The Wall Street Journal: Street Sleuth: Oil-Price Forecasts Seem to Miss Upward Trend

Wall Street Tends to Make

Wrong Calls in Rallies;

History Is No Assurance

By JUSTIN LAHART

Staff Reporter of THE WALL STREET JOURNAL

May 19, 2004; Page C3

Most energy watchers agree: The price of oil won't remain above $40. But if the recent history of oil prices is any sort of guide, the slippery slope may lead upward.

Wall Street's consistently wrong call on the price of a barrel of oil at any given time is perhaps most notable when the commodity is rallying. Recent lowball oil forecasts, in turn, help explain why Wall Street has produced low earnings forecasts for energy companies. The problem: Energy analysts by profession appear to assume a reversion to means when analyzing oil -- that is, they assume any big move in price soon will be followed by a move back to the historical norm.

According to estimate tracker Thomson First Call, analysts expect the price of West Texas Intermediate crude -- the benchmark oil -- will fall from the current $40.52 a barrel to average $28.91 in the fourth quarter of this year. In 2005, it's expected to trade at $27.74.

Oil futures tell a similar story. On the New York Mercantile Exchange, light, sweet crude for delivery next month finished trading yesterday at $40.54 a barrel. If one is willing to forgo delivery until June of next year, however, that barrel of oil can be had for $34.81. While such "backwardation" has come to be the norm in the oil market, the difference between near-future prices and prices a year out tends to be much narrower. The market is betting on a price drop.

But forecasts for a drop in oil prices have been going awry in recent years. At the start of 2002, analysts expected West Texas Intermediate crude would average $21.16 a barrel. It averaged $26.09 instead. In 2003 they reckoned on an average of $23.67. It clocked in at $31.07.

These low oil-price expectations caused Wall Street to seriously underestimate energy companies' earnings strength. In 2003, profits from energy outfits in the Standard & Poor's 500-stock index grew by 63%, more than double what the analysts expected.

It's not hard to find reasons for Wall Street's flubbed oil expectations. The impending war in Iraq drove prices in late 2002 and early 2003, and soon after Saddam Hussein's regime fell it became clear that the country's oil infrastructure was in sorry shape. Bringing Iraqi oil back online has been hampered by insurgency in the region. And unrest in Nigeria, Venezuela and, most recently, Saudi Arabia has further threatened supply.

If such exogenous factors are behind the rise in oil, it's natural to think, as most analysts do, that the price of crude will dip back down to the mid-$20 a barrel levels that prevailed over the past decade. But the analysts may be ignoring lasting shifts in oil fundamentals.

"Mean reversion is not an applicable model," says Tom Petrie, head of Petrie Parkman, a Denver investment bank specializing in energy. "That doesn't mean oil prices can't come down, but they're going to settle materially higher than what most people think."

Mr. Petrie sees two factors pushing oil prices higher. First, with the Chinese and Indian economies growing rapidly, global energy demand has taken a step higher. More important, a maturing resource base means that production will not be able to easily rise to meet that fresh demand. In fact, says Mr. Petrie, world-wide oil production looks set to peak either in this decade or the next.

The idea of peaking was most notably put forth by M. King Hubbert, a Shell Oil geologist who predicted in 1956 that oil production in the U.S., excluding Alaska, would peak in the early 1970s. Mr. Hubbert's thesis was that once half the oil in an oil field was depleted, it would become increasingly difficult to get the other half out, and production would decline. It's akin, says Ken Deffeyes, a former Princeton geology professor who worked with Mr. Hubbert at Shell, to a fishing hole getting half fished out.

As U.S. oil production continued to rise, Hubbert's Peak, as the theory was called, was widely ridiculed. But Mr. Hubbert was right: Oil production in the Lower 48 began to slip in the early 1970s, and even with advances in oil-field technology it couldn't bounce back.

"I spent the first 12 years of my career thinking Hubbert was wrong," says Mr. Petrie, "and I spent the next 20 realizing how right he was."

According to Matthew Simmons, a Houston oil analyst and energy investment banker who heads Simmons International, Mr. Hubbert's thesis also applied in the Soviet Union, which hit peak production levels in 1987, and in the North Sea, which peaked around 2000. Mr. Simmons caused a stir earlier this year when he argued that production in Saudi Arabia, the world's leader in oil production, was on the verge of peaking. Saudi officials dismissed Mr. Simmons's argument as wrongheaded.

If Mr. Simmons turn out to be right, it could spell bad news for SUV-driving consumers and energy-dependent companies. But energy-company shares, whose low price-to-earnings multiples suggest investors expect oil prices to slip, could rise.

Mr. Deffeyes, the geologist, believes world oil production is set to peak on Thanksgiving Day 2005, give or take a few weeks -- an estimate whose false precision is meant, he says, mainly "to annoy the economists."

But he thinks the idea of peaking is worth taking seriously. In 1962, Mr. Deffeyes saw that by the time he was set to retire from Shell, U.S. oil production would be down to half of its peak levels. "And I didn't want to live in Libya or Iran," he says. "I'm an American."

Mr. Deffeyes became an academic.

Write to Justin Lahart at justin.lahart@wsj.com