Oil
Page 32
Chapter Fifteen
The Gamble
AFTER 30 YEARS in the business, Andy Hall was confident about his unique gamble. Since the 1973 crisis he had traded oil from troughs to peak prices and down again. He believed that 2003 was the beginning of another watershed.
The previous year, the EIA had estimated that the world enjoyed a healthy excess of production of about eight million barrels a day, confirming John Browne’s assurance, “Supply will always meet demand.” In 2003, the chief executives of the oil majors spoke confidently about oil continuing to trade between $25 and $30, the average price since March 1991. Hall disagreed. Prices, he was sure, were about to increase. The market suggested the opposite. Buoyed by the EIA estimate, “spot” prices on Nymex were higher than prices for oil to be delivered five years in the future. Convinced of ample supplies, investors could buy oil for delivery in 2008 at 15 percent less than the current prices. Technically, the market was in backwardation. Hall believed that the market was wrong. “It’s a layup,” he said. But before placing an enormous bet against the market, he immersed himself once again in the recent history of oil.
The Big Four majors, Hall believed, were losing control of supplies and prices. In 2001, Shell’s economists correctly forecast that oil prices would fall below $20 by the end of the year, and with herd instinct, BP’s and Exxon’s economists had echoed that prediction. But careful study of oil statistics showed that overall production was static, and in some places was actually falling. Unable to increase production in the North Sea, Alaska, the USA and other mature areas that were now beyond their peak, the Western industrialized countries had increased their dependence on OPEC countries and Russia. Silently accepting their decline, each of the Big Four chief executives preferred to buy oil supplies cheaply from the producers, boast about huge profits and return capital to their shareholders, rather than drill for oil in risky areas. That was in the public domain, but a new truism, Hall noticed, had infiltrated the market. For generations, sages had preached that knowledge was power. In the new millennium, access to unlimited information was easier than ever before, but Hall’s rivals lacked the skill to profitably use that benefit. Unlike him, they were ignoring glaring inconsistencies.
The West’s principal intelligence agencies for oil, the EIA and the International Energy Agency (IEA) in Paris, were using “murky data” and guesswork. In 2003, the IEA, dependent on OECD forecasts of economic activity, estimated that $16 trillion would need to be invested over the next 30 years to maintain existing supplies of energy, and even more as demand for oil surged from 77 million barrels a day in 2002 to 120 million in 2030. Low oil prices, complained the IEA, were limiting investment. Within its report there was a passing reference to China’s increasing demand, but ignorance about its extent. Data showed that the inventories of oil had fallen since 1998, but the consequences were ignored. That vagueness fed Hall’s interest. In his opinion, there would be shortages.
Oil prices at $20 to $30 a barrel were encouraging the heads of the major oil corporations to misinterpret the market. In unison, they ignored the size of China’s growing demand and the Kremlin’s use of oil to increase its political influence. None anticipated that the cushion to cope with increased demand was diminishing.
Most traders and oil producers had foreseen a glut of Iraqi oil after Saddam Hussein’s defeat in 2003, but contrary to expectations, prices rose after the invasion of Iraq. OPEC countries, anxious for more revenue, had increased their production by 900,000 barrels a day (to 25.4 million barrels a day), but shortly after the Allied victory prices unexpectedly fell to $23. Based on his conviction that oil prices would permanently hover between $20 and $30, John Browne sold BP’s Forties field in the North Sea to Apache, a company based in Houston, for $630 million. In Browne’s opinion, the North Sea’s decline, low oil prices, the eventual cost of abandonment and an oil major’s overheads made Forties peripheral and offended his “more for less” philosophy. Outwardly, he blamed repeated huge tax increases levied by Gordon Brown, which would hasten the North Sea’s decline. Unspoken was BP’s deficit of reliable in-house expertise in the North Sea, a corrosive consequence of Browne’s outsourcing. The Forties sale signaled the end of the oil majors’ investment in the North Sea. Events over the following five years would challenge their assumptions. BP had estimated the field possessed reserves of 150 million barrels of oil. Apache would invest $2 billion, and five years later the amount of recoverable oil had increased to 200 million barrels; the corporation had similar results from old wells in Utah that it had bought from Chevron and Exxon. But even Apache was cautious. Advised after buying Forties that prices would fall, it sold production for the following two years at $26, a decision that would cost it about $650 million.
Andy Hall knew few details about the technical problems in the North Sea, the Gulf of Mexico or elsewhere around the globe. The glory of pinpointing a reservoir through thick salt and extracting oil from miles beneath the seabed aroused no curiosity in him. He was interested solely in the statistics of supply and demand, which would enable him to calculate whether to bet long or short. Naturally, his rivals’ bets were significant. In 2003 he was amused that Morgan Stanley, after brokering the purchase of oil wells in the Gulf of Mexico on behalf of Apache for $1.5 billion, had invested in the well and then presold the oil over the following four years for around $35. Hall intended to bet that Morgan Stanley was wrong. “Be long or be wrong,” some believed was his motto: his rivals thought he always betted the same way. “It’s so dangerous,” retorted Hall. “I’d be bankrupt if I did that.” But he hated going short, which meant risking a total loss. The beauty in 2003 was the safety of going long.
John Browne, Lee Raymond and Phil Watts all believed that regardless of the turmoil in the Middle East, the supply of oil from Russia and other non-OPEC countries was safe. “Nations need not be overconcerned about energy security,” Browne had said in July 2002. Relying on BP research that the world’s energy consumption had grown in three consecutive years by less than 0.5 percent, Browne’s confidence was bolstered by the diversity of supply, neutralizing OPEC, and by the 2.4 percent decline of consumption in America. But his reasoning, like that of Raymond and Watts, was flawed. None of their experts had accurately assessed all the consequences of President Putin’s seizure of Yukos in November 2003. Nor did they link it with the decision by Gennadi Timchenko, Putin’s key adviser about oil during the Saint Petersburg era, to move Gunvor, a Russian oil trader, from London to Switzerland. Some said the move was executed to exploit an easier climate, beyond regulatory reach. Others assumed that Gunvor was involved in transfer pricing by reselling Russian crude oil and refined products bought for artificially low prices. A few in Moscow felt that Timchenko was doing the opposite, paying top prices for Russian crude to get market share and earning his profits from refined products. In more prosaic language, Albert Helmig, a former vice chairman of Nymex, who had served on several of its regulatory committees, explained that in Switzerland Timchenko could benefit from “better vocabulary texture.” Few in Europe and America spotted a more important development: the Kremlin was positioning itself for an anticipated increase of oil prices.
Mikhail Khodorkovsky’s arrest had flummoxed the Chinese. Beijing’s agreement with Yukos to build a pipeline and supply crude to Daqing, China’s oil capital in the northeast, was canceled after six months by Putin. “When we faced them at first,” said Yang Bojiang, a Chinese observer, “I thought they were sincere. Now, I think they are probably playing a game.” Despite the arguments advanced by the Chinese to the Kremlin that Japan’s economy was fading, Putin and his advisers were tempted by Tokyo’s offer of additional money and opportunities. Putin did not however ignore China’s urgent need for more oil. The Chinese economy, Beijing’s emissaries explained, was expanding, and electricity produced from the country’s coal-powered stations had become unreliable. As a substitute, households across China were buying diesel generators. To feed the increased domesti
c and industrial demand, China would need an extra million barrels of oil every day. Reports in New York and London about China’s increased demand were common, but Andy Hall knew that no one could accurately predict the amount. Even he did not anticipate that China’s demand would increase during the following year by 2.6 million barrels a day. Global demand for oil in 2004 would increase to 82.4 million barrels a day, and in 2005 would hit 84 million. “How did we miss that?” American experts would ask three years later, contemplating the consequences of a decade of underinvestment. Hall could claim to be the exception. He had been “mulling the idea for around two years.” The best oil traders spent more than half their day reading — everything on the Internet and in trade journals about politics, geology, economics and commodities. Unlike traders who feared competing against the intelligence fed to BP’s, Shell’s and Vitol’s traders, Hall trusted his own sources, and everything he read suggested it was “a perfect opportunity to place the bet.”
Hubbert’s Peak (2001), a book by Kenneth Deffeyes, a retired petroleum engineer, had been inspirational. Deffeyes described the prediction in 1956 by M. King Hubbert, a Shell geophysicist, that oil production in the “Lower 48,” the United States excluding Alaska and Hawaii, would peak between 1966 and 1971, and would decline thereafter. Hubbert’s prediction was based upon an equation using accurate production and reserve figures. The image he drew — a bell-shaped curve showing the rise and fall of oil production — was unusually accurate. In 1970, oil production in the Lower 48 did peak at nine million barrels a day, and then declined. Hall was impressed by that accuracy, and by a wave of other predictions, based upon Hubbert’s assessment, published under the umbrella of the “peak oil theory.”
Like-minded petroleum experts and environmentalists grasped at peak oil to support their campaign against the excessive use of fossil fuel, and in particular against what President George W. Bush would call America’s “addiction to oil.” Emboldened by his new fame, Hubbert had predicted in the early 1970s that production in the North Sea would peak in 1985 at 2.5 million barrels a day, and global crude production would peak in 2000 at 12.5 billion barrels a year. Neither Hall nor other “oil peakists” would be disillusioned when Hubbert proved mistaken. In 2000, the North Sea was still producing over two million barrels a day, and in 2007 global crude production reached 31 billion barrels. Like religious fundamentalists, orthodox peak oil theorists, alias “potheads” or “oil peakists,” preferred to ignore any evidence that contradicted their beliefs.
Hubbert’s gospel had been embraced by Colin Campbell, a retired geologist who had worked for most of the major oil companies, including Amoco, Fina, BP, Texaco, Shell, Chevron and Exxon. “The peak of oil discovery,” wrote Campbell in June 1996, “was passed in the 1960s, and the world started using more than was found in new fields in 1981. The gap between discovery and production has widened since. Many countries, including some important producers, have already passed their peak, suggesting that the world peak of production is now imminent.” He was convinced that the reserves reported by the oil majors and producers were inaccurate, and concluded, “It is now evident that the world faces the dawn of the Second Half of the Age of Oil, when this critical commodity, which plays such a fundamental part in the modern economy, heads into decline due to natural depletion… Petroleum Man will be virtually extinct this century, and Homo sapiens faces a major challenge in adapting to his loss. Peak oil is by all means an important subject.”
In Campbell’s scenario, oil production would peak in 2000 with 24 billion barrels and the ultimate recovery of 1,750 billion barrels. In 1998, he reassessed, or in his words “updated,” his forecasts. Once the oil-price slump had passed, world production had risen to 65.74 million barrels a day, and “proven” reserves were assessed by the IEA and EIA to be about 1,830 billion barrels, sufficient for the 21st century. Campbell challenged that estimate of “proven” reserves, pronouncing that oil production was flat, and would begin to decline in 2010. His argument was enhanced by an allegedly irrefutable “fact”: no major oil reserves had been discovered since the 1960s, and since 1985 consumption had been higher than all the combined new discoveries. Every area in the world, he claimed, including Greenland and Tibet, had been surveyed, and no more great discoveries could be anticipated. Therefore, he wrote, since the depleted reserves were not being replaced, “we conclude that the decline will begin before 2010.” In 2000, Campbell wrote: “The myth of spare capacity is setting the stage for another oil shock.” In his alarming scenario, he asserted that at the current rate of consumption, the new deepwater discoveries of oil would be depleted in just four years, not least because Norway’s reserves were not 49 billion barrels, but 28 billion. Firmly, he predicted that by 2009 the Middle East would be “close to its depletion,” and that oil supplies would peak between 2005 and 2010.
By 2003, both Hubbert’s and Campbell’s methodologies were proved to be flawed. Hubbert’s original forecast of production in America was based upon data compiled over decades, but new technology and prices challenged its accuracy. While America’s “proven” reserves in 1991 were 24.7 billion barrels, the total reserves, including tar oil, oil shale and other condensates, were estimated in 2003 to be 204 billion barrels, and some experts calculated that new technology could increase the ultimate recovery to between 263 billion and 368 billion barrels; and that figure did not include an estimated 180 billion barrels of oil in the Canadian tar sands. Price and technology would determine the amounts: producing a barrel of oil from bitumen cost about $35. “Let me know when we reach peak technology, then we can talk about peak oil,” Paul Siegele, Chevron’s vice president in charge of strategic planning, would say.
Studies in 2005 would show that even Hubbert’s original estimate of America’s peak was mistaken. Between 1970 and 2005 he had underestimated production in the Lower 48 by 15 billion barrels. Failing to take technological advances into account, his mistake in the North Sea was more profound. Production did not peak at 2.5 million barrels a day in 1985, but rose to 2.7 million in 1999, and only fell to 2.09 million in 2003. Hubbert’s — and after his death in 1989, Campbell’s — calculations had been based on inaccurate estimates of the reserves reported by the oil majors and the national oil companies. Despite Hubbert’s tarnished reputation and the error of his own previous predictions, Campbell issued a revised forecast in 2003 that the peak would be reached between 2010 and 2015. Maximum production, he said, would be 87 million barrels a day, including 7.2 million barrels from the four major deepwater areas. He disputed the forecast issued by Guy Caruso, the head of the EIA, that the producers would be able to supply 116 million barrels a day in 2020. The gap between Campbell’s and Caruso’s estimates was enormous. Campbell believed in 2003 that the world’s maximum annual production would peak in 2004 at 28 billion barrels, while the EIA reckoned it would eventually rise to 43 billion barrels. The EIA reported that the oil price would remain at $20, while Campbell was certain there would be rises and “shocks.”
The huge disparity sprang from differing interpretations of statistics. Campbell relied on information provided by OPEC and non-OPEC producers who, for political and financial reasons since their expulsion of the Seven Sisters, had deliberately published inaccurate information and underestimated reserves in order to resist pressure for extra production. Russia’s production and reserves were unreliable on both counts. Similarly, forecasts of supply were always mistaken, because no one could accurately foresee the influence of prices. Campbell’s insistence that no new major discoveries of oil had been announced since the 1960s was also questionable. Although the discovery of new oilfields had declined since 24 were pinpointed during the 1970s (there had been 46 during the previous 20 years), 26 new supergiants had been found in the last 20 years, and four giant fields since 2000 — in Cantarell, off Mexico; Tengiz and Kashagan in Kazakhstan; and Azadegan in Iran. While the bald statistic that 20 percent of oil production came from just 14 fields appeared alarming, the rate of repl
acement, said Campbell’s critics, was underestimated.
Using contradictory data, the two sides clashed over a core issue. Peak oil’s advocates required a “final” estimate of the ultimately recoverable reserves. Campbell estimated that the world had consumed about a trillion barrels of oil, and that a further trillion barrels remained, which at current levels of consumption would last about 34 years. The EIA, BP, and the optimists calculated that there were at least three trillion barrels still to be recovered. ExxonMobil disagreed. Based on its “global geoscience toolkit” and an online database with over 100 terabytes of data, five times more than the entire Library of Congress, Exxon calculated that 15 trillion barrels of energy equivalent remained to be extracted. The combination of Exxon’s exploration costs falling from $2.75 to find a barrel of energy in the 1980s to 44 cents in 2004, and the power of its computer comparisons between the tectonic history of the earth’s plates and the rock distribution in specific locations, magnified the disagreement between it and Campbell about the amount of oil that could be recovered in the Arctic and other remote areas.
Drilling in deep water was akin to putting a man on the moon. In the 1960s and 1970s, Hubbert could not imagine successful drilling six miles beneath the sea’s surface. By the end of 2003, over 1,800 wells had been drilled in deep water in 70 areas, with 990 in the Gulf of Mexico, 383 in Brazil, 111 in Angola and 84 off Nigeria. By then, 47 billion barrels of oil had been found, but most experts believed that at least 100 billion, and possibly 700 billion, barrels of undiscovered oil remained in the Gulf of Mexico. Peak oil, according to Hubbert and Campbell, meant that after 50 percent of the oil had been produced from a major discovery, any further discoveries in the same area would be small and immaterial. The statistics appeared to support their argument. The number of offshore wells in Brazil had declined since 1987, and had almost halved in the Gulf of Mexico, from 109 in 2001 to 56 in 2003. The contrary opinion was equally plausible. Low oil prices during those years deterred expensive exploration, but once prices improved, new possibilities arose. Undiscovered reserves were still likely to be found in the vast, uncharted deepwater areas. Explorers using 3D and 4D seismic, horizontal drills, multilateral wells and smart infill drilling were likely to increase production from mature wells and revive dry ones to extract over 50 percent of the oil, as BP had accomplished at Thunder Horse. The world consumed about 30 billion barrels every year. Contrary to Campbell’s scenario, the problem was not how much oil was in the ground, but how much the producers would spend to extract it.