Understanding crude oil prices

That’s the title of my latest research paper. Here’s the summary from the paper’s introduction.

How would one go about explaining what oil prices have been doing and predicting where
they might be headed next? This paper explores three broad ways one might approach this.
The first is a statistical investigation of the basic correlations in the historical data. The
second is to look at the predictions of economic theory as to how oil prices should behave
over time. The third is to examine in detail the fundamental determinants and prospects
for demand and supply. Reconciling the conclusions drawn from these different perspectives
is an interesting intellectual challenge, and necessary if we are to claim to understand what
is going on.

In terms of statistical regularities, the paper notes that changes in the real price of oil
have historically tended to be (1) permanent, (2) difficult to predict, and (3) governed by
very different regimes at different points in time.

From the perspective of economic theory, we review three separate restrictions on the
time path of crude oil prices that should all hold in equilibrium. The first of these arises from
storage arbitrage, the second from financial futures contracts, and the third from the fact
that oil is a depletable resource. We also discuss whether commodity futures speculation
by investors with no direct role in the supply or demand for oil itself could be regarded as
a separate force influencing oil prices.

In terms of the determinants of demand, we note that the price elasticity of demand
is challenging to measure but appears to be quite low and to have decreased in the most
recent data. Income elasticity is easier to estimate, and is near unity for countries in an
early stage of development but substantially less than one in recent U.S. data. On the
supply side, we note problems with interpreting OPEC as a traditional cartel and with
cataloging intermediate-term supply prospects despite the very long development lead times
in the industry. We also relate the challenge of depletion to the past and possible future
geographic distribution of production.

Our overall conclusion is that the low price-elasticity of short-run demand and supply,
the vulnerability of supplies to disruptions, and the peak in U.S. oil production account for
the broad behavior of oil prices over 1970-1997. Although the traditional economic theory
of exhaustible resources does not fit in an obvious way into this historical account, the
profound change in demand coming from the newly industrialized countries and recognition
of the finiteness of this resource offers a plausible explanation for more recent developments.
In other words, the scarcity rent may have been negligible for previous generations but is
now becoming significant.

You can read the entire paper here. Comments welcome from all.

Technorati Tags: ,


41 thoughts on “Understanding crude oil prices

  1. Jim

    I just skimmed part of the paper and found it very readable and 75% plus understandable for me a non-economist. Thanks! I’ll print it out and study it over the next few days. It threw me for a moment when you talked about Iranian production as falling by 5400 barrels per day on p17 but these are all kbarrels (per fig. 5) so its making sense to me again. Thanks again!

  2. Boswell

    I wonder if the combination of tiny price elasticity of demand and intertemporal calculations with respect to prices on the part of producers could provide incentive for them to lower current production for purely financial reasons.

  3. shan

    It seems a good overview of the situation, and was fun to read. I’m a big fan of oil market commentary, so much of it was not new to me, but reading it in the language of an economist certainly filled in a few gaps in my understanding.

    One point I’m not quite sure if I understand: “an update of the CERA methodology would leave one still quite optimistic about near-term oil supplies”. …

    Isn’t that more of an update of the forecast using the same method as CERA, rather than an update of the methodology? If it didn’t produce accurate results before with CERAs data, how likely is it that it will do so now with the wikipedia data?

  4. kio

    The price index for energy (crude proxy to oil price)intercepted the core CPI in the second part of 2007 and now at the level of 233 (core CPI at 214 and CPI at 213). So, one can partly relate the oil price jump to some recovery process. This observation is supported by other price indices which are close to the CPI.
    In 1981, teh differnce between energy price index and core CPI was -10. In April 2008, it was -19. Not too big by all means.
    Similar recovery has been observed for food price since 2000. The index for housing had an excellent recovery from 1996 to 2007. The index for transportation started its recovery in 2003. If to consider turning points in the past for the energy and other indices, one can assume that the index for energy is close to its turning point and core CPI soon will be growing faster than oil.
    More interesting is that the index for medical care was at the level of 362 in April 2008 with the same relative importance for CPI (~7%) as energy (~8%). The index for education grows even faster but has lower effect on CPI.

  5. MG

    I enjoyed the paper. One minor comment: the elasticity of demand for Saudi oil (p 24), would also be affected by the elasticity of supply of other producers. Perhaps this is assumed to be zero, but that wasn’t stated.

  6. John Mashey

    Very useful paper, thanks! A few notes & questions:

    1) Re: Saudi Arabia leaving oil in ground for future generations, i.e., the piggy bank. US might well apply this to ANWR, etc, if only to answer a question a friend’s 13-year-old daughter asked him:

    “Daddy, are you adults going to leave any oil for us?” He had no answer beyond “not so much”.

    2) I found the data for Figure 6 and recreated the chart. Minor nit wishes: URLs would be helpful, also, putting in a few more year markers along the path could be helpful.

    Figure 6 invites some scenario-spinning in the face of Peak Oil, and comparison with GDP projections elsewhere (in climate-change mitigation or damage effects on GDP).

    The question: what are plausible bounds on trajectory of Figure 6 going forward?

    a) It is hard to believe that the US oil consumption can ever go much higher. However, with efficiency improvements, avoidance of waste, and substitutions (electric cars, biofuels, etc), hopefully the line at least keeps moving to the right, i.e., GDP continues to grow.

    b) If one believes typical Peak Oil curves, and is not as optimistic as CERA, one would expect that sometime between 2050 and 2100, US oil consumption (as defined by EIA) would be back at 0 on the chart, i.e., the same level as 1949.

    c) Many projections use at least 2% CAGR for US:
    2050: GDP 2.4X larger than 2006 (+.9 log)
    2100: GDP 6.4X larger than 2006 (+1.9 log)

    d) I think that means that the trajectory crosses the y=0 axis somewhere between 2.9 and 3.9, adding 2006′s 2.0 to each of the above. Of course, on that chart, anything in the range [2.9-3.9, 0] is rather far outside any GDP/oil ratio seen for a long time. I.e., 2050′s ratio of GDP:oil would be ~9X higher than 2006 [if 2050 is back to 1949]. 2100′s is 23X higher.

    e) Alternatively, maintaining 2006 GDP with a 1949 level of oil requires improving GDP:oil ratio by ~4X over 2006. Certainly 2X seems relatively straightforward, albeit a radical change to the US vehicle fleet.

    f) but that seems to offer an envelope of plausible trajectories, i.e., a line starting at [1.94, 1.27], and ending somewhere between [1.94, 0] and [3.9, 0], with the latter assuming that it’s 2100 before oil returns to 1949, and 2% growth continues.

    Are those reasonable interpretations of this sort of chart? I do keep seeing projections that use GDP CAGRs of 2% or more, with apparently no effect from oil. They still make me nervous.

  7. bill j

    The contrast with the period after 1997 is interesting. If you consider the decisive change in geo-politics in this period it was the transformation of the former centrally planned economies of the USSR, Eastern Europe and China into capitalist ones.
    The impact of this transformation can be broken into two distinct phases, the phase of destruction between 1991-2000 and then the period after that.
    During the period of destruction the plan collapsed and alongside it the demand of these economies for oil. Therefore, you had a massive increase in supply to the capitalist world market, combined with a massive reduction in use by an important new component of it. Hence there was a dramatic fall in raw materials prices across the board during the 1990s.
    From around 2000 onwards the process of capitalist restoration was complete and these economies began to recover. Hence there was a massive increase in demand but only a relatively smaller increase in supply, as the oil sector of these countries may have declined in the 1990s, but not as much as the general decline in the output of these economies.
    So what you’re seeing today is a huge increase in demand with a lag in supply, and this is the fundamental driver behind high petrol and raw materials prices today.
    In addition there are the other well known factors, the depreciation of the dollar and the role of speculators etc.

  8. Dan Weber

    Hopefully I’m not being too picayune, but “begs the question” means “assumes the proposition,” not “raises the question.”
    Or perhaps I’m tilting at windmills and language has marched on. ;)

  9. Bruce Hall

    An interesting paper. Just a couple of questions:
    1. Because of recent potential discoveries of oil off Brazil and Nova Scotia [among others] and the vast tracts of tar sands/shale that are now economically feasible to develop, can the use of drilled wells from existing fields be a valid indicator of the actual oil available?
    2. There is not much referenced about the political aspect of U.S. supply mentioned. This has led to 40 years of no real development of oil resources for this country. An example is the restriction of offshore drilling in the GOM and Florida. While the U.S. prevents domestic supply increases from these sources, other nations, including China, are actively pursuing these resources. What impact would the elimination of those restrictions have on oil prices… when brought on line?

  10. Josh Stern

    I listened to an interview with an oil/natgas production company (EOG) executive last week who claimed that the largest bottleneck limiting his long term ability to increase production was a shortage of qualified geologists. I certainly wouldn’t take his remarks as gospel, but they do fit with other suggesting that the supply response to demand in this field may have a much larger lag than what one would typically assume for modeling. For this factor mentioned, the lag involves step 1) companies perceive there will be higher price/tight supply, step 2) budgetary plans are modified, step 3) salaries for geologists go up, step 4) more people decide to train in that area, step 5) more people graduate and begin looking for new drilling prospects, etc.
    This factor might has a 10-20 year lag.

  11. ramster

    “But the algebra of compound growth suggests that if demand continues to grow in China and other countries at its current rate, the date at which the scarcity rent will start to make an important contribution to the price, if not here already, cannot be far away.”
    That’s the last line of your paper. In the US, gas consumption is dropping in response to the recent price increases (consider the recent data on US vehicle miles driven). My understanding is that in many developing countries, gas is subsidized so that consumers aren’t seeing the price increases that you would expect given the large increase in oil prices over the last few years. This is the reason for the continuing increase in demand that you refer to. These subsidies are busting the budgets of the countries that offer them and probably won’t continue.
    So it seems that a key issue is not an economic question but a political one. When will the subsidizing countries end the subsidy? If you can answer that question, you’ll know the path of future oil prices.

  12. JDH

    John Mashey, you ask extremely important questions, and I must apologize that I do not have good answers. I think trends like these are helpful to predict what could happen in situations similar to those we have seen, but I am very nervous about using them to describe something so different from anything within this historical experience.

    All I am confident in taking out of Figure 6 is that the income elasticity has fallen over time and is now clearly less than unity, and that the price elasticity is much less than unity and has also fallen over time. One thing that’s extremely clear from economic theory is that the price elasticity must increase from those low values as the dollar share of energy in the total expenditures grows. I believe we’ve already reached a point where we should be expecting historical price elasticities rather than the recent lack of response.

  13. John Mashey


    But many (not all) of the subsidizing countries. are oil exporters, and China has a lot of money to spend this way.

    Hence, at least some of this subsidization seems likely to continue for a while. I’d guess Saudi Arabia will subsidize it for quite a while.

  14. ramster

    That may be true of the lightly populated oil exporters such as Saudi but Russia, Venezuela, Iran, Mexico, Nigeria and Indonesia (combined population of roughly 700 million) have lots of spending needs other than cheap gas. Even if you’re an oil exporter, subsidizing gas has an opportunity cost and that’s going to get a lot harder to ignore as the cost of the subsidies baloons. My guess is that things will change sooner rather than later.

  15. JDH

    Bruce, any oil in ANWR is still there. Isn’t a key question here one of when is the optimal time to develop this resource? Wouldn’t you agree that the U.S. would have been better off economically if we produced whatever oil is in ANWR in 2008 rather than in 1998? Are you sure we wouldn’t be better off yet producing it in 2018?

    You are right that there may be more offshore resources as well, but did you have a reaction to the graph I produced showing that the fraction of U.S. production that is coming from offshore sources has been increasing over time, despite your description of this as a key reason for falling production? I think it is very hard to reach any conclusion other than depletion from the major conventional U.S. oil-producing areas set in long, long ago.

  16. JDH

    Jim, Shan, MG, and Dan, thanks for your suggestions, which I’ve incorporated in the updated version now posted. Further comments from others most welcome.

  17. Valuethinker

    You say thousand (000) on quoting production (thousand per day) when you mean million.
    Sorry I don’t have the page refs in front of me, but it was in the first half.

  18. Colin Clarke

    Very interesting paper but the real analysis may be simpler.
    As an ex supply manager at an oil company…
    In commodity markets when demand got above 90-95% of supply prices rose sharply.
    Whether this was natural gas, or petrochemicals OR CRUDE.
    In most years the ethylene business was just about cash breakeven , negative to zero positive accounting profit.
    But one year every 7 years or so the market tightened , selling prices skyrocketed….profits were 1 million dollars per day on our plants.
    Oil from 1980 to 2005 has typically had a production overhang….now everything possible is being produced.
    While there is some speculative premium , the short run inelasticity means that consumers in the USA can afford 4 dollars per gallon although consumption is now falling.
    Oil and other minerals have 3-10 year investment lead times…..the Saudis 3 years, gulf of mexico 10 …

  19. Knowledge Problem

    Are oil prices going higher?

    Michael Giberson Today a Washington Post article discusses the most recent oil price forecast from Goldman Sachs analysts Arjun N. Murti and Jeffrey Currie, which have oil prices averaging $141/bbl for the second half of 2008. As usual, it isn’t…

  20. George Barwood

    The paper is good. A few thoughts of my own.
    The oil market, at least until the last few years, has been dominated by spot prices, and the balance of short term supply and demand.
    If speculators are (at considerable risk to their capital) driving current prices higher, this is in fact for the good of all, in that oil is being saved for use later when it will be put to better use at a higher price. So the speculation is good.
    Hence the response to politicians who blame speculators for higher prices is that this is a form of saving for the future, which is praiseworthy.
    I don’t think there is any easy answer to the question of how much should be saved. It is clear to me though that historically not enough has been saved – oil was sold only a few years ago for a price that was clearly too low, indicating that the futures market was not powerful enough to raise prices then.
    I suspect the reason for this is that the risks of commodity investment have made such arbitrage too risky for most investors.

  21. John Mashey

    Q for Colin Clarke:
    Your timeframes are useful reminders that if an oilfield is producing at its maximum safe flow rate, drilling more wells and building more infrastructure takes a while, even in Saudi Arabia.
    Do you have any comments on lag times induced by need to hire drilling rigs these days? Bloomberg says that Petrobras has leased about 80% of the world’s deepest-water drill rigs.
    This highlights an issue for those who actually produce oil:
    1) They incur substantial costs in paying for leases, finding oil, drilling, and building pipelines, maintaining machinery.
    2) Hence, once someone has built that, even if one expects oil prices to rise, one wants to be pumping at least some oil to cover costs and pay back the effective debt incurred, and certainly cover the company overhead. ExxonMobil cannot just decide to ship zero oil some quarter…
    3) Hence, the best way (especially for a national oil company like Aramco) to save oil is to run existing wells at maximum non-damaging flow-rates, and simply defer drilling some new wells.
    4) But, one has to be very careful about being over-optimistic about the availability of resources, like drill rigs and teams, especially the rarer, more expensive ones, as they may not be available at the time one would like to have them.

  22. Bruce Hall

    Certainly there has been good production coming out of existing leases in the GOM. My point was that there are significant areas offshore that have been “permanently” off-limits and that represents lost opportunity.
    Because of the development time required, it may be 2018 before we could see any of the resources I mentioned… if Congress allowed the development to go forward now. But just because Annie loves Tomorrow doesn’t mean that is a valid strategic approach to energy.
    And there is no conceivable reason why Congress should have prevented the publication of the regulations necessary to begin development of shale oil/tar sands in the great white north.
    I suspect that at some point this “hands-off” approach to domestic resources [which is truly a market manipulation by the government just as much as it would be if oil companies were hoarding the resources] will raise the ire of voters in a big way.

  23. Colin Clarke

    i have been out of the industry for 4 years…which gives me time to peruse things like this.
    several comments….
    1. the only change from 1980 to 2008 regarding technology options is regards to natural gas.
    the efficiency of combined cycle plants and turbine size is much greater than forecast then.
    LNG has become more feasible as a traded commodity instead of a dedicated field to a market.
    ethanol, biofuels, wind, solar, coal to liquids, coal to gas , batteries were all major reaeaerch areas.
    2. over a beer with friends still in the industry drill rig availability and ALL equipment costs are major hassles….this was same 1980 to 1982.
    3. when you go back to 1999 and oil was 15 dollars per barrel and i was still in industry no one was using anything higher than 20 dollars as econonmic basis for project justification.
    there are no experts who can truly speak with certainty…
    best analysis is 2007 NPC study……

  24. Jeremy

    110/100 for your paper. It is so refreshing to read a true academic work presented in true balanced academic fashion.
    In a world of “hype” it seems most people have sold their academic soul to either sensationalist causes or pharmaceutical companies with the next “wonder drug”. (A good example being the ludicrous claims by the Global Warming Camp and their “pseudo-scientists” who love the media attention and funding that catestrophic predictions engender.)
    Your observation that Governments are now completely in control of Oil Production and that “Big Oil” is a minor factor is bang on. Your observation about cheating and “swing producer” behaviors reminds me of a piece in the book “Thinking Strategically” by Bary Nalebuff. It reminds us that even though Governments are in total control they can behave independently and in a self-interested unenlightened manner – it is a classic case of the “tragedy of the commons”.
    No one Government feels responsible for supply restrictions and extremely high oil prices – so all have been acting with self interest (raising royalty rates/taxes and generally restricting access to reserves or limiting the expansion capital needed by their NOC’s…from environmental protection, to NIMBY, to “save the oil for our children” the justifications are varied but near universal)
    Of course, inevitably at some price, global recession and oil demand destruction will follow, and the individual “lack of responsibility” will have lead to a “collective crisis”. Jeff Rubin at CIBC has identified what high oil prices may do to global trade.
    Another good example of the tragedy of the commons is when many countries allow overfishing of cod (with the preaviling attiude that my individual take won’t affect the overall outcome) which then leads to a fisheries collapse of Eastern Canada. (Jared’s book Collapse is much about the tragedy of the commons concept with regard to resources)
    What a shame that recent oil inelasticity has left governments very smug with their restrictive policies – perhaps that is until now…of course Government Officials globally will be only to happy to blame the speculators (its “them” not “us”) for the recent run up and the subsequent glbal recession (when it inevitably occurs)!
    It is little known, but a certain consultant called Pedro Van Meurs advises many countries at extremely high Government Energy Policy making levels. As food prices continue to rise and little children are left suffering with swollen hungry bellies – it will be an important reminder to Government Energy Policy makers everywhere that the “tragedy of the commons” has just as important ramifications as filling Government coffers. I hope Mr Meurs sleeps well at night.

  25. Nemo

    I can suggest a typo fix: On page 6, you introduce a notation as “Et(Pt+1) but then drop the parens and use it as EtPt+1 through the rest of the paper. I think the original notation (with the parens) makes more sense, but at least you should be consistent. :-)
    I have a couple of questions about the paper.
    In the case of a product for which the Hotelling Principle applies, Jovanovic (2007) noted that self-fulfilling bubble paths could be indexed by the residual quantity of oil that never gets produced.
    I do not understand what this is trying to say. Can anyone provide a brief elaboration, or do I have to read Jovanovic’s paper?
    For this reason, an ongoing speculative price bubble would have to result in continuous inventory accumulation, or else be ratified by cuts in production. The former is clearly unsustainable, and if it is the latter, one might make the case that the supply cuts rather than the speculation itself has been the ultimate cause
    of the price increase.

    I have asked this before but I have not seen a satisfactory response. Couldn’t an artificially high long-term futures price itself create an incentive for producers to limit supply (i.e., to sell later rather than sell now)? Could that not create a “self-fulfilling prophecy” of higher spot prices, as the institutional money rolled forward into new futures contracts? Could that not last indefinitely, like any other speculative bubble?

  26. Jeremy

    Did anyone notice the recent backtracking of Russia and UK on petroleum related taxes.
    For Prime Minister Brown it appears to be a complete U-turn from his own tax policies as Finance Minister. Both nations are seeing declining production.
    Are we finally seeing the emergence of politicial awareness that taxation has played a critical role?
    Oil supply/demand is by no stretch of the imagination behaving as a “free market”.

  27. JDH

    Thanks for the suggestion, Nemo. I’ll make the change.

    The Hotelling constraint is derived from the condition that the sum of total oil extracted over all time equals the total quantity in the ground. Jovanovic’s bubble path is one in which the difference equation followed by the price of oil is that derived by Hotelling but the sum of oil extracted is less than the total quantity– some is still left in the ground when the sun becomes a nova.

    Your thought that producers respond to the increase in price by cutting production is precisely what I am referring to when I say that the price increases could be ratified by cuts in production.

  28. Nemo

    Hi, Professor. Thank you for the reply.
    When you say, “ratified by cuts in production”, I am not sure you mean the same thing I am trying to say. I am not suggesting that the producers might “ratify” the futures price by cutting production; rather, I am suggesting that futures speculation might directly cause cuts in production.
    The observation underlying the Hotelling principle is that, for a finite resource, the question a producer asks is not “How many barrels should I produce today?”, but rather, “When should I produce each barrel?”
    The answer depends on the future price of oil, the present and future production costs, and the prevailing rate of interest. But the futures markets have a direct influence on the future price of oil!
    Imagine a long-only index fund that only invests in long-term oil futures; say one year out. As those futures near expiration, the fund sells the (now near-term) futures and rolls the proceeds into new long-term futures. Assume they never take delivery. Then imagine that large amounts of money enters this fund; enough that all the other long/short speculation is negligible by comparison.
    Even without taking delivery — i.e., even without any inventory build — the mere act of pushing the long-term futures contract higher would, by Hotelling’s reasoning, create an incentive for producers to withhold production. In the steady state, the fund would be engaging in a long-term money-losing strategy… But it would be a very slow bleed, especially if interest rates were low. The constant flood of money into the long-term futures would constantly induce producers to curtail current production, keeping the spot price high and thus making the futures bet into a self-fulfilling prophecy (almost).
    People do buy the GLD ETF, and that actually charges a storage fee. Why would they not buy such a “slowly bleed” fund for exposure to oil?

  29. JDH

    Nemo, again I don’t recognize a fundamental difference between what we are saying. But if the Hotelling constraint was satisfied initially (at the end of time the oil will be exactly exhausted), and if from the scheme you are describing the oil is at every date at a higher price relative to that initial scenario, then the result would be that under the new scenario, at the end of time we still have some oil left. So I think you are indeed talking about a Jovanovic type of bubble. By using the word “ratify” I am trying to make the point that the physical supply response of producers is a critical element in any such scenario. This response is only in their interests to the extent we’re ignoring the role of the transversality (or “end-of-time”) condition.

  30. Nemo

    Again, thank you for the reply.
    From what you say, it sounds like I am, in fact, describing a “Jovanovic type of bubble”. I will track down his paper…
    I think my issue was with your use of “ratify”. I find it an odd way to characterize an action purely in one’s own economic interest. (If a competitor offered me a higher salary to leave my current job, I would not say I was “ratifying” their offer by accepting it.)
    If long-only commodity funds really are adding an enduring premium to long-dated futures prices, then the curtailing of production is simply a rational economic response by producers.
    How to tell whether this is happening is the question. The answer, I think, depends on whether producers today are unable to increase production, or unwilling to do so. I do not know where to go to answer this question. The quotes from King Abdullah are interesting, but not sufficient. :-)

  31. Nemo

    I have now read Jovanovic’s paper (“Bubbles in Prices of Exhaustible Resources”, available at his Web site).
    I think what I am proposing is that long-only futures funds can create a Jovanovic type of bubble, by creating a incentive for the indefinite withholding of some amount of the resource. Put another way, investment in long-only futures contracts could, by itself, move prices from the Hotelling equilibrium to one of the other equilibria as described by Jovanovic. So buying and rolling the long-dated futures contracts can have the same ultimate effect as buying and hoarding the physical commodity…

  32. Juan

    While my post and your reply seem to have been removed, I would like to take the moment to say my use of ‘patently absurd’ was, first, very impolite and secondly, an implication that I know exactly what price curves the various physical oils would have followed within a price regime that does not exist but can only be assumed.
    As I believe you touch upon in your recent paper, production management decisions may well have resulted in greater-than-otherwise supply (in)elasticities. Simultaneously, products demand side inelasticities have, I believe, been greater due to both direct and indirect subsidies, both of which are coming under pressure.
    Perhaps it is correct to think of house-as-asset price inflation and the ability to monetize this as having provided a multi-year indirect subsidy to many within the largest consuming nation? (Related; relatively inefficient car fleet may provoke greater behavioral changes than expected).
    If the above paragraph is correct, some portion of the 2002-08 rise in prices has no doubt been rational. I would be interested in the terms of government to government contracts; private sector long-term contracts, and transfer pricing. Not that these are perfectly divorced from the present “market-related” price regime but degrees of contradiction.
    Since, in large part, it depends on the shape of the forward curve, the oil storage trade cannot be considered an ‘indefinite withholding’, nevertheless, it does relate to your ‘Addendum’.
    NB, a number of the largest financial institutions also own and/or lease storage facilities here, Europe and I believe also Asia.

  33. John Mashey

    A few more questions [Continuing my May 25 7:11pm & your May 26 12:20pm comments].

    1. Figure 7 is interesting, especially when combined with your comment about price elasticities, i.e., it looks crude oil as % GDP could rise some more. Can you point at any good studies that would calibrate the short-term oil price that would stop US GDP growth, given the current domain, somewhat different from the earlier oil shocks?

    2. At EIA, I see an outlook to 2030 of smooth growth in US consumption of oil, following a slight flat spot around 2006. Their oil supply&demand spreadsheet shows 2006-2030 CAGRs of 0.4% for Domestic Crude Production, Imports, and Total Crude Supply.

    The price spreadsheet shows little sign of scarcity rents.

    I’m having trouble making sense of these numbers when compared to yours.

    3. I’m trying to understand various models for determining costs of climate change mitigation. As noted earlier, they all seem to use reference BAU cases that assume GDP and GDP/capita growth rates of several percent for the next 50-100 years.

    One example would be the DICE model of William Nordhaus & co, as in his book that just arrived, “A Question of Balance”. It analyzes different policy options, claiming an economically optimal policy of a carbon tax equivalent to:

    2015 $0.11 / gallon

    2025 $0.14 / gallon


    2105 $0.59 / gallon

    I.e., carbon prices above these are seen as suboptimal. However, oil price changes due to supply/demand don’t seem to factor in anywhere in these (or other models). It seems odd that more than $.11 is an economy-damaging tax, but going from $3 to $4/gallon in US has no effect. Perhaps that’s because it is a world GDP model? I.e., if the elasticity remains low, people pay, but the money shows up in someone else’s GDP?

  34. JDH

    John Mashey, I’m afraid that neither the EIA nor I nor anyone else would be able to offer you definitive answers to your questions. No one even knows whether real GDP will grow for 2008 or not. Let’s see if the EIA predicted oil price for 2008 of $84/barrel proves to be correct before putting too much confidence in their 2030 prediction of $70.

  35. John Mashey

    JDH: thanks. Needless to say:

    1) Given that to get an average of $84 this year, we’d have to average ~$70 for the second half … my confidence in $70 oil in 2030 is not high. I’d certainly love to see that happen, since the only way I can imagine that is that by 2030 almost all cars (and many smaller trucks) were either BEVs or PHEVs.

    2) I keep asking this stuff because I find that most climate mitigation – economics models seem to rely on this or similar data, and major policy decisions are being argued based on it, without usually considering that “predicting is hard, especially the future.” This especially seems true across major inflection points / state changes, which just don’t seem to appear in the models.

    This makes me nervous, especially when I can’t make sense of the base data. If I hadn’t already been sensitive to the dangers of over-confident forecasting, “Flaw of Averages” Sam Savage’s short course at Stanford would have made me so.

    I sure hope the EIA is right, but so far, I have reservations.

  36. Darren

    I’ve discovered something strange when running the numbers in Prof. Hamilton’s report… I am unable to
    duplicate the figure of 1.12% average quarterly growth in the natural log of real oil prices (page 3) (I get a
    figure of .4% average quarterly growth) – and no, I don’t think it’s an artifact of monthly/quarterly conversion :)

    Given the relative experience levels of myself and JDH, the error is likely mine, but if anyone has any light to
    shed on my probable error I’d be much obliged.

    As a sanity check, here are some of my numbers
    Columns are
    real oil price in April 2008 dollars using CPI-U,
    100*LN(real oil price),
    % change in 100*LN(real oil price) since last quarter)

    • Mar-70 18.69 292.81 -0.54
    • Jun-70 18.45 291.51 -0.44
    • Sep-70 18.04 289.28 -0.76
    • Dec-70 19.11 295.05 1.99
    • Mar-71 19.02 294.55 -0.17
    • Jun-71 18.78 293.30 -0.42
    • Sep-71 18.65 292.57 -0.25
    • Dec-71 18.51 291.83 -0.25
    • Mar-72 18.38 291.11 -0.25
    • Jun-72 18.24 290.38 -0.25
    • Sep-72 18.07 289.43 -0.33
    • Dec-72 17.90 288.48 -0.33
    • (snipped for brevity, email for full set)
    • Mar-06 67.34 420.98 1.19
    • Jun-06 75.11 431.89 2.59
    • Sep-06 67.27 420.87 -2.55
    • Dec-06 65.20 417.75 -0.74
    • Mar-07 63.10 414.47 -0.79
    • Jun-07 69.58 424.25 2.36
    • Sep-07 81.92 440.57 3.85
    • Dec-07 92.61 452.83 2.78
    • Mar-08 105.76 466.12 2.93

    If I average the full column of % changes I get .4%, not 1.12%

  37. JDH

    Not quite sure what you’re doing with these numbers, Darren. The logarithmic percent change is calculated as 100*ln(Y/X). Thus for example, with the data you’ve provided here, the percent change in nominal crude oil prices for 2008:Q1 would be 100*ln(105.76/92.61) = 13.28. The percent change in real oil prices would be 13.28 – 100*ln(466.12/452.83) = 10.38. Don’t know where you’d get the claim that real oil prices only increased 2.93% in 2008:Q1.

    There are some other slight differences in our raw numbers, though this shouldn’t make much difference for these kind of calculations. I took the monthly WTI from FRED, who give the March 2008 average WTI as 105.56 rather than the 105.76 number you use above. You’re also using something different for the CPI.

  38. Darren

    Thanks for your reply, Prof. Hamilton. That works much better. I was simply calculating the percentage change in the logs themselves
    ie: 100*(4.6612-4.5283)/4.5283, which in retrospect doesn’t make much sense anyway.

Comments are closed.