« Do not buy art on cruise ships | Main | Spend More Today »
How can the price of oil move so much in one day?
Over the two previous days oil fell $10.50 a barrel. By definition this is driven by news about supply and demand but has so much news come out so quickly?
Here are two ways to think about the mechanisms at work. First, some producers could supply more but they figure that China will be buying more tomorrow so maybe it is better to wait. If they see a signal that future global demand will be lower, they are less likely to let oil sit in the ground. In other words the market develops the expectation -- true or not -- that oil supply will rise more rapidly than had been thought (or "decline less rapidly" may in some cases be a more accurate phrase but the net direction of the effect is the same). Lower expected demand is thus paired with greater expected supply and that tends to make price volatile. Higher expected demand is paired with lower expected supply in similar fashion. noting in either case that you can make lots of different assumptions about the relative timing of the expected changes.
Second, any new information leads to more trading and to more trading at different ranges of price and quantity. This trading reveals more information about the elasticities of market supply and demand curves and that information in turn feeds back into the market price. In a nutshell, some initial price and quantity movements lead to further price and quantity movements.
Neither of these phenomenon are correctly called "bubbles" but neither do they fit the story where the price of oil is determined by fundamentals alone. "Expectations" is a central word here, noting that only time will tell whether or not the expectations are rooted in reality or not.
Posted by Tyler Cowen on July 17, 2008 at 07:09 AM in Economics | Permalink
Comments
china, india, the middle east are the real culprits. How they keep quiet and suck up oil is suspect.
Posted by: tony at Jul 17, 2008 8:40:37 AM
Or, oil could actually be a bubble. And with the lessons of dot-com and housing behind us, you should get out before everyone else decides to do that. So there would be an exaggerated reaction to any signal that prices are going to drop.
Posted by: Michael at Jul 17, 2008 9:17:09 AM
I'm seeing some news coming out of the US in the past few days as the culprit. First, gasoline consumption has actually dropped in response to prices. Second, gas guzzling vehicles are selling very poorly, and GM and Ford are flailing as a result. Third, the US economy is taking a very big hit, which should mean less frivolous spending, which should mean even lower fuel consumption. That's, respectively, short, long and medium term drops in fuel consumption right there.
There are two other whimsical reasons for the price drop. One, Batman came out. Two, the price of oil could be moving away from the dollar to another, more stable, currency. (That's right, I said it, the dollar is unstable.)
Posted by: Xmas at Jul 17, 2008 9:35:32 AM
It is not that all of a sudden. We had almost an identical move week before last. But other indicators of a significant weakening of world economic growth and oil demand are emerging. The DryDock index is down sharply, industrial commodity prices appear to have peaked and real US oil imports collapsed last month and are down some 14% from last year. All three of these indicators tend to lead oil prices. In the stock market month to date the S&P energy sector is down at double digit levels -- until now this has been masked because the press was focusing on the plunge in financial stocks. Conditions for a peaking of oil prices are falling into place. Note also that as oil fell financial stocks rebounded.
Posted by: spencer at Jul 17, 2008 9:41:52 AM
In response to tony (with the first comment,): do you know how incredibly condescending that sounds to people from China, India and the Middle East? Together they account for half the people in the world, yet they probably equal the oil consumption of the US which has only 300m people. Who is sucking up oil then?
Posted by: sunbomb at Jul 17, 2008 9:51:23 AM
"First, some producers could supply more but they figure that China will be buying more tomorrow so maybe it is better to wait."
Many people play this into their IOC conspiracy theories. As, I understand it though, the IOCs that are catching all the blame can't really do this anywhere (which I wish you would have mentioned in the post if I'm correct).
In the US, state laws generally prevent oil & gas companies from "shuttting-in" wells unless they are not making money. Abroad, host countries seem to always require production, and similarly don't permit IOCs to shut-in wells to wait for a better market.
I suppose it's possible to hold off on drilling in the first place, but for IOCs there's bound to be another company that's okay with the current level of profits who will take a lease from a US mineral owner or a foreign government to drill it now.
Posted by: effay at Jul 17, 2008 10:29:54 AM
The situation is unstable. There isn't a lot of excess capacity. Consumption can be expected to keep going up -- china and india are poised to increase consumption as well as increase industrial production etc, and they will need more oil. Problems in the USA might dampen their economies, but in the medium run why should they let the US economy determine theirs? Iraqi oil production could plummet any time if the other shoe drops and violence goes up. Iranian oil production could plummet quickly if the USA or israel attacks them. The US dollar is unstable, nobody knows how fast it will depreciate. Six months from now will we be starting a partial pullout from iraq or will we be digging in for at least a 4-year stay?
Given all this, what market-maker could be expected to stabilise oil prices? The idea that prices will be stable unless new information causes them to move is backward. In a free market prices are inherently unstable unless a market-maker works competently at damping fluctuations.
Posted by: J Thomas at Jul 17, 2008 11:02:56 AM
Perhaps this is a naive/stupid comment, but I believe that much of the short term movements of any commodity price have a strong element of randomness (noise?).
Also, let us assume that the "natural" price of oil is today's market (given existing production and consumption patterns, etc.) is $105 per bbl. (this is a pure guess on my part -- any number would do in this example). Would not any movement well above or below this range (say $102 to $108, though again this range is a pure supposition) be characterized by frequent, somewhat "random", and often significant shifts?
Finally, can we identify the major market players responsible for this price shift? Different groups of investors and/or speculators typically have varied goals and motivations, and it is often not an easy process to sort these out.
I think that I am saying essentially what Tyler is in slightly different terminology.
Posted by: Chris Pepin at Jul 17, 2008 11:29:39 AM
There are two other whimsical reasons for the price drop. One, Batman came out. Two, the price of oil could be moving away from the dollar to another, more stable, currency. (That's right, I said it, the dollar is unstable.)
The Fed’s broad index of the dollar’s value was 95.97 on February 28 and 95.97 on July 8. The broad index of the dollar reached its low in March of this year. It is currently around what it was in December of 2007. The fall in the dollar is something that happened throughout last year, for the most part. It has not been happening this year since March (and was only a very small decline from December to March, almost all of which has been recovered), despite oil prices going up.
Another possible reason, I suppose-- President Bush did lift the executive branch restriction on offshore drilling this week. Of course, that has no immediate effect, especially since the restriction on drilling has a legislative restriction as well that much be repealed. However, he did also call on Congress to do that same; it's possible that oil traders, mistakenly or not, were responding to that possibility.
Posted by: John Thacker at Jul 17, 2008 11:44:50 AM
I disagree with Tyler that oil companies try to "time" the market from a supply perspective. I work for an oil company and nobody here has any idea where prices are going. Oil companies are price takers, despite what all the never-ending grandstanding government inquiries. If we did, we would be trading futures and be gazillionaires. The necessary investments are very large with very long lead times and many uncertainties. Even current production is pretty much impossible to "turn off" rapidly.
I think it has more to do with a tightly balanced market combined with perceptions of future demand. There are signs that demand is being affected by price, as it rightly should.
Posted by: asparagus at Jul 17, 2008 11:46:33 AM
@John Thacker:
The legislative restriction on drilling does not have to be repealed. It sunsets on September 30. If Congress does not pass another one, offshore drilling will be open for business on October 1. In the current political environment, the chances of passage of a new restriction are considerably less than one, and I think that is part of what the market is reacting to.
Posted by: Peter at Jul 17, 2008 12:10:44 PM
No comment on the big 1-day drop, but here's my prediction: at some point in the near future, oil is going to drop almost 50% in a few months for no apparent reason. Everyone's going to buy the dips, seeing this as a once-in-a-lifetime opportunity. Only after several months will we get the figures showing that Chinese GDP growth, construction and industrial investment have slowed to a crawl. We will slap our collective foreheads as we realize that all our fancy charts and predictions forgot to account for the mere possibility of a Chinese recession. Oops!
Posted by: Lawrence at Jul 17, 2008 12:16:05 PM
Why would today's oil prices react strongly to something that might marginally increase oil supply 10 years from now?
Far more likely that they're reacting to the news that Bush is making a last-ditch attempt at diplomacy with iran before the attack, and the attempt might "succeed" and we won't after all have a war.
Or it could be anything. Without a market-maker adequate to stabilise prices, whyever would we expect prices to be stable? We should expect prices to fluctuate wildly in a free market in the absence of an actor whose job is to stabilise them.
Posted by: J Thomas at Jul 17, 2008 12:22:54 PM
Why would today's oil prices react strongly to something that might marginally increase oil supply 10 years from now?
For the same reasons that we should conserve more now if we're going to run out 10, 20, or 40 years in the future. Sen. Schumer has claimed that half a million barrels of day extra now would decrease the price by much more than $10. If we're going to have a million barrels of oil a day more ten years in the future, then perhaps we can get away with consuming slightly more now. If we don't have those million barrels of oil a day in the future, then we need to save more oil today, and perhaps it took $10 a barrel to get the world to use the fraction of oil less now that means that there will be enough in the future.
If anyone sees that "oh, we have enough oil now, but we don't have enough in the future," well, it makes sense for everyone, not just the speculators but society, to make the price higher now so that we use less now and the price is lower later, when there's both more and we're used to conserving. If future supply increases, then it doesn't make quite as much sense to conserve now.
That doesn't mean that it's necessarily the answer, but it's not all that bad a theory.
Far more likely that they're reacting to the news that Bush is making a last-ditch attempt at diplomacy with iran before the attack, and the attempt might "succeed" and we won't after all have a war.
The odds of an attack on Iran, even just an airstrike by Israel and nothing by the US, have always been pretty small. (Including as judged by betters at Intrade.) Someone who thinks that an attack is so certain is very much not based in reality.
Of course, as you say, it could be anything. People could suddenly think that breakthroughs have been made in alternative fuel sources.
Posted by: John Thacker at Jul 17, 2008 12:50:30 PM
You can see on Intrade's site that there's been a lot more movement in the "US to lift offshore drilling ban" contract (which is moderately high, around 40%) than in the "US or Israel to overtly strike Iran" contract (which remains low, around 8%).
Posted by: John Thacker at Jul 17, 2008 12:57:30 PM
The (rational) expectation of the Fed's response to galloping inflation? Stop printing money, slow the economy, less demand for oil.
Posted by: PJ at Jul 17, 2008 1:05:04 PM
If you want to give Bush credit, the announcement that the US is sort of opening relations with Iran probably has a much bigger impact on oil market than the lifting of restrictions on off-shore drilling.
Posted by: spencer at Jul 17, 2008 1:05:55 PM
Or could it have anything to do with this?
http://www.ericmargolis.com/archives/2008/06/at_last_some_tr.php
Posted by: Michael at Jul 17, 2008 2:24:47 PM
John Thacker, your first hypothesis above is that the markets are rationally handling expectations about oil supplies 10 years down the road.
This does not fit my experience, observation, or reading. We have no market mechanism to achieve the right amount of oil production 10 years from now. For most of the time of oil production we had no efficient guess how much oil there would be -- a series of unexpectedly large oil strikes revised the estimated reserves at random intervals. When each strike was announced did oil prices immediately fall a lot to reflect the increased supply in coming years? It's been awhile since we had large unexpected bonanzas, and total reserves are more predictable than they used to be though improvements in recovery, oil shale, etc might still have hard-to-predict effects on supply in coming years. I don't think anybody is making workable predictions 10 years out. Among other things we can't predict the stability of middle-east governments, or the venezuelan government.
We *should* conserve more now, but how does that translate to our actually doing it? All the oil that gets pumped will be consumed, except the little bit of it we pump back into the ground for our strategic reserves. Every barrel of pumped oil the USA fails to consume will be consumed by somebody else. It's the market at work, man.
The odds of an attack on Iran, even just an airstrike by Israel and nothing by the US, have always been pretty small. (Including as judged by betters at Intrade.)
And how can you prepare for it? It's like planning for a Richter 10 earthquake, you do better to just shrug your shoulders and hope it doesn't happen. You can't bet oil futures on the possibility even if you think it's likely -- when you win the markets are likely to shut down and your oil gets confiscated by worried governments etc.
So the markets can factor in a small or moderate disruption in trade in a fizzle of a war, which isn't all that likely, and a change in the likelihood won't affect oil prices that much.
Is there another explanation? How about this?
http://afp.google.com/article/ALeqM5hkLCJIKdipOHqJ8bwdjN1JUZSFtA
------
Oil surges to new heights after Israeli warning on Iran
Jun 6, 2008
NEW YORK (AFP) — Crude oil prices went on a record-setting surge Friday as fears of a new Middle East conflict were fanned by comments from a top Israeli official about Iran.
New York's main oil futures contract, light sweet crude for July delivery, leapt 10.75 dollars a barrel -- its biggest one-day jump ever -- to close at a record 138.54 dollars.
------
Last month we had the biggest one-day rise in oil prices ever, which some people attributed to the israel/iran thing. Now we get the biggest-one-day drop in many years, smaller than that rise, which pretty much cancels out that increase. Over 5 weeks the price of oil hardly rose at all!
If last month's war scare was the reason for that one-day rise, maybe people have gotten used to the idea now and have started to discount it.
But my own favorite theory is the null hypothesis. There's nothing special going on except the market is increasingly volatile. We can expect bigger highs and lows because nobody is putting many resources into moderating those fluctuations.
Posted by: J Thomas at Jul 17, 2008 2:30:15 PM
I never thought I would disagree with Tyler Cowen and the fact that I am going to lends me to believe that I am wrong. I would have very little chance to hold his water bucket in the corner of a fight ring let alone to spar with him but I am going to make a comment that I believe is a disagreement.
I believe that physical commodity spot prices trade at a settled price that is based on the immediate supply and demand balance. Expectations are not incorporated into the price. The spot price of oil is relatively high because of the relative scarcity of oil at this moment. The clearing price is based on what a buyer is willing to pay for that next unit at this time given the supply. The producer will get the market price now and in the future.
If supply is being withheld from the market it would be for non-econmic reasons (government). A supplier would make a high margin to sell now and would be able to supply oil in the future by producing more in the future. The assumption of withholding supply for economic reasons would only make sense if there is an absolute scarcity of the product for the long term. Stocks and bonds trade on expectations of future cashflows but commodities trade on supply and demand. Oil is a commodity not a monopoly/oligopoly (or is it?).
Another reason the spot price would trade in a volitile way would be from an event risk that would cause increased immediate demand for inventory as a buffer to a supply shock (big convenience yield).
Posted by: SA at Jul 17, 2008 2:43:08 PM
The short-run elasticity of oil is extremely low -- the supply is fixed, and consumers give it high priority in their budgeting. Large price swings are therefore required to balance even small changes in quantity.
Posted by: David Wright at Jul 17, 2008 2:59:43 PM
Of course it's expectations! Expectations of fundamentals. This is how all markets work.
Don't we make every decision in our lives based on what we expect the outcome of that decision to be? And don't we change our minds when new information makes us rethink the expected outcome.
Why would oil markets be different. When traders talk about "fundamentals", they're always talking about expectations of the change in fundamentals. It's still about fundamentals. I thought everybody understood that.
I'm waiting for congress to explain to me how the oil price dropped so much what with all those rascally speculators still about and uncontrolled by the saintly Friends of Anegelo Mozilo [sarcasm].
Posted by: Methinks at Jul 17, 2008 6:40:35 PM
Of course it's expectations! Expectations of fundamentals. This is how all markets work.
Don't we make every decision in our lives based on what we expect the outcome of that decision to be? And don't we change our minds when new information makes us rethink the expected outcome.
Why would oil markets be different. When traders talk about "fundamentals", they're always talking about expectations of the change in fundamentals. It's still about fundamentals. I thought everybody understood that.
I'm waiting for congress to explain to me how the oil price dropped so much what with all those rascally speculators still about and uncontrolled by the saintly Friends of Anegelo Mozilo [sarcasm].
Posted by: Methinks at Jul 17, 2008 6:41:05 PM
According to Intrade.com, the odds of an attack on Iran by September 30th, 2008 have dropped from 25% as of July 8th to 7% today. Let's say a trader expected a 100% increase in the price of oil in the event of an attack with probability = 25%, and zero change if not. Then a weighted average return would have been 25% as of July 8th and is down to 7% today. That would justify a large position reduction for any sensibly managed hedge fund.
Also, since volatility in oil prices has been rising, a risk model based on mean-variance theory would suggest an even larger position reduction in oil futures. Finance geeks will appreciate that a doubling of volatility, with no change in the expected return, implies cutting your position size not in half but by 75% -- because the denominator of the mean/variance calculation contains volatility squared.
The combination of a sharply lower expected return on holding oil and higher volatility in the price would mean one thing to an astute trader -- take your money and run.
Interestingly, odds of an attack by March 31, 2009 have not changed much at all and still hover near 40%. A more meaningful reduction in geopolitical tensions could therefore imply an even larger reduction in speculative demand for oil.
Posted by: Bill S at Jul 17, 2008 8:14:15 PM
Bill, that's an excellent post. Just to add....
Oil inventory figures, released Wednesday morning, showed an increase of 3 million barrels when a small draw was expected. Draws are typical in summer months when most people travel. This is a relatively big increase in inventories and has obviously negative implications for changes in demand.
Posted by: Methinks at Jul 17, 2008 11:40:03 PM
