrupt a company that drills too many wells based on an assumed one-to-one ratio of investment to net present value.

The claims that the United States has 100 years of recoverable natural gas as a result of the newly accessible shale basins has no meaning without attaching a price to it, Doyle contends. The fact that major shale gas producers have trimmed their active drilling fleets to a fraction of what they were during the 2008 boom in natural gas prices proves that price is a critical factor in determining whether to drill. And, where there is no drilling, there are no additions to reserves. The natural gas market has shown itself to be highly volatile which has not surprisingly led to wide swings in natural gas drilling. The notion that somehow there will be a consistent accretion of natural gas reserves from year to year or that all discoveries from previous years will still be considered reserves in a low-price environment is pure bunkum.

The same logic applies to oil discoveries. But these days no one is claiming the United States has enough oil left to supply the entire country for 100 years. And, so hype about oil reserves is less of an issue.

The upshot is that expected cash flow determines what areas will be drilled, not the size of potential reserves. Most companies won't drill a prospect unless they believe they can get their money back within two to three years, Doyle says. If it takes four or five years, the prospect is not very attractive. Cash flow is king.

It turns out that the NPV of the first three years of cash flow from my hypothetical well mentioned above is $2,401,831, less than the initial investment. Most companies would or should pass on such a prospect, and it would therefore never become part of anyone's reserves, he explains. Part of the hype over shale gas has to do with the claim that the wells may be very long-lived, he adds. Even if that turns out to be true--not a certainty as of now--the low flow rates expected after the initial burst of production and the distant payoffs would actually work against any decision to drill such wells. No wells, no reserves.

Doyle says that given modern technology, oil and natural gas are easier to find than ever before. But he doesn't believe that in North America at least, there is that much more to find. He thinks that shale gas in North America my indeed prove to be plentiful. But it will not be both plentiful and cheap.

And, of course, if we succeed at expanding natural gas production to meet the needs of a new natural gas-powered vehicle fleet--an idea advocated by one of the leading producers of shale gas--and expand other current uses such as the generation of electricity, we can expect that natural gas prices will soar. That may provide the necessary incentive (i.e. cash flow) to extract the shale gas that lies below the American landscape. But it will also certainly mean that the 100 years of supply that has been so frequently touted in the media will rapidly shrink to perhaps 30 or 40, and that the peak in production will come much sooner.

A peak in natural gas production in, say, 20 years would not exactly be a useful talking point for those advocating the wholesale conversion of key parts of the U. S. economy to run on natural gas. Just as we would be finishing such a conversion, we could find ourselves on the downslope of the natural gas production curve and faced with the urgent need to adapt our costly and newly completed natural gas infrastructure to run on some other energy source.

19 comments:

sv koho said...

Very fine post Kurt. You did a great job explaining something very important. Clear and concise and a pedagogic tour de force which I have moved to my permanent file on what I feel are important economic articles. You didn't actually use the EROEI term in your article although the spirit of it flows through it. I think I follow the math involved but a few examples would have been appreciated. I have been skeptical of T Boone and his supportive minions on a conversion to CNG for the transportation industry. This post seals its fate in my opinion. I hope this post gets wide circulation and I will do my small part to ensure that it does.

Anonymous said...

Very interesting post, Kurt. It is important for a reader to understand your orientation to peak oil. Because much of your article could be a strong argument against peak oil, despite your obvious orientiation.

Economics makes all the difference. Not necessarily just oil or gas economics. The entire economic picture. That is what most people in the peak oil community fail to see, but you probably do understand.

john personna said...

I thought the most interesting progression was from "100 years of supply" down to "perhaps 30 or 40" and finally "in, say, 20 years."

Neat rhetorical progression, on nothing more than a fear of too-rapid conversion.

This despite the fact that there is not a glimmer of conversion in the domestic automobile fleet. (Buses, taxis, and trash trucks, but not enough to consume 80 years of supply, 100-20=80)

I like you Kurt, because in each post you almost get there. As Anonymous says, much of your article could be a strong argument against peak oil.

Kurt Cobb said...

Let me respond to comments so far:

I think SV Koho is right that the question of EROEI is lurking in the background of this piece though it is expressed in financial rather than energy terms.

I think Anonymous is on to something about the way people define peak oil. In its most literal and simplest meaning, it is the all-time highest level of production followed by an irreversible decline. Now some will say the reason is geology. And some will say it will be the result of shrinking demand as other fuels replace oil. But many will rightly recognize that peak oil production, in fact, peak energy production will be the result of many factors, one of which is financial.

In this case, I am pointing to a financial model and a set of incentives that have important implications for how much oil or natural gas is going to be drilled. But more important is the effect of that model on the rate at which drilling will take place since it the rate of production which is most important. Flows are far more important to the functioning of modern civilization than stocks.

John Personna should not take my example of a possible conversion of the North American vehicle fleet to natural gas as a prediction. It is merely an illustration showing that if you increase demand for a limited resource exponentially, you will deplete the resource much more quickly than any rate predicated on current production. Forecasts based on static rates of consumption of a resource such as natural gas are ludicrous, and they are especially ludicrous when they are coupled with a call to vastly increase that consumption.

For the record, I don't think any such conversion is going to take place. But if it does, I believe it will prove to be a disastrous policy mistake based on a naive reading of the audacious resource claims being made by the shale gas players.

Kurt Cobb said...

As I reread John Personna's comments I realize that he has a mistaken view of the meaning of peak production. He says that my scenario for peak natural gas production assumes that total supplies will run out after 20 years. Of course, that is not what peak means. In this scenario (which is not a prediction) the rate of production begins to decline after the peak even while large stocks of natural gas remain. As I said above, the rate of production is all important for society just as the rate of oxygen intake for a human is all important, not the total stock of oxygen in the atmosphere. If you have an infrastructure predicated on a certain minimum rate of production of a critical energy resource such as natural gas and the rate falls below that minimum and then keeps falling even gradually, you are in for terrible difficulties.

john personna said...

Perhaps we started with mixed measures. "100 years of supply" was a quote. As was "in, say 20."

I'm fairly confident in the fundamentals, and actually think your summary of them NPV & etc. was very good.

No, I'm stepping back and looking at the essay as an example of how 100 years can be whittled down, with coulds and mights.

As to the original 100 years claim, I think all we can really be is a little agnostic. We can neither prove it or disprove it. But the important thing is that if true, it seems to put this energy resource out lasting longer than useful prediction.

john personna said...

As I continued my morning reading I found something that might interest you all:

"Another answer, offered by Blanchard and Gali (2007) is that the vanishing correlation between oil prices and the business cycle is the result of important structural changes modifying the transmission mechanism of oil shocks of similar sources and magnitude. They point out that the effects of a given change in the price of oil, identified as exogenous with respect to the US business cycle, have decreased substantially over time and conclude that the post-1984 effects of the price of oil on either output or the price level were roughly one-third of those for the pre-1984 period."

http://www.voxeu.org/index.php?q=node/4341

Anonymous said...

sir-
I am amazed at the 12% discount rate you cite. Given negligible interest rates and ongoing deflation (not inflation) how can such a number be justified. Is this an artifact of prior years very different economic environment?

Kurt Cobb said...

The discount rate, as I said in my piece, is partly predicated on the riskiness of the venture. Certainly, oil and gas exploration counts as a highly risky business. That means that those lending capital to it for exploration purposes are going to demand much higher returns to compensate for that risk. The low discount rates implied by, say, the treasury bond yield really apply only to something that would be very low risk since treasury bond yields represent the risk-free rate of return.

Kurt Cobb said...

John Personna keeps shifting back and forth between stocks and flows and does not seem to recognize the important distinction between them. A claim of 100 years of supply is false on its face if that claim is based on the current rate of consumption and consumption is rising exponentially during the period cited. A good summary of the effects of exponential increases in growth is available in Albert Barlett's lecture entitled, "Arithmetic, Population and Energy" (available both on the web and in DVD form).

John is right that we cannot know how much natural gas will be available in North American over the next century. That is reason enough to be cautious and not commit ourselves to a strategy of profligate use of this resource that could end in disaster.

john personna said...

Right Kurt, I quote you, and I'm the one going back and forth.

Beautiful sentence:

"A claim of 100 years of supply is false on its face if that claim is based on the current rate of consumption and consumption is rising exponentially during the period cited."

You've just done an explanation of how costs matter, and I've provided a reference on how much they matter changes over time.

How should I interpret "rising exponentially" in that light? A peaker would take the exponential as an unalterable law of the future ... but you seem to recognize the levels of economic feedback.

Half Empty said...

Anonymous (Comment #2)- would you care to set out the "entire economic picture" argument against peak oil?

Let's look at this wider economic picture.

The coming decades are shaping up to be a collision between EROEI and a global infrastructure built using cheap fossil energy and still 90% dependent on continuing supplies of cheap energy for it to function.

Neoclassical economics is fine when it comes to showing how resources respond to price signals but doesn't bother to concern itself with whether those resources exist in the first place.

OPEC now feels it necessary to issue repeated reassurances that it can hold oil prices down to a certain level, in order to prevent further worsening of the global recession. The oil people don't seem to want higher prices. They fear they will kill economic activity and deprive their industry of investment to fund (increasingly risky) exploration and development work.

But if energy prices stay low, investors won't pony up for new drilling because they won't get their money back. If energy prices rise, the cheap energy economy crashes, so there won't be enough capital to invest anyway.

To me, this is the economic explanation for the "undulating plateau" in oil production. Unable to sustain increases in output but still able, thanks to recession and past discoveries, to postpone decline. Take a finite resource and a system that breaks under the strain of high energy costs and what have you got?

And, to John personna (Comment #7):

2007: Blanchard and Gali see correlation between oil prices and business cycle vanishing.

2008: Correlation jumps out of a cupboard and says "Boo!".

john personna said...

Don't worry guys. I'm going to lose interest very soon. You can say "boo!" but I'm confident that I can go off, live a life with entirely different day-to-day concerns.

Yes, PO is an uncertain risk.

My health and fitness are other uncertain risks. The safety of my retirement portfolio is an uncertain risk. The health and direction of the international economy provides several uncertain risks, as do guided and misguided responses to the same. There are a few terrorists out there. There are certainly bad drivers.

The difference between someone who stays in PO and someone who leaves is that one sort can use PO as their organizing principle for the uncertainty of the world. And, this being the internet age, they can find confirmation and community in that.

This really sets up the permanent relationship between PO and the MSM.

The MSM has wider, and more balanced concerns, PO is about picking one, and sticking with it.

john personna said...

BTW, if you think 2008 and the financial crisis was about energy, try taking off the PO goggles.

Wikipedia ranks oil as a lesser cause, it is true:

http://en.wikipedia.org/wiki/Late-2000s_recession

but you need to see items 2.1 through 2.5 to have the balanced (dare I say MSM) worldview.

Richard Gibson said...

Among the clearest explanations of this idea I have seen. I'll be referring people to this. THANKS!

Anonymous said...

sir-
the explanation is now clear. You have conflated a discount rate with a risk premium. the former discounts a stream of payments in the future back to the present. It always uses a rate based on the risk-free (is there such a thing anymore?) rate of return. The increment of income demanded in consequence of risk is a separate factor: a risk premium. The 10% "discount rate" you refer to is undoubtedly largely a risk premium.

Henry Warwick said...

What everyone is missing is the following:

We might talk of 20, 50, 100 years of natural gas production. This misses the point : SHOULD we be developing these resources, especially given their deleterious effect on the environment?

john personna said...

Ideally we'd be shifting from coal to NG.

Half Empty said...

Mr Persona.

I'm fully on board with all Wiki's likely causes of the 2008 financial debacle and the subsequent recession. Do you read John Michael Greer or Steve Keen? It's quite possible to retain PO as just one part of a balanced portfolio of views on what is driving events and risks.