ditropan
Forumer storico
Robert Feldman (Tokyo)
With oil prices surging, investors are wondering whether the fundamentals in world oil (and energy) markets have changed. Decoding the message from markets requires geology, statistics, and economics. The first installment of this series discussed geology and statistics. This second part discusses the economics.
Economics
How do the geology and statistics in the first part of this series help predict oil prices? By warning us to re-think the way we make oil price forecasts. For example, one approach to oil price prediction is used by my colleagues Ben Funnell and Teun Draaisma (see Selling Oil, Adding Risk, Morgan Stanley, August 5, 2004). They think that oil prices will decline from current levels (US$40+/bbl) because (a) current prices are very far above a linear trend line, and (b) that oil is a mean reverting commodity because demand is price elastic (they mean in the longer run). The first argument is solid as long as the basic supply/demand conditions allow the linear trend to remain intact. However, this trend developed in a period where both demand and supply were rising. If our statistical analysis on oil production potential is correct, and production has peaked, then the world would face rising demand and falling supply. The trend would change. The model of oil as a mean-reverting commodity is subject to similar criticism. The mean can move.
Another model of oil prices is provided by industry analyst Doug Terreson, who focuses on normalized inventories to predict prices. (See Doug’s excellent set of materials, “Crude Oil Fundamentals,” Morgan Stanley Global Energy Teach-In, June 2, 2004, pp. 39-43). Looking at the correlation between inventories and prices, Doug concludes that crude should revert to US$33/bbl oil.
But let’s turn these two models on their heads. What would the market have to be using as inputs to the Draaisma/Funnell and Terreson models in order to generate the current oil price as the equilibrium price? For the former, different trend lines generate different answers. For linear trend lines starting in 1986, 1994 or 1999, the deviation of the current price is still more that 3 standard deviations above trend. Using quadratic trend lines for the same sample periods, the deviations are only slightly smaller. There is no plausible trend line, either linear or quadratic, that would keep today’s oil price within two standard deviations of the predicted value. For the latter model, my eyeballing of Doug’s chart suggests that the market would have to be looking at only about 45 days of inventories in order to generate today’s prices. This level would be significantly below any level in the historical figures of the chart.
The conclusion is that unreasonable values of inputs would be necessary for either the Funnell/Draaisma or the Terreson models to generate the crude oil prices of today.
So Why Is Crude So High?
Is it possible that the market is very far-sighted, sees a Hubbert-like peak of production while demand continues to rise, and is pricing tighter world oil demand into the market now? Yes, this is possible. But there are other explanations as well, as some of my colleagues have pointed out:
(1) It’s extra demand. China demand remains strong even with a China slowdown, as the IEA has recently predicted (see “Investment Highlights: IEA June Oil Market Report, Morgan Stanley, European Oil and Gas team, June 14, 2004). Hence, prices are rising given low elasticity of supply. Demand from India is also rising.
(2) It’s speculation. The proliferation of hedge funds has created a large pool of money that moves in herds. As Ben Funnell and Teun Draaisma point out, “It’s a crowded long.” When the Fed hikes rates and raises the cost of carry for these positions, some may close and trigger price declines of oil futures. Such declines would in turn trigger further selling, and the market would come back to normal. In addition, Ben and Teun point out that the forward curve has moved up with the spot price, with a beta of about 0.7. Either the market is pricing in a new paradigm, or has simply over-reacted.
(3) It’s geopolitics. Iraq and much of the Middle East remain unstable. There are uncertainties in Russia, Venezuela, and Nigeria. Until the world calms down, there will be a risk premium on oil.
Doug Terreson has also brought up two other factors that get less attention. First, the oil industry has undergone a huge reorganization in recent years. Spurred by low returns on capital, the majors merged into Super-Majors. This reorganization has already born fruit in higher returns for the merged entities. Building on Doug’s point, my sense is that higher oil prices could bring further fruits from this reorganization, since the restructured industry is more able to bear the risk of large oil exploration and development projects. Second, Doug points out the major improvements of drilling and recovery technology, so that recovery rates from oil fields have been rising. Although the next set of such technologies is not obvious, the economics of innovation suggest that higher oil prices will stimulate good ideas.
Despite these soothing reasons to think that oil prices may come down some, I find myself in the same camp as my colleagues Steve Roach, Dick Berner, and Eric Chaney, although for an extra reason. Looking at India and China -- the “laboratories of globalization” -- Steve wrote, “As economic development spreads, the real oil price could actually rise over time” (“The Great Oil Debate,” Morgan Stanley, May 27, 2004). Thus, Steve emphasizes the long-term demand reason for higher oil quotes. Dick and Eric wrote, “Long term, we are convinced that the equilibrium price in crude markets has shifted from the US$20/bbl region towards US$30 to US$35, on a Brent basis” (see “Oil Prices: Once Again Marking to Market,” Morgan Stanley, July 26, 2004). They see both demand (from China and India) and supply (lack of investment over the last 15 years) as reasons. While agreeing with the reasons that these colleagues cite, I add another, long-term supply reason: Geologically recoverable production capacity will peak soon -- even if several more large fields emerge.
One last aspect of the debate is also important. It is easy to talk about US$80/bbl oil in the short run, but sustaining such high levels is a different matter. One of the reasons for the failure of the Club of Rome’s dire predictions about the world running out of oil is that the price went up, triggering conservation and innovation. Above US$40/bbl, alternative energy sources become more attractive. After all, people do not buy gasoline because they want gasoline. They buy it because they want transportation. When transportation can be provided more cheaply without gasoline, they will abandon gasoline.
Revenge of the Oil Bears: The Abiogenic Theory of Petroleum Origin
There is one more bizarre twist to the story, one that argues for a more relaxed view of the long-term supply of oil. This twist is the revival of the abiogenic theory of the origin of oil. This theory has excellent scientific pedigree, going back more than a century to Mendeleyev, the inventor of the periodic table. In the last few decades the theory has been espoused chiefly by Russian scientists, and by the late Thomas Gold, a professor-emeritus of astrophysics at Cornell. (See “The Origin of Methane (and Oil) in the Crust of the Earth,” by Thomas Gold, USGS Professional Paper 1570; available at the website www.people.cornell.edu/pages/tg21/usgs.html). These scientists believe that hydrocarbons have origins that predate organic material, in the materials incorporated into the earth when it was formed. The bad news for the abiogenic crowd is that commercial application of the technology for finding and recovering oil (or gas) from potential abiogenic sources is still very far away. It cannot help us in an economically relevant horizon.
Conclusion
My conclusion from all of this is that the world oil market has entered the Crisis phase of a CRIC cycle -- the cycle of Crisis, Response, Improvement, and Complacency that characterizes the interaction of structural reform and economic performance. (See my “Cobwebs and CRICs,” Morgan Stanley, April 4, 2001.) In a nutshell, the oil market is giving the world a swift kick in the pants, in order to stimulate exploration, substitution, and -- the only long-term solution -- innovation. (For my recommendation on what to do in the subsequent Response phase of the CRIC cycle, see “Replacing Pessimism: A CRIC Cycle Approach,” Morgan Stanley, March 27, 2003).
With oil prices surging, investors are wondering whether the fundamentals in world oil (and energy) markets have changed. Decoding the message from markets requires geology, statistics, and economics. The first installment of this series discussed geology and statistics. This second part discusses the economics.
Economics
How do the geology and statistics in the first part of this series help predict oil prices? By warning us to re-think the way we make oil price forecasts. For example, one approach to oil price prediction is used by my colleagues Ben Funnell and Teun Draaisma (see Selling Oil, Adding Risk, Morgan Stanley, August 5, 2004). They think that oil prices will decline from current levels (US$40+/bbl) because (a) current prices are very far above a linear trend line, and (b) that oil is a mean reverting commodity because demand is price elastic (they mean in the longer run). The first argument is solid as long as the basic supply/demand conditions allow the linear trend to remain intact. However, this trend developed in a period where both demand and supply were rising. If our statistical analysis on oil production potential is correct, and production has peaked, then the world would face rising demand and falling supply. The trend would change. The model of oil as a mean-reverting commodity is subject to similar criticism. The mean can move.
Another model of oil prices is provided by industry analyst Doug Terreson, who focuses on normalized inventories to predict prices. (See Doug’s excellent set of materials, “Crude Oil Fundamentals,” Morgan Stanley Global Energy Teach-In, June 2, 2004, pp. 39-43). Looking at the correlation between inventories and prices, Doug concludes that crude should revert to US$33/bbl oil.
But let’s turn these two models on their heads. What would the market have to be using as inputs to the Draaisma/Funnell and Terreson models in order to generate the current oil price as the equilibrium price? For the former, different trend lines generate different answers. For linear trend lines starting in 1986, 1994 or 1999, the deviation of the current price is still more that 3 standard deviations above trend. Using quadratic trend lines for the same sample periods, the deviations are only slightly smaller. There is no plausible trend line, either linear or quadratic, that would keep today’s oil price within two standard deviations of the predicted value. For the latter model, my eyeballing of Doug’s chart suggests that the market would have to be looking at only about 45 days of inventories in order to generate today’s prices. This level would be significantly below any level in the historical figures of the chart.
The conclusion is that unreasonable values of inputs would be necessary for either the Funnell/Draaisma or the Terreson models to generate the crude oil prices of today.
So Why Is Crude So High?
Is it possible that the market is very far-sighted, sees a Hubbert-like peak of production while demand continues to rise, and is pricing tighter world oil demand into the market now? Yes, this is possible. But there are other explanations as well, as some of my colleagues have pointed out:
(1) It’s extra demand. China demand remains strong even with a China slowdown, as the IEA has recently predicted (see “Investment Highlights: IEA June Oil Market Report, Morgan Stanley, European Oil and Gas team, June 14, 2004). Hence, prices are rising given low elasticity of supply. Demand from India is also rising.
(2) It’s speculation. The proliferation of hedge funds has created a large pool of money that moves in herds. As Ben Funnell and Teun Draaisma point out, “It’s a crowded long.” When the Fed hikes rates and raises the cost of carry for these positions, some may close and trigger price declines of oil futures. Such declines would in turn trigger further selling, and the market would come back to normal. In addition, Ben and Teun point out that the forward curve has moved up with the spot price, with a beta of about 0.7. Either the market is pricing in a new paradigm, or has simply over-reacted.
(3) It’s geopolitics. Iraq and much of the Middle East remain unstable. There are uncertainties in Russia, Venezuela, and Nigeria. Until the world calms down, there will be a risk premium on oil.
Doug Terreson has also brought up two other factors that get less attention. First, the oil industry has undergone a huge reorganization in recent years. Spurred by low returns on capital, the majors merged into Super-Majors. This reorganization has already born fruit in higher returns for the merged entities. Building on Doug’s point, my sense is that higher oil prices could bring further fruits from this reorganization, since the restructured industry is more able to bear the risk of large oil exploration and development projects. Second, Doug points out the major improvements of drilling and recovery technology, so that recovery rates from oil fields have been rising. Although the next set of such technologies is not obvious, the economics of innovation suggest that higher oil prices will stimulate good ideas.
Despite these soothing reasons to think that oil prices may come down some, I find myself in the same camp as my colleagues Steve Roach, Dick Berner, and Eric Chaney, although for an extra reason. Looking at India and China -- the “laboratories of globalization” -- Steve wrote, “As economic development spreads, the real oil price could actually rise over time” (“The Great Oil Debate,” Morgan Stanley, May 27, 2004). Thus, Steve emphasizes the long-term demand reason for higher oil quotes. Dick and Eric wrote, “Long term, we are convinced that the equilibrium price in crude markets has shifted from the US$20/bbl region towards US$30 to US$35, on a Brent basis” (see “Oil Prices: Once Again Marking to Market,” Morgan Stanley, July 26, 2004). They see both demand (from China and India) and supply (lack of investment over the last 15 years) as reasons. While agreeing with the reasons that these colleagues cite, I add another, long-term supply reason: Geologically recoverable production capacity will peak soon -- even if several more large fields emerge.
One last aspect of the debate is also important. It is easy to talk about US$80/bbl oil in the short run, but sustaining such high levels is a different matter. One of the reasons for the failure of the Club of Rome’s dire predictions about the world running out of oil is that the price went up, triggering conservation and innovation. Above US$40/bbl, alternative energy sources become more attractive. After all, people do not buy gasoline because they want gasoline. They buy it because they want transportation. When transportation can be provided more cheaply without gasoline, they will abandon gasoline.
Revenge of the Oil Bears: The Abiogenic Theory of Petroleum Origin
There is one more bizarre twist to the story, one that argues for a more relaxed view of the long-term supply of oil. This twist is the revival of the abiogenic theory of the origin of oil. This theory has excellent scientific pedigree, going back more than a century to Mendeleyev, the inventor of the periodic table. In the last few decades the theory has been espoused chiefly by Russian scientists, and by the late Thomas Gold, a professor-emeritus of astrophysics at Cornell. (See “The Origin of Methane (and Oil) in the Crust of the Earth,” by Thomas Gold, USGS Professional Paper 1570; available at the website www.people.cornell.edu/pages/tg21/usgs.html). These scientists believe that hydrocarbons have origins that predate organic material, in the materials incorporated into the earth when it was formed. The bad news for the abiogenic crowd is that commercial application of the technology for finding and recovering oil (or gas) from potential abiogenic sources is still very far away. It cannot help us in an economically relevant horizon.
Conclusion
My conclusion from all of this is that the world oil market has entered the Crisis phase of a CRIC cycle -- the cycle of Crisis, Response, Improvement, and Complacency that characterizes the interaction of structural reform and economic performance. (See my “Cobwebs and CRICs,” Morgan Stanley, April 4, 2001.) In a nutshell, the oil market is giving the world a swift kick in the pants, in order to stimulate exploration, substitution, and -- the only long-term solution -- innovation. (For my recommendation on what to do in the subsequent Response phase of the CRIC cycle, see “Replacing Pessimism: A CRIC Cycle Approach,” Morgan Stanley, March 27, 2003).
