Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


The 2008 oil shock

by Brad Setser
July 19, 2008

Dubai is rather frothy. Landon Thomas of the New York Times reports that rents are up 40%. New supply has to be coming onto the market, but I guess it has yet to catch up with demand. And it isn’t hard to see why. Oil is — thankfully — off its recent highs. But at close to $130 a barrel, it is still well above its average 2007 price of around $70 a barrel.

The rise in prices between 2008 and 2007 has, obviously, come a lot faster than the rise in prices from say 2003 to 2007. The following chart* shows the estimated export revenue of the world’s major oil exporting economies as a share of world GDP if oil averages $120 a barrel this year. An average price of $120 a barrel requires $130 oil for the remainder of the year.


I also plotted the y/y increase in oil export revenues — both in billions of dollars and as a percent of world GDP. If oil averages $120, the 08 rise in oil export revenues would be comparable in size — relative to world GDP — to the rise in 74 and 79. An average oil price of $120 a barrel would increase the export revenue of the oil exporters by about $900 billion.


This calculation assumes that the oil exporters will export about 45 million barrels a day of oil.
Each $5 increase in the average price of oil increases the oil exporters’ revenues by about $80 billion, so if oil ends up averaging $125 a barrel this year rather than $120 a barrel, the increase in the oil exporters revenues would be close to a trillion dollars.

I consequently was a bit slow putting this post up. If I had put it up a few days ago, I could have talked about the “Trillion dollar oil shock.” I learned the value of a catchy headline last week.

* Gaurav Tiwari of the Council on Foreign Relations helped gather the data for this chart. It is based on the BP data for oil production and consumption. Net exports is the difference. I used BP’s prices as well — which likely overstate oil revenue somewhat. But the biggest problem with the data series comes from Russia. BP only has data on Russia from 1985 on. And it almost certainly is the case that Russia didn’t sell its oil at the average world market price until after the Soviet Union collapsed. If any reader has a good suggestion for how to incorporate Russia more smoothly, please let me know. I also would appreciate advice on the average discount that should be applied to the price of sweet light crude to get a better estimate of oil export revenue.


  • Posted by Pallj

    Massively illuminating post. Thanks!

    Still, 3.5% not being worse than in ’79 is strangely comforting.
    If 140 was the peak one could perhaps hope for further decline, as 3.5% is surely not a sustainable level! I’d hope.

    Thanks for putting this into perspective.

  • Posted by Rien Huizer

    Good to see the historical perspective.
    Re Russia: it did export some oil outside Comecon and a few more distant areas like Cuba, the latter in an indirect way. Suggstion: Non ex USSR Comecon uses probably about as much oil as i did in the 1980s and 1970s (output has gone up since collapse but also energy-efficiency). So adding all of ex-Comecon’s current oil imports to the years prior, but at market prices, will probbly get you a resonably accurate picture of ex USSR oil exports. Problem then would be to break up ex USSR into three oil producers (Rus, Az, Kaz) and many consumers. I doubt that that can be done (not pthe supply side, but the demand side. Another way to do this would be to adopt the method for Eastern Europe and reconstruct the USSR for the post collapse years. That would lump Rus, Az and Kaz together as producers, and may be a bigger distortion going forward, than the historical one.

    Incidentally, I would have liked to see a set of two graphs with net export volumes (my hunch is that oil demand (and of course supply) was , on average, flat between 1973 and 1988 (with fluctuations of course, perhaps a trend growth of .5% p.a.. That must have changed in the mid 1990s (sharp decline in consumption in ex-Comecon, little effcet yet from China etc), US economy recovering from early 1990s recession etc.
    In hindsight it seems a miracle that it took oil so long to start rising. I guess that the late 1990s would show the first serious demand increase in two decades, initially matched by supply from spare capacity and gradully eating into more expensive (and more greedy/needy) sources.

  • Posted by RE


    Here is a chart of world oil consumption since 1966. Your assumption is not quite correct.

  • Posted by Rien Huizer

    You are right! However, a time series portion of only 1977-1988 shows strong peaks and a deep valley.but a horizontal linear trend. I had not realied that production (and of course consumption) grew so much between the two oil crises. Wonder if we will see a similar supply response again (once prices start to come down a little, inflation erodes export proceeds and alternatives cloud the horizon and the consumer dusts off his SUV), now prices are again very high. So many alternatives (with poor greenhouse chracteristics but that is for later) are now affordable, just a matter of time and money.

    This chart must have the correct volume figures for the USSR.
    And by the way, re the USSR/Comecon again: their oil numbers (In money terms) are probably dodgy, but so r their GDPs ..

  • Posted by mitchell porter

    Speaking of trillions of dollars, the GCC’s planned infrastructure spending over the next ten years is on that scale:

  • Posted by bsetser

    Rien — what is certainly true is that non-Gulf supply increased significantly after 75 — so by the mid 80s, the combination of conservation and new supply meant the Saudis had a significant resevoir of spare capacity. For a while they were only producing something like 4 mbd. The collapse of the Soviet Union also may have had the effect of increasing the effective supply for the rest of us in the 90s.

  • Posted by Rachel

    Brad, the collapse of the soviet Union opened up new sources of global oil supply – but the real increase from FSU countries didn’t kick in until about 99/00. For most of the 90s, oil production fell significantly from the levels of the late 80s meaning that net exports were lower despite lower domestic consumption.
    net oil exports from Russia and Kazakhstan were almost 6mbd in 85, falling to just over 3mbd in 94, surpassing 6mbd in 2002 and reaching over 8mbd in 2005 before leveling off more recently.

    The real increases in supply in the 90s came from the Gulf – and to a lesser extent norway. The data on consumption in Africa is not consistent.

    all data comes from BP.
    The real increases in supply in the 90s came from the Gulf – and to a lesser extent norway. The data on consumption in Africa is not consistent.

    all data comes from BP.

  • Posted by Rien Huizer

    Brad, Rachel,
    The thing has many moving parts. It is easy to lose onself in small part of this mechanism keeping in mind also that medium term production capacity are simply not in the public domain. And both your comments are really interesting. I cannot help being fascinated by two things: (a) the apparent inability of net oil exports (which reprsent probably more than 70% of OECD oil consumption, even with large producers like Mexico, US, Canada and Norway in that group) to move outside +/- 1% of world GDP, and most of the time staying within half that range.
    b) the other fascinating thing is the symmetry of the spikes.

    Hence I wonder (re a) what happened in the past 10 years
    to make net exports move towards the upper band.and (b) will the current spike also have steep downslope. Generous reward available for correct answer..

  • Posted by Pallj

    End consumer price is significantly higher than the export price, and counting for consumption of locally produced oil we’d be looking at a much higher percentage of GWP for oil.

    Unless oil exporting economies beef up their local consumption of imported goods to the tune of a trillion, oil importing economies will grind to a halt.

    Assuming that at 3.5% a proverbial pain threshold has been reached, you might as well start shorting oil as vigorously as you can.

    There is nothing that suggests the steep trajectory of price increase will be followed by a level price. Everything points to a sharp drop, unless there were some quick reduction on the supply side.

    My bank details available upon request 😉

  • Posted by bsetser

    rachel — thanks for correcting me. i guess the fall in production in the former soviet union was bigger than the fall in demand.

  • Posted by Stefan, Tallinn

    If oil producers reduced production by 1%, prices would probably rise 10% (ceteris paribus).

    So why don’t they do that?

    Answer: Because it works only short-term. That is also why the graph of oil income in relation to GDP is interesting. It is VERY easy for it to rise SHORT-TERM, but very difficult/impossible long-term. Thus it will fall back.

    A number of mechanisms will ensure that oil’s relative value to other goods and services stays comparatively stable.

  • Posted by Frank Gifford

    What is different this time (from the 70’s) is no new finds like Alaska and North Sea coming oline plus robust demand growth from oil producing countries while most of worldwide oil is coming from fields more than fifty years old now in production decline. Meanwhile, new discoveries are not able to replace the declines from existing fields. Furthermore, the world is awash in dollars. Think of the current price decline as a short term correction. Long term, we are headed for higher oil prices, much higher. Somewhere I read that within five years, the amount of exportable oil worldwide will be less than fifty percent of its current level. I suggest reading and understanding the ideas of L King Hubbard in regards to the depletion of oil fields/countries/world.

  • Posted by Rien Huizer


    Thanks, as said, many moving parts: certainly less near term available supply of crude of the right kind (and far too little refining capacity, and not enough product tankers, etc) but several million bpd (perhaps as much as Saudi now) locked up in well-explred areas in Iran and Iraq. Visited the peak oil site and have a strong snse of deja vue (been around for a long time. The truth is: nobody knows, but big areas have been unavailable for exploration (US especially), oil production in the middle east ouside GCC declined for political reasons (war, de facto boycott Iran). Lots of current exporters are not nterested in icrasing production beyound current levels even if they can, etc
    Then there is the option to turn NG (reatively plentiful) into variuos forms of synthetic liquids (succsfull plant in Bintulu Malaysia makes superclean fuels, technology now stable) and methanol. So there is a little bit to say about peak oil but it is inherently impossible to make accurate predictions. Nobody knows enough about long term oil/substitutes availability and demand response to higher prices .I think even a layman like myself could make a convincing case about peak oil being at least ten more years into the future based on the same data.. But, we know, thanks to graph 2 above) at least that in the past, sharp increases in net oil exports share in world GDP have hardly pierced the noise floor in the GDP data, and have generally ben followed by equally fast declines. That makes oil look like a pretty normal good to me. Markets in the past had apparently better information than the many speculating theorists

  • Posted by Pallj
  • Posted by Sinomania!

    RE, I wonder how much of the increase in oil consumption that began after 1993 was just China alone?

  • Posted by Rien Huizer

    Palj, thanks. Slide # 14 is most interesting.