February 22, 2011
Lots of action in oil prices today, as the unrest has spread from Tunisia and Egypt-- which produce relatively modest amounts of crude oil-- to Libya, the country sandwiched between them, and producer of over 2% of the world's crude oil supply. Rather than try to guess where those developments are going to lead, I wanted today to try to make sense of another equally striking development in oil markets over the last 6 weeks-- the disparity between the price of oil in the Midwest United States and that elsewhere in the world.
The graph below compares the price of West Texas Intermediate for delivery in Cushing, Oklahoma with that for North Sea Brent in Europe. Usually you can hardly tell the two apart on a long-term graph like this.
But in the last few weeks, the discrepancy has at times exceeded $15 a barrel.
The prices of other high-quality international crudes, like Nigerian Bonny Light and Algerian Sahara Blend, have hugged close to Brent as they also pulled away from the WTI.
Stuart Staniford and Donald Marron have discussions of this, with the Oil Drum having the most thorough treatment. An influx of Canadian crude through pipelines delivering to the Midwest, along with impressive production gains from North Dakota, are bringing a lot of new oil to Cushing, depressing the price there relative to other locations.
But the fact that more oil is being produced and delivered locally is not an adequate explanation for the price discrepancy. There is a very fundamental economic principle that goods that are close substitutes and easily transported should sell for a similar price even if the fundamentals of local supply and demand differ across regions. The economic force that ensures this is arbitrage. Whenever price differentials emerge, there are very powerful monetary incentives for anyone to buy in the market where the price is low in order to sell in the market where price is high. The consequences of such actions are to drive the low price up and the high price down until profitable arbitrage opportunities no longer remain.
The first line of defense of the Law of One Price is financial arbitrage, or what some observers refer to as trading in "paper oil." If I simultaneously buy WTI futures and sell Brent futures, I can take my profit before anybody on the ground gets their hands dirty moving actual oil. Plenty of people make their living doing just this, and this is in fact what is usually responsible for keeping the relative prices in different markets instantaneously synchronized in response to news that arises anywhere in the world.
But for some reason, the folks who usually play this game have taken their marbles and gone home. Maybe they were burned by trying to maintain a tighter Brent-WTI spread than true physical arbitrage warrants given the changing supply conditions on the ground, maybe they ran into other financial constraints, or maybe something else in the market scared them off.
But that leaves lots of money on the table right now for physical arbitrage-- buy the physical product where it is cheap and sell it where it is more expensive. The Oil Drum notes that the most efficient way to do this would be to start running the Seaway pipeline, which is currently delivering oil from the Gulf of Mexico to Oklahoma, in reverse. But it seems ConocoPhillips, part owner of the pipeline and of a refinery in Oklahoma, is not interested in that plan. As an integrated refiner and producer, it can take its profits either as refinery margin or producer-seller. Transporting oil out of the Midwest might leave it vulnerable to political criticism as domestic oil prices rise in response that it perhaps avoids by maintaining the status quo.
But there are lots of other arbitrage opportunities here for other players. It apparently costs $6 to ship a barrel from Cushing to the Gulf of Mexico by rail or $10 by truck, either of which sounds pretty tempting with a spread of $15 or more. More fundamentally, there are lots of other suppliers in the chain with an opportunity to make money in easier ways. Basically anybody further upstream currently feeding into the supply that ends up in Cushing has a strong incentive to be looking at a better place to sell their product, and given time, they're surely going to do just that.
And I wouldn't expect it to take a whole lot of time. Those physical adjustments are going to occur and the price differential between WTI and Brent is going to be reduced. That could take the form of an increase in the price of oil at Cushing, a decrease in the price of oil in Europe, or a combination of both.
Or, if significant threats to European oil supplies drive the price of Brent way up, the price of oil in the U.S. Midwest should increase even more than the price in Europe. And, it seems, today it did just that.
Posted by James Hamilton at February 22, 2011 09:06 PMdigg this | reddit
These divergences have taken place before, most noticeably in 2008 when Saudi crudes were trading @ a large discount to the (inflated) futures price. (This was noted in the Saudi Wikileaks cables released a couple of weeks ago.)
Also, there was a number of times when futures contracts failed to converge with spot prices in some of the ag commodities in 2008. This seems to be a function of the how the markets settle and a large overhang of paper positions.
With regard to the WTI it makes sense that the differential reflects large short positions and the CFTC stats indicate this. This could be the outcome of the large physical position in Cushing but might also represent shorts in USA hedging long positions in Brent mkts. In this way the price differential is shifted overseas to some degree and away from domestic refiners.
Even with the so- called blockage the same amount of Cushing crude is sold now as was last year and the past ten years. This means a bit of cheap(er) oil available to refiners and less $$ pressure on the gas pump price.
Posted by: steve from virginia at February 22, 2011 10:08 PM
The folks at Conoco said it would take two years to reverse the pipeline at Cushing. I find that hard to believe.
The question is, "what is the cost that the US is paying for crude?" is it WTI or is it Brent?
I think it is at least Brent at this point. Look what is being paid for spot delivery of crude at Gulf of Mexico. Yesterday is was $111 at the close. A significant premium to Brent. $20 over WTI.
Forget WTI it is a worthless indicator. Louisiana Light Crude (LLS):
Posted by: Bruce Krasting at February 23, 2011 03:26 AM
Appears to be mostly a logistics bottleneck at Cushing. I think Gail "the Actuary" Tverberg would merit noting here by name; she's done a good bit of the work with her post on the Oil Drum.
Posted by: Steven Kopits at February 23, 2011 05:31 AM
Even if all the arbitrageurs "take their marbles and go home", and even if ConocoPhillips behaves irrationally for political reasons, the prices still should converge simply because oil is fungible.
Who in the U.S. will buy imported oil if domestically produced oil is cheaper? The answer is nobody. So this sort of price differential cannot possibly persist unless U.S. oil imports fall to zero. (Otherwise somebody in the U.S. is buying expensive imported oil when cheaper domestic oil is available.)
Put another way, the question is not "Why is arbitrage not happening" or "why is Conoco importing oil from Mexico". The question is, who in the U.S. is buying imported oil when domestically produced oil is cheaper? And why?
It seems to me that the only plausible explanation is that the domestically produced oil is not truly available at the WTI price. So "a logistics bottleneck at Cushing" makes some kind of sense. Most of the other explanations I have read do not.
Posted by: Nemo at February 23, 2011 06:07 AM
Have you looked at the same trend when gold and oil took a huge dip in 2008? I have not done this, but my conjecture is you will see a similar trend. Gold declined significantly in recent months pulling oil prices with it. Recently the oil price has bounced up but so has the gold price.
Actually, as I observed a few weeks ago, the price of oil was high relative to the price of gold and so the pressure should have been down. That in fact did happen as oil declined to the mid-80s. With recent events the price of gold has started a move up and the price of oil became relatively cheap. The recent increase in oil has actually moved it to a slightly over priced condition. If gold continues to rise then the price of oil will be verified, but if gold returns to its decline the increase in the price of oil will come to a screaching halt.
Granted the middle east turmoil is rattling markets but you will find that both oil and gold will show the effect.
Posted by: Ricardo at February 23, 2011 06:32 AM
Wow, a lot of the comments here and the article itself really gives me a headache.
First of all, there is no futures/spot market convergence in the price of oil. The spot market takes its cue from the futures market, through a weighted average of the futures price.
The article addresses how one can normally just go long WTI and short Brent and profit through that convergence. This of course ignores the dynamics of what normally keeps WTI and Brent close to each other: psychology, or market manipulation by the sellers of crude oil. There really should be no convergence whatsoever between these two light crudes, BECAUSE IT IS ILLEGAL TO SHIP WTI OVERSEAS!!!
Some of you then make the value judgement that because WTI is lower in price that it is somehow "worthless" as an indicator of what the "true" value of oil should be. First of all(an aside), whose side are you on here? Why are you rooting for oil to go up? The article itself makes the implicit value judgement that they HAVE TO! converge without understanding the real mechanics behind the price differential. In other words, just because historically the price has been within a few % points of each other does not mean that it should always be a few % points from each other. Fundamental factors have affected the futures price and there is no convergence thats going to fix it.
So this differential should by all means seen as normal. Its nice to comment on it, but don't make any value judgments.
Posted by: Payam at February 23, 2011 09:19 AM
The discussion reeks of academic ideology.
Brent is not WTI. They are slightly different sulf/specific gravity of crude.
Not directly intergchangeable.
Further, the transport and storage costs are real and carry a risk.
Even further, they are just benchmarks and depending on benchmark prices, better profits may exists for refineries at different crude categories based on location and demand.
To get a real understanding of this, one would have to have a complete breakdown of the papers being traded, all storage and tanker capacity utilization and demand for each specific gravity by location of refinery.
But to say that "There is a very fundamental economic principle that goods that are close substitutes and easily transported should sell for a similar price" sound just like what it is - academic bookworm pondering about the real world, which the real world has shown to be wrong in this case.
This is not your Econ 101 lemon stand sales case.
Oil is not 100% fungible across all distances and oil compositions.
Posted by: vasra at February 23, 2011 09:24 AM
Payam: It's not necessary to export WTI, only to import less, to implement the arbitrage. You sell crude that was headed to Cushing to a domestic refiner who was importing.
vasra: 100% fungibility across all distances and oil compositions is a straw man. The question is whether there are profit opportunities when the divergence is as large as it is at the moment. I claim that there are.
Posted by: JDH at February 23, 2011 09:39 AM
This post appears to have generated a number of knowledgeable comments on the oil market. I wonder if I could pose a question to these commenters concerning the relationship between oil speculation and actual supply and demand.
Does the futures market in oil influence the actual price paid by consumers? Paul Krugman says that as long as real inventories aren't rising, then real supply and demand, not speculation, determines price for consumers.
But if people in the futures market think that price is rising, then as long as they are selling into a rising market and making money, don't those who want to actually use the stuff have to buy it at the higher price?
In the long-run the futures market may crash if price far exceeds ability of consumers to pay. Eventually someone will take a big loss and consumers will pay a lower price. But, taking housing as an example, the period during which consumers have to pay a premium could be quite long.
For any commodity, inventory size should be a function of leads and lags, responding to supply or demand. But as long as supply is limited, and demand is steady and inelastic relative to price, why should inventories rise if futures prices were influencing prices paid at the pump?
I am not asking for a judgment about whether this is currently the case, but trying to understand the relationship between futures prices and real prices, and whether there would have to be rising inventories if speculation were involved.
Any information would be appreciated!
Posted by: jcb at February 23, 2011 11:20 AM
This just goes to show you how ridiculous it is to have Cushing, OK as the delivery point for the benchmark contract. I looked into this in 2008, and many years ago it was recommended that physical settlement be allowed by tanker in the Gulf Coast (where most of the nations refineries are). Undoubtedly there are players who benefit from the current arrangement and the political will for change is not there.
Posted by: Basho at February 23, 2011 12:03 PM
It is an interesting indication of how close we are living on the edge when just the mere possibility of disruption of 2% of production can cause such a dramatic increase in the Brent price. There seems to be virtually no excess capacity available.
Posted by: Joseph at February 23, 2011 01:01 PM
Another look at oil markets from a net export point of view:
Egypt, a classic case history of a rapid net export decline, and a look at global net oil exports
Posted by: Jeffrey J. Brown at February 23, 2011 01:07 PM
jcb: I would describe it this way. Current supply/demand factors influence the current spot (real) price and the current spot price influences the futures price.
A futures contract is simply one in which person A promises to buy a good from person B at the spot price in the future. So on that date, the future price and spot price must converge. I would think of it as the spot price is pulling the future price closer to it as the expiration date nears, not the other way around. After all, today’s futures price is irrelevant to someone actually trading the good today, but today’s spot price today is very relevant to someone trading the futures contract today.
What some people have noted is that if a trader only deals in futures contracts and never holds it through the expiration date (i.e. a speculator), then he will never influence the supply/demand of the underlying good, and thus never influence the spot price.
On the other hand, if this person ever decided to hold the future contract through the expiration date, he will physically trade the good, influence the supply/demand and thus the spot price of that good. But if this person doesn’t actually want the good and only wants to sell it later at a higher price, he must store it (i.e. increase inventories). And for things like oil, this is expensive and there isn’t much evidence of it happening at a significant level.
So the short answer is that if a speculator (someone who doesn’t use the good) wants to affect the spot price he must store the good.
Posted by: Jeff at February 23, 2011 01:38 PM
JCB: Do not listen to Jeff, he is largely incorrect in his assertions. I recently wrote an article on this at naked capitalism, here it is:
That should answer all of your questions.
Posted by: Payam at February 23, 2011 02:32 PM
Continuing with net exports . . .
ANE = Available Net Exports = Net Oil Exports not consumed by Chindia
ANE were about 41 mbpd in 2005 and may be down to about 35 mbpd in 2011, probably trending down to about 27 mbpd in 2015.
So, a loss of 1.5 mbpd of global net oil exports from Libya would be on the order of 4.3% of ANE, based on the foregoing.
Posted by: Jeffrey J. Brown at February 23, 2011 02:57 PM
Maybe someone is afraid the Saudi Kingdom will fall overnight and leave a huge disparity between U.S. and world oil prices. Though short-lived, it wouldn't have to last very long to ruin most highly-levered speculators.
Posted by: don at February 23, 2011 05:24 PM
Jeff and Payem: Thanks for both your very different inputs. Haven't settled anything, but at least clarified the problem.
I would like to know, though, whether or not every futures contract has to ultimately result in a transfer of a barrel of tangible oil at the end of the contract. Or if the contract can be traded and re-traded, and extended ad infinitum. If so, it would seem to create at least the expectation of higher or lower prices, depending on the price at re-trade.
To take a related example. Today's (2/24) NYT reports:
"Oil prices jumped more than $12 a barrel this week, rising above $98 a barrel in New York on Wednesday, after the violent uprising in Libya, a major oil producer. In London, Brent crude futures rose to $111.25 a barrel."
So, both spot and futures prices rose in view of expectations that the Libyan uprising would result in lower supply. Neither demand nor supply has changed. If they move together, there must be an underlying common cause. And that cause could be either rational foresight (if you're going to take delivery and use the oil) or speculation (if not). No?
I suppose another possibility is that someone along the line of production has created a larger stockpile. But my point is that all this is based upon changes in expectation, not in ultimate supply or consumer demand.
Posted by: jcb at February 24, 2011 09:08 AM
Less than 2%(actually much less) of all futures contracts actually have physical delivery. In other words, they are just bets that are cash settled/rolled over into a new bet.
Given that the spot market uses a weighted average of futures prices you see why this can be very abusive if everyone is betting the same direction.
Posted by: Payam at February 24, 2011 10:46 AM
If you look at the futures spread, you find that it anticipates the gap shrinking at about $2.20 per month. That is the amount you would lose each month if you made a futures bet that the price gap would shrink, but it stayed the same. Looked at another way, if the gap shrank by $2.20 in one month, you would just break even on your bet. The oil futures are one-month contracts. Whether you take delivery or not, the settlement price of the futures contract at expiration is the spot price of oil (after miniscule brokers' fees).
Posted by: don at February 24, 2011 11:52 AM
Couple of thoughts here:
First, the North Sea (brent) has peaked and now it is starting to bite.
Second, it is illegal for most US crudes to be sold overseas. Its illegal, so the paper arb may work short term, but you cannot deliver the commodity from Cushing, OK to the UK or anywhere else.
Methinks the UK has a problem here. Gonna be interesting to see it unfold.
Posted by: Scott M. Koser, CFA at February 24, 2011 12:13 PM
The futures price does in fact influence the spot price: I produce oil, I will not sell today for $100/barrel if the futures price of $120 next month indicates an easy extra profit. Demand is not very elastic in the short run, so the consumer has to pay more today.
In short: The speculator doesn´t have to hoard oil in order to influence the price. The physical traders do the hoarding for him (and profit also).
Posted by: Olaf at February 24, 2011 02:05 PM
jcb: That excerpt states the the future price of oil traded in NY and London both increased because of conditions in Lybia. There are not talking about spot prices. The future prices rose on the worry that troubles in Lybia will impact supply in the future and thus the spot price will increase in the future. There is nothing inconsistent about this statement and what I said earlier.
Payam: I found that post rather unhelpful. You seem to just talk a lot about the current situation and never articulated exactly how futures prices influence spot prices. And as to your claim that 2% of all contracts are never settled, even if true that doesn't support your claim. I can open and close an oil future contract tomorrow and thus never settle the good, but that doesn't imply that I influenced the spot price in any way.
Olaf: Your example implies that oil was hoarded and there is no evidence of this happening.
Posted by: Jeff at February 24, 2011 02:52 PM
Just as a side note, I have been looking at this for a while, and tried to trade this recently using ETF's, and it's not particularly easy. The ETF USO tracks WTI and BNO tracks Brent, but liquidity in BNO is fairly low. My broker asked for a minimum $50K position plus a high borrowing rate to borrow BNO.
Posted by: IHC at February 24, 2011 05:52 PM
"BECAUSE IT IS ILLEGAL TO SHIP WTI OVERSEAS!!!"
But the U.S. is a big net importer of oil, so it doesn't need to export oil for price arbitrage to work.
Posted by: don at February 24, 2011 07:10 PM
Jeff: Where did you find "no evidence of hoarding"? U.S. Storage is at a very high level, even ships are used for it. I learnt that Producers are net sellers in the futures market. They sell for the inflated (speculators!) futures price and will not ship for a lower price. If there was just one producer, they would stop the prices from rising now and avoid another recession. But they are many, and urgently need the windfall cash to silence their uprising countrymen. So 2008 will repeat in a couple of months. Only with weaker, more indebted governments.
Posted by: Olaf at February 25, 2011 02:28 AM
1. WTI infrastructure more efficient then Brent.
2. Unrest in Libya (and, thus, contaigon stress in all of MEA)
3. Significant Domestic oil production in USA coming online.
4. Massive Domestic oil production in Canada coming online.
5. Cushing bottlenecks from sluggish US demand combined with #3 & #4.
6. CHINDIA contracts (where REAL demand growth is) priced in Brent.
This is why there is a divergence. It's a fundamental shift.
Posted by: KANE at February 25, 2011 03:30 PM
it appears consumers are being screwed again by you greedy bastards with all our elected officials grinning and counting thir money
Posted by: dennis battles at February 26, 2011 08:31 AM
Scott, do you have a source to support that it is illegal to export oil from the US? Because that sounds like total bullshit to me. It is certainly bullshit for Mexico and Canada, where free oil markets are written into NAFTA. And it is almost certainly going to bullshit almost everywhere else, by the simple fact that America depends more from a free market than anywhere else, and so should be its biggest supporter.
And indeed, a brief google search led me to ">this site which shows that the US exports significant quantities of mostly refined petroleum to countries in NAFTA and beyond, a surprising 2 million barrels per day in 2009.
Posted by: Alex at February 26, 2011 03:47 PM
Sorry, the link got mangled somehow. Here it is again:
Posted by: Alex at February 26, 2011 10:14 PM