Oil: Are We In A Bubble?

Asks Megan McArdle. The short answer to this is that we don’t know.

Let’s start with some facts. A barrel of light, sweet crude oil today costs about $130 – $135, which is up 80% from the beginning of the year and is about 5 times the price in 2001. On an inflation-adjusted basis, this is higher than the previous modern peak, achieved in 1980, which would have been a little over $100 in today’s money. Because we’ve gotten a lot wealthier in the last 25 years, however, the percentage of GDP spent on oil and the numbers of days work required to buy a barrel of oil today are less than they were in 1980.

We don’t know if we’re in a bubble because we’re not sure what drove this increase.

It’s important to start by disaggregating the effects of the depreciating general value of the dollar versus the appreciation of the cost of oil. According to the Dallas Fed, as of early this year, the price of a barrel of oil would have been about 20% lower had the dollar held its 2001 value.

So, general inflation has been significant, but most of the rise has been created by other factors. What are they? A fantastic academic paper from a couple of months ago thoroughly reviews several non-mutually-exclusive candidates:

1. Supply and demand is usually the right place to start, and it is surely central here. As everybody knows, global demand is growing rapidly. While the US + Europe account for almost half of all worldwide oil consumption, about 80% of the growth in consumption over the past several years has been created by other countries. China alone accounted for about one-third of all consumption growth between 2003 and 2006. At the same time, production has remained fairly flat over this period.

Barring some major change in the world’s political and economic situation, demand will continue to grow. If China, for example, continues to grow its consumption at current rates, it will consume twice as much oil as the U.S. by 2030. This is entirely plausible, as it would still only be using about as much oil per capita as Mexico does today. Basically, as the world moves a billion or more people through the development pipeline, it will create large increases in the demand for oil.

Can we meet this demand? Smart experts predicted a few years ago that with oil at $60 per barrel, we would see large increase in production capacity as investments were made in new and existing fields to respond to this. CERA President and “The Prize” author Daniel Yergen predicted in July, 2005 that we could easily see 16 MMBPD (million barrels per day) of incremental capacity come on line by 2010. So far, we’ve only gotten about 1.5 MMBPD, and almost none of that in the last couple of years. When you look field-by-field, there appear to be explainable technical reasons for this, and as I’ve written in prior posts, expert bodies remain confident that we will be able to grow supply substantially at least for the next several decades. The recent failure to achieve production increases over, however, has naturally caused worry.

2. Somewhat counter-intuitively, this fundamental supply-and-demand driven price increase is exacerbated in the U.S. because we are so much more efficient in our use of oil than we were at the time of the prior oil shock in the 1970s. Europe and Japan are, in general, even more efficient than the U.S. This is why the already low short-run elasticity of demand, as far as it can be crudely measured through these natural experiments, is lower today than it was 20 – 30 years ago. In effect, we have already wrung out the easier-to-get efficiencies in the economy, so it takes yet more price movement to cause the developed world to use one less barrel of oil today.

3. If crude oil were a manufactured entity like cars, rather than a mined entity like gold, then we could simply estimate the supply curve (i.e., the lowest cost field will produce X MMBPD at such-and-such cost per barrel, the second lowest cost field will produce Y MMBPD and such-and-such cost, and so on all the way out to the world’s most expensive field), and make the tolerable estimate that price at any given time would be set by the marginal cash cost of the highest-cost field needed to meet demand. In fact, if you think there is a practically-infinite amount of oil and a lot of competition among oil producers, then this is pretty close to how prices get set. The era of cheap oil wasn’t too terribly far from this, adjusted for the many weird facets of the oil business. Once producers begin to become concerned, however, that there is a foreseeable and relevant timeline after which we would begin to approach a drawdown of a fixed supply of irreplaceable oil in the ground, then prices will begin to move above the marginal cost of the marginal producer. This was described by the economist Harold Hotelling in the 1930s, when he labeled it “scarcity rent”. The combination of huge price spikes, obvious long-term growth in demand, and lack of success in achieving the production gains we expected over the past 2 or 3 years have likely combined to lead some market participants to believe that we have moved to position in which scarcity rents have become a significant component of prices. Recent statements from major Middle Eastern producers indicate that they are starting to think this, or at least consider it to be in their strategic interest to signal that they believe it. At a minimum, any producer handicapping the odds almost certainly would put higher odds on this today than five years ago.

4. All of these factors combine to create s a fertile market for speculation. Almost all bubbles start with valid fundamental drivers of a price increase. They are abetted by uncertainty: many tried-and-true valuation tools start to fail, and many theories of valuation begin to compete. You start to hear talk of ever-increasing prices. This is the fire; then you need to throw the gasoline of lots of dumb money onto it to get a real bubble going. The amount of money dedicated to commodity index trading strategies is estimated to have increased from $13 billion at the close of 2003 to about $260 billion by this past March. In theory, money should be equally-likely to go on the long (“I bet prices will go up”) as on the short (“I bet prices will go down”) sides of this speculative market, but for a lot of practical reasons it is much more likely to go to the long side. This is the most plausible argument for the mechanism by which a bubble might currently be operating, and if history is any guide, there is probably at least some of this. What nobody knows is whether this is worth $1, $10 or $100 per barrel (though it seems very unlikely to be at either of these end-points).

What’s the bottom-line? We don’t know the relative importance of each of these factors, or how they interact. It is true and important, however, that demand is growing and production capacity is not. Demand growth is highly likely to continue for some time. The big question is therefore how fast the world can ramp up production.

The market is, as always, making bets in the face of uncertainty. When and if this uncertainty is resolved, the price of oil will respond. The most obvious positive solution would be for global oil companies to achieve current development plans, and begin to grow production capacity again at a high rate. All fundamental analyses by expert bodies that I have seen support the view that this is, for at least several decades, achievable through normal engineering. What is not as clear is the cost basis for this new production that would be central to establishing the new price floor, though it is very likely to be higher than the oil price that the world experienced prior to the big run-up of the past several years.

This engineering solution would not only have the direct effect of changing the short-term balance of supply and demand, but would likely reset the “scarcity rent” and speculative psychology to the extent that it is a significant driver of current prices. Independently, U.S. monetary and other policies that strengthen the dollar would be helpful in reducing dollar-denominated oil prices. These would also be easier to execute in a world of improved oil supply, as we saw in the virtuous cycle of the 1980s and 1990s.

A more speculative way (at least in the short-term) to go about doing this is to find oil substitutes. This effectively de-links the demand for energy from the demand for the specific energy source of oil. More likely, of course, it would partially de-link them. Without going into this huge topic in a lot of detail, it is inevitable that this will happen in the long run, and, while it certainly will create unanticipated problems while solving currently-known problems, it will be a great thing.

If this happens, would this be popping a “bubble”? I guess that depends on the definition of the term. I think it’s more accurate to say that right now a lot of uncertainty is priced into a barrel of oil. If it turns out that the experts are wrong about our ability to grow supply, we could as easily then see the era of cheap oil as a kind of bubble in reverse.