Elliptical Research Contribution 2008.5:

Amid the carefully choreographed political conventions last month, there was a spontaneous chant that will be
familiar to anyone who watched the Republican festivities:

 “Drill, baby, drill!  Drill, baby, drill!  Drill, baby, drill!”

For a party firmly associated with politically unpopular “Big Oil”, having this chant rise up—most vociferously from
the ten-gallon hatted Texas and Oklahoma delegations—was not exactly what Republican leaders wanted in
prime time. John McCain’s ambivalence was palpable when this chant briefly interrupted him during his
acceptance speech.

Yet the chants of “Drill, Baby, Drill!” do have real populist appeal these days. It was not only Texas oilmen who
joined the chorus. Consumers are upset. Gas prices still hover near $4 a gallon. Crude oil, even accounting for
its recent swoon back into the sub-$100 zone, has tripled in price since President Bush took office less than
eight years ago. The idea that our energy crisis could be fixed by the simple measure of more drilling is tempting.

But will increased drilling really bring us cheaper oil? Will it bring us independence from foreign oil? Will it even
bring us much more oil at all?

Common sense says it certainly won’t hurt domestic production to drill more. But the truly important question is
this: will more drilling really substantially improve our domestic energy supply? Common sense alone cannot
answer this question rigorously. However, we can answer this critical question if we take a careful look at hard
numbers from the oil patch.

Drilling Activity and Oil Production - The Big Picture

First, we need the hard numbers. The Baker Hughes Corporation (BHI) is not only a leading manufacturer of drill
bits, the company also compiles comprehensive data on oil drilling for the entire industry. We’ll be looking at
their historical drilling data in this analysis.

There are two main classes of common oil rigs, when someone chants “Drill, baby, drill!” they are referring to
rotary drill rigs. These are the workhorses used to punch new holes in the ground. Rotary drill rigs are used both
to explore for new fields and also to enhance extraction from known fields. The other main type of rig, the
“workover rig”, is primarily used for recasing, data logging, fracturing wells and the like. Workover rigs are also
an essential part of oil production, but the focus of people on drilling is quite appropriate, because it is arguably
the critical “rate-limiting” step in the search for new oil.

So the first step in determining the relationship between drilling activity and oil is to look at the big picture. The
following graph shows United States drilling activity and United States oil production, going back as far as we
can (1949) with the publicly available Baker Hughes drilling data.

Why “Drill, Baby, Drill!” is Not a National Energy Policy

An Analysis of the
Relationship of Oil Drilling and Production

Timothy D. Kailing
Elliptical Research
September 2008
Looking at nearly 60 years of data in this plot, four important things are clear:

(1) People debate the likelihood of a global oil production peak heatedly but, from a domestic point of view, peak oil
is very old news:
U.S. oil production peaked in 1970—a peak that we’ve never since exceeded, despite the
subsequent 38 years of technology improvement, major events like the ramping up of North Slope oil production,
and the economic incentive of increasing oil prices (remember the average price of oil at the 1970 peak was about
10 dollars a barrel—and that’s in inflation adjusted dollars!)

(2) Over this large time scale,
there is remarkably little correlation at all between drilling activity and oil
For example, from 1955 to 1970 drilling activity in the U.S. decreased steadily, from over 2500 working
rotary rigs all the way down to 1000. Yet during this time oil production increased by over 40%. These were the
good old days when boys were boys, men were men, and the U.S. domestic oil industry was still a young industry—
with large and easily accessible fields just waiting to be drilled—rather than the mature, literally “over the hill”,
industry it is today.

(3) The best case one can make for the benefits of drilling are during the period from 1971 to 1981 when—spurred
by the oil embargo, the huge price increase of oil, and Jimmy Carter’s infamous cardigan—the number of active
rotary drill rigs more than
quadrupled. This herculean increase in drilling activity did manage to reverse the decline
in U.S. production for several years (after a five year delay) but, overall, production in this period of increased drilling
still declined: oil production at the secondary peak was still below the 1970 peak. In fact, to take the most optimistic
assessment of the benefits of drilling,
the 300% increase in drilling bought just a 10% increase in production from
the 1976 intermediate nadir to the secondary 1985 production peak.

(4) Recently, the benefits of increased drilling in expanding oil production are even more anemic. With our domestic
resources increasingly mature,
the recent increase of drilling—a three-fold increase from 1999 to 2008—has
not only brought no increase in production, it has not even stopped the continuing
decrease in domestic oil
over the last decade.

Looking Closer - Does More Drilling Activity Lead to Increased Oil Production?

It’s possible to analyze U.S. drilling and oil production more closely and quantitatively. We can do this by graphing
the relationship of the number of working drill rigs in a year, and the domestic oil production
per rig. Keep in mind
that this is something of an abstraction. Rotary rigs do not directly produce oil, there is generally a great deal of
work to be done under a workover rig (hence the name) before the oil comes to market but, as I said, drilling is
arguably the rate-limiting step to the production of oil and one would expect a tight relationship. Here is the graph:
This is quite an extraordinary graph—and a sobering one. Unlike the loose overall historical relationship between
domestic drilling and production, there is quite a tight relationship between drilling activity and oil production per rig.
However, the relationship is precisely the last one you would want if you hope that drilling will solve our nation’s
energy problems. The relationship shows that more active drill rigs translates quantitatively into less oil per rig.
Even worse, as the red data and fitted curve show, this law of diminishing returns has gotten both worse, and
statistically tighter, recently, including during the increase in drilling over the last decade.

The relationship is a power law, and the little negative exponent is the critical bit. The fact that it is negative, and that
its magnitude is greater than one, ends up being very important. This fact is easier to see on a logarithmic graph,
which “takes all the curves out” of a relationship like this, and linearizes it:
With everything straightened out this way, the relationship becomes very clear. The little exponent in the
untransformed data is now just the slope that you (hopefully) remember from high school math class. As you can
see, this slope is -1.16 for the data before 1990. For the most recent years the line is both lower and more steeply
negative, with the slope now nearly -1.30 and, furthermore, the relationship is statistically even more robust as of
late—the earlier R2 of 0.91 is good, but the R2 of 0.95 for the most recent data means that 95% of the variation is
explained by the relationship, which is very strong indeed.

This strong quantitative relationship between drilling and oil production has a sobering upshot. Since 1949 the
overall pattern has been this: the harder we look for oil and the more we drill, the harder it gets to find. And during
the last 17 years, with the most current technology and under the most recent economic conditions, this
relationship has only gotten worse. The fact that the recent (red) data are lower means that recently we are finding
less oil per working drill rig. The fact that the slope has become more steeply negative means that as we look
harder for oil in the United States, the oil production we actually get has become
even more paltry.

U.S. oil production is in late middle age. This is what the decline looks like quantitatively. And you see those last
four red triangles marching forlornly down the red line? Those are the last four years of data.

One Proviso – Time Delays From Drilling to Production

Some clever readers may have noticed that I have simplified things a bit by ignoring the time delay between drilling
and production in some of this analysis. Drilling proponents tend to underemphasize this delay, because they want
the public to think that current drilling will bring oil online very quickly. It turns out that empirically, the best-fitting time
delay between drilling and production is 4-6 years. This delay is fairly apparent just by looking again at the graph of
U.S. drilling and U.S. production, and inspecting it closely:
For example, as the up arrows indicate, there was a five-year lag from the drilling nadir in 1971 and the production
trough in 1976. And, as the down arrows indicate, there was a slightly shorter four-year lag between the enormous
spike in drilling that peaked in 1981 and the following anemic secondary peak of oil production in 1985.

The Pollyannas out there might conclude that the slight shortening of the two lags with time might be because
improved technology and increased efficiency have shortened the lag from exploration to extraction. I have heard
many oil lobbyists make this claim when they try to downplay the lag between drilling and production. However, if
you believe this, then the fact that
despite a three-fold increase in drilling from 1999 to 2008, our domestic
continues to decline, is just that much more sobering.

I have done same the same analyses accounting for the time delay. The basic picture is entirely the same. The
graphs tell the same story, with the slopes of the power law relationship negative and steeper than -1. (I’ll explain
the implications of these slopes in the next section.) So, even accounting for the time delay between drilling and
production, the conclusions are identical: the data do not support the hope that even a huge increase in drilling will
do anything but, at best, slightly stem the overall rate of
decline of our domestic oil production. Hope for a
substantial increase is just that,
hope, and it is hope blind that is blind to the data and to history.

[Anyone undaunted by the additional complication of accounting for the time delay can find my discussion of the
relationship of
future oil production and current drilling activity here.]

More Straws in Increasingly Smaller Cups

The overall quantitative picture for United States oil production today is one where the effect of increased drilling is
essentially like putting more straws into the same cup.

Actually, it’s
worse than that:  if it were only that bad the slopes in the graphs would be -1, and they are actually
steeper than that. This is a bit counterintuitive, because what’s actually happening is rather complicated. Some
drilling activity really is just putting another hole into the same old reserves. But drilling certainly does find new,
previously untapped, resources, too. The problem, as Ken Deffeyes has pointed out in his book
Hubbert’s Peak, is
that all our new technology, and the recent increase of drilling activity, is mostly going into smaller and smaller
discoveries. Our domestic oil supplies are pretty well picked over, and the “low-hanging fruit” remaining—the
shallowest, lightest, most-permeable, and largest reserves of domestic oil—are few and far between.

The fact that the slopes of all the power laws are more steeply negative than -1 can be illustrated by a short
parable. You’re on a first date in high school, and your date orders a chocolate malted right as you both slide into
your booth. When the malted arrives you find that you’re
really hungry and it looks really good. You’re tempted to just
pop another straw into her shake and suck as hard as you can. But, no, you don’t really know her that well and,
besides, you don’t want to look cheap. So when the waitress swings by you order a malted yourself. America’s a
land of plenty, after all: there are malteds enough for everyone, right? However, when it finally arrives, you find that
your malted is actually less than half the size of your date’s, it’s in this little bitty cup, and what’s more, it’s also too
thick somehow and it tastes really gritty and nasty.

That’s where the United States is in regard to oil: most of the untapped reserves we have left are smaller, deeper,
farther offshore, less permeable, increasingly sour, and generally more expensive to bring to market. And the more
we drill, the more this will be the case.

Diminishing Returns of Oil, But Not Of Money, in the Oil Patch

Upon a little reflection, anyone who did not entirely sleep through their introductory economics class should be able
to make sense of this. With a nonrenewable resource, of course one will reach a point of sharply diminishing

Does this mean the oil industry is finished? Not at all. There is good money to be made in drilling for a long time.
As prices go up, oil drillers and producers can make very good money on the diminishing resources. The scarcity
value can make up for the fundamental scarcity for a very long time. That is, of course, the economic reason why
drilling activity has recently increased despite the diminishing return of oil production: the price is higher.

For example, say you think ANWR is likely to be opened to oil companies, and you think significant oil reserves will
be found. Because even the largest likely finds there are unlikely to significantly lower world oil prices (even the
most optimistic daily production estimates are OPEC rounding error),
precisely for that reason, there is a lot of
money to be made. Likely beneficiaries of such an event include many of the majors, especially current North Slope
producers: names like Chevron (CVX), ConocoPhilips (COP), Shell (RDS.A), Eni (E), Veritas (CGV) and BP (BP).
Similarly, if you think previously untapped offshore waters are likely to be opened, offshore specialists like
Transocean (RIG) and Diamond Offshore (DO) are likely to benefit, long term. Also, because in an over-the-hill
region it takes that much more drilling to produce the same amount of oil, drillers and equipment makers are likely
to be busy. A company like Baker Hughes (BHI) will find it easy to increase its top line, and a basket like the Oil
Service Holders (OIH) is likely to be a good long-term investment. With $100-a-barrel oil, a little goes a long way,
when it comes to making money.

The parts of our economy most vulnerable to the domestic oil decline are not the oil companies themselves, but
are things like airlines, Hummers, and jet skis—the consumers, not the producers—which will be in trouble long
before the drilling industry winds down (like now, for instance).

But the hard facts do have a very important strategic implication for the United States. The facts clearly show that
people who believe that by simply ramping up drilling we will substantially improve our diminishing domestic oil
production, well, these people—to use an expression not uncommon in oil country—are very much pissing in the
wind; in fact, they are pissing directly into a
very strong geological headwind.

Given the diminishing returns on domestic oil (in energy terms, if not in terms of profits) I believe that developing
alternative energy is our nation’s only real option to significantly improve our domestic energy production and
strategic energy independence. For this reason, I think that baskets of alternative energy stocks, despite their
inherent risk and volatility, are likely to be good long-term bets. The best way for most individual investors to invest
in this young and volatile industry is through relatively diversified ETFs like the PowerShares WilderHill Clean
Energy ETF (PBW), Progressive Energy ETF (PUW), Cleantech ETF (PZD), or the Market Vectors Global Alternative
Energy ETF (GEX).

Strategic Implications of the Increased Drilling’s Inability to Solve our Energy Problems

The data show that the sentiments behind “Drill, Baby, Drill!” do not translate into a rational or effective national
energy policy. In fact, aggressively developing our nation’s late stage oil reserves is arguably
precisely against our
long-term national interest. The data are clear: ramping up domestic drilling will at best only slightly slow the rate of
decrease of our domestic oil production, but it will rapidly exhaust our remaining precious domestic oil reserves.
the faster we use up the little oil we have left, the quicker OPEC will be the only one at the table with any
chips left.
 Strategically, this is a loser’s strategy.

Increased drilling is not the basis of a reasonable strategic energy policy, it is simply the chant of special interests
who want to make money for themselves from our nation’s limited oil resources. I am sure that many drilling
proponents genuinely would like to believe that increased drilling will solve our energy dependency, but they are
simply wrong. The data are shockingly clear.

Being skeptical about our ability to drill our way out of energy dependency is often portrayed by oil lobbyists as an
inherently politically position—a position that only environmentalists with ulterior motives would take. But the data I
have analyzed are oil industry data and the conclusions are unequivocal. Though some of my quantitative
techniques are novel, I am certainly not alone in my general conclusions. Similar logic led some smart veterans of
the oil industry, men like T. Boone Pickens and Ken Deffeyes, to see the writing on the wall some time ago. And it is
a welcome development that, recently, more conservatives have come to understand the reality about oil. Not long
ago, an article by former CIA Director R. James Woolsey and the brilliant Anne Korin argued forcefully that we are
not in a position to drill our way out of dependence on foreign oil. They made this argument in the
National Review,
which is hardly an environmentalist hotbed.

So “Drill, Baby, Drill!” may have a nice ring to it, but Woolsey and Korin put it nicely in their
National Review piece:
“Speechwriters’ tropes shouldn’t be taken as serious policy proposals. Geology will not cooperate in any such

Disclosure: The author currently (September 2008) has long positions in PBW, XTO, XEC, COP, and SU in the
energy sector.
Timothy D. Kailing is the principal at
Elliptical Research.

He wears another hat as an
advocate for the benefits of early
literacy in children; in this effort he
authored the book:
Native Reading: How to Teach Your
Child to Read, Easily and Naturally,
Before the Age of Three.
Copyright © 2008 Timothy D. Kailing