This weekend, we are in Chicago for a family event. After dinner (over Kir Royals), we discuss the economy, tax cuts and Oil. The subject of whether Oil is a bubble comes up. What’s become a now standard argument — that Oil, inflation adjusted, is relatively cheap (see chart at left) — gets made.
My B-in-L Bob, a very senior BP exec (now retired), is the one who initiates the "Oil is a bubble" discussion. All the inflation adjusted charts seem to only go back to include the 1970s — and that’s not far enough to show the true price trend of oil. Bob argues that Oil has been in a very long downtrend, and the 1970s price spike was an aberration. So too, the 2003-05 run up. A longer, inflation adjusted chart would reveal that the present spike is aberrational, and unlikely to be sustainable. I am somewhat incredulous of this claim.
His point however, is well taken. While he is expecting an eventual mean reversion, simply base dupon price, I have a similar expectation based upon market cycles. The next recession (there’s always a next recession, just as there’s always a next recovery) will see reduced demand for Oil, and that will allow prices to fall.
By coincidence, an article in the NYTimes the next day has this chart in it of the real price of Gasoline for the past century. Surprisingly, it supports Bob’s position pretty well:
Real Gasoline Prices, U.S., 1920 – 2005 
Graphic courtesy of the NYT
Let’s look at this technically: This chart is of a very long downtrend. There was a minor spike in the 1930s, which eventually reverted back to trend. There was a major spike in the 1970’s; It subsequently collapsed in price some 50%, from $3 / gal to $1.50, over the 1980 to ’84 timeframe. It would not be outrageous to call that price action bubble-like.
This certainly supports the argument that Gasoline/Oil prices are unsustainably high at present, and will eventually revert back to trend. However, as the prior two spikes make clear, the key is the word eventually. It typically takes about a half of a decade or longer for this mean price reversion to occur. With Oil prices still in an unptrend at present (i.e., no reversals indicated yet), this implies that $30 a barrel (real) oil may be a phenomenon no sooner than 2010. Work that data into your econometric models — high oil persisting at least another few years — and the 2006-2008 period is an even more pronounced recession than I have previously suggested.
Incidentally, consistent with our views on multi variant analysis — a single factor should not get all the credit or blame for major economic or market events — the significant strength of the 1980’s can be as easily traced to the plummeting cost of real gasoline as it could to tax cuts, interest rates, or the stock market. Looking at this helps to explain to some small degree why the 1980s were such a powerful economic period.
>
Sources:
At the Pumps and on the Web, Drivers Check for Lowest Prices
KIRK JOHNSON
NYTimes, August 13, 2005
http://www.nytimes.com/2005/08/13/national/13gas.html
The Plot Thickens
ALAN ABELSON
UP AND DOWN WALL STREET
Barron’s MONDAY, AUGUST 15, 2005
http://online.barrons.com/article/SB112389049076212298.html
But the price can’t go to zero, so there must be a limit price.
Price at the pump (local retail product) doesn’t really tell you much about whether oil (international raw commodity) is a bubble. Taxes, regulations, refining capacity, etc. all fuzzy the issue.
Secondly, bubble suggests speculators buying high now with the expectation of selling higher later. If you expect prices to decline because a recession reduces demand, that’s not a bubble, that’s a supply/demand imbalance.
Thirdly, the price of oil is largely a concern for its broader economic impact. Here’s a chart I prepared that shows oil consumption in dollars as a percent of GDP: Chart [flickr.com]
This chart makes a better argument for the economic growth of the 90’s than yours does for the economic growth of the 80’s.
Here’s the same chart as the previous one, but with a second data series showing the barrels of oil consumed per $1000 GDP (Y2000 constant dollars): Chart [flickr.com]
This makes a pretty picture of economic adjustment to a crude oil supply shock.
The real problem with this type of analysis is that it presumes that oil is essentially an economic good that can be generated at infinite amounts. In reality, oil is a phyiscal finite consumable good – every barrel we suck out of the ground not only leaves one less barrel in the ground for tomorrow, but the next barrel out of the ground is usually more expensive to get at.
The geological analysis of whether ‘peak oil’ is coming sooner or later is arguable, but the fact that oil is finite (at least in the timeframes we deal with) means that you can’t simply look at a past technical analysis from purely economic grounds and learn anything useful about the future.
The thing most people don’t get about the current situation is that everybody except for the Saudis is pumping as much oil as they possibly can, and the Saudis’ spare capacity is a low-grade crappy crude oil that nobody can practically refine. Huge investments in additional oil fields can bring some more sources online, but by the time that happens, some existing fields will have already started to decline, or so the PO theory goes.
While in the PO community I’m on the conservative end, I do think that our grandchildren are going to look back at our (US) decision to subsidize suburbia as the same variety of empire-killer as the lead-pipe story for the Romans. There’s no way for us to unwind suburban sprawl with expensive oil; we won’t be able to afford to rebuild our rail network, much less build buildings where they need to be.
Following up that last post, it seems pretty obvious that crude oil consumption has remained relatively constant at 2% GDP only because of significantly increased efficiencies. This seems inconsistent with the “crude just keeps on getting cheaper” line of argument.
In fact, the raw price information I used pretty much refutes Bob’s assertion that “oil has been in a very long downtrend”.
This whole analysis is based on the assumption that oil prices and inflation are independent variables. In fact the opposite is true. The rise in oil prices always leads to inflation in the long term – any plot of oil price adjusted to inflation would show the trend that you have plotted. Only a short term severe disruption in oil supplies would show up in the above plot. The 70s spike fits this profile perfectly.
“any plot of oil price adjusted to inflation would show the trend that you have plotted.”
So, what you’re saying is that if the price of oil doubles, then the price of everything else doubles, too. Right?
Did your discussion with the BIL somehow avoid the phrase ‘peak oil’? To my mind, the key factor interplay here has already come up in the comments thread: demand for energy, productivity, and oil production.
If oil is extended at $60, you believe demand increases will not swamp improvements in productivity and the eventual production peak.
The other two words I was surprised not to hear from Bob were ‘India’ and ‘China’ — but that’s another story.
Technical analysis and charts would help if oil were a renewable resource or something we could manufacture in large quantities (say, like more bozo Internet retailers..). It is neither. We can make a little from various forms of tar sands, etc, but not in the quantities we currently consume.
Maybe we’re in a bubble, maybe it’s supply constraints due to years of underinvestment and explosive demand from emerging markets, or maybe it’s Peak Oil manifesting itself.
As they say: You bets your money and you takes your chances.
See you in 10-15 years for the answer.
PS. I think Clinton clearly has cheap oil to thank for the 1990’s prosperity.
First time I’ve seen the price of gasoline influencing the type of car an auto enthusiast is looking to buy on the autoweek.com forum: http://forums.autoweek.com/thread.jspa?forumID=14&threadID=19245
Perhaps we can now say that it is not $50/barrel and $2.00/gal, but prices north of $60/barrel and $2.50/gal that will affect consumer’s car buying choices. And two years of today’s prices will likely see the big impact on automakers and the whole auto retailing industry. E.g., two years from now if today’s gas prices remain in effect, how much lower than cost for new full sized SUVs (it will take lower prices than “employee prices” to move product) and how much lower still to sell a used full size SUV?
If you look at a 100 year log chart of oil you can see how significant the break of $40 was. Oil is not a bubble…..at least not yet. Im looking for a parabolic rise in oil and the stocks over the next 6-9 months which will be followed by a sustained correction. Oil has followed the market and will retreat when the market gets back to bear mode. The next rise in oil will be because of supply and not increased demand. Watch the oil stocks closely. They are still trading at 10-15x earnings, with the best numbers out there and with most of the street still decidedly bearish.
Is Oil A Bubble?
Barry Ritholtz:My B-in-L Bob, a very senior BP exec (now retired), is the one who initiates the “Oil is a bubble” discussion. All the inflation adjusted charts seem to only go back to include the 1970s — and that’s not…
What’s the deal with the spike in the 1930s? That can’t have anything to do with a roaring economy. Could it be that large movements in the price of oil have little to do with the state of the economy as a whole?
Hi Barry,
I know you are a huge fan of great graphics.
An excellent graph that provides a long-term perspective on oil prices can be found at BP Statistical Review 2005. At the bottom of page 14 of the Review is an excellent graph showing oil prices in real and nominal dollars for over 100 years.
I hope this helps.
Best regards,
Kevin
royce,
The problem with looking at oil that far back is that the energy economy pre-WWII was essentially a completely different animal to what we have today. Much more local; much more guesswork (nobody knew where the oil was or even might be). Not a lot of transoceanic shipping of oil going on, for instance.
Today, we have a pretty good idea where we’ve already looked and where there still might be some; and how little chance there is of another huge find like the Middle East.
BP data: WW Oil consumption growth growing steady around ~3% for the last few years Oil prices however are up over 100% in the last year. The Economist and other sources are also reporting that Oil is not in shortage in Asia – rather refined gasoline is in shortage in North America due to the flat refining capacity,oil grade mismatch, and recent short term production disruptions. With the above in mind it seems pretty clear (100% vs. 3%) that most of the oil price run up in the last year is due to short term factors such as speculation, fear, and refining capacity, not secular demand.
I’ve been bullish on energy stocks for the last few years, but it now seems like a good time to adjust my stops and be ready for a correction. That’s not to say that a speculation binge driven by leveraged hedge funds couldn’t drive prices higher in the near term – the “Oil is running out” story has a nice emotional pitch to it – I’m looking forward to the ride.
How does the identification of a long-term downtrend constitute evidence that the downtrend is likely to continue? A purely empirical model of the world based on extrapolation makes no sense at all. Unless you can identify the reasons for the trend and argue that they are likely to continue, I would be inclined to draw the opposite conclusion: oil has become more and more undervalued over time, and its price is likely to continue rising (though not monotonically).
Price at the pump would include more than the raw cost of oil anyway. Improvements in refining technology, pipelines for cheaper transportation, etc, etc. So the downtrend would be more from the improvements than the straigh tcost of the oil itself, I would think.
KNZN: There is a “persistency” to markets — its a component of chaos theory — and its why trends tend to remainin effect far longer than can be explained by randomness or other theories . . .
Is Oil’s Rise Sustainable?
Oil’s powerful ascent to $67/barrel has been downright awe-inspiring. And while still not at an inflation-adjusted all-time high, it’s definitely starting to get up there. So, is the price of oil akin to a bubbly, overbought tech-stock, or is this
Is Oil’s Rise Sustainable?
Oil’s powerful ascent to $67/barrel has been downright awe-inspiring. And while still not at an inflation-adjusted all-time high, it’s definitely starting to get up there. So, is the price of oil akin to a bubbly, overbought tech-stock, or is this
In discussing the 1980s don’t forget that oil played a much larger role in the collapse of the USSR then miititary spending. I’s seen a chart that shows Russian real GDP growth has a 0.95 correlation with
russian oil production.
The argument that oil is just another commodity that is seeing another big cyclical swing around a long term downtrend is supported by comparing the price of oil to other commodity prices. For example since 1990 oil prices and copper prices have had a 0.97 correlation.
But given the long lead times for oil supplies, oil prices should continue to rise until demand drops — and in the short to intermediate run the only way that will occur is a world wide recession.
Kevin — thanks for reference to BP info.
the BP long term chart of oil prices provides a very different and challenging story. One reason theat I have been sceptical of the peak oil argument was the real price of oil was still well below prior peaks.
Buy the BP data tells a very different story.
The IEA expects oil demand to rise to 85.9 million barrels a day by the fourth quarter of this year, which is higher than the global refining capacity. Goldman Sachs’ even predicted $105 per barrel. I have been checking out dailyreckoning.com for my info and their opinions on oil as a bubble. The U.S. is near the bottom of the barrel for many of America’s oil fields with declines in output of 160,000 barrels a day.
An interesting analysis would be looking at the 5 year futures (if they existing going back long times). that ought to help at least say if a spike was based on short term issues (like a war) or if it was based on a view of continued shortage.
Invert that plot and show Gallons/Dollar and you will see Peak Oil written all over it.
The peak in Peak Oil occurs at maximum global production which equates to inflation-corrected cheapest per volume delivered. This is essentially Gallons/Dollar. If it keeps going, we are seeing the backside of the peak, no two ways about it
Barry, I enjoy your occasional comments over at Altercation concerning the radio & music biz. I will definitely check in here again as you have quite the wide-ranging perspective on things!
kir royales /after/ dinner not particularly good for the digestion?
I think long term we are going run into a peak oil situation. At the same time in the short term their could be a downward pressure on prices. While we obviously won’t find another Middle East size resource its possible we could open some more taps like those in the gulf of mexico that were at least first widely reported a few months ago http://abcnews.go.com/Business/wireStory?id=2395433. Long term though we are dealing with a finite resource. Hopefully we will start looking at energy alternatives so that we have a soft long transition instead a fast hard transition. One thing that is somewhat positive is I talked to someone in the oil exploration industry recently. They said that their are alot of areas that it was not profitable to explore and drill because they were hard to reach. As oil prices rise these will become profitable. So basically as oil prices increase more oil will become available. And hopefully as prices increase more effort will be put into exploring alternatives.