Dispellng Economic Myths

[J Larsen]

research has shown that the coefficient of correlation is - 0.7 and not - 1.0 , meaning that the dollar price of oil needn't increase at all in such a case , though more often than not it does .

But even in a case where the dollar depreciates against the euro and the dollar price of oil does not increase, it is not true that oil exporters would have been better off if they had priced their product in euros. Commodity prices are set entirely through supply and demand; the currency used for pricing is irrelevant. I thought that was the whole point of the argument.

[Chas]

Commodity prices are set entirely through supply and demand; the currency used for pricing is irrelevant. I thought that was the whole point of the argument.

If the currency used for pricing were irrelevant , how then to explain the - 0.7 correlation between movements in the dollar/euro cross and the dollar price of oil ? When you consider that almost all of the oil exporters' exports are priced in dollars , while a great deal of their imports are priced in euros , and that they have an understandable desire to diversify their investments , the relevancy of the currency used to price oil becomes apparent . Recall that Guy's initial claim was that at present , the use of dollar-pricing for oil does not hurt oil exporters . I pointed out , contrary to Guy's assertion , that this is a contingent fact , not a necessary consequence of the very meaning of dollar depreciation .

[Guy Berger]

Commodity prices are set entirely through supply and demand; the currency used for pricing is irrelevant. I thought that was the whole point of the argument.

If the currency used for pricing were irrelevant , how then to explain the - 0.7 correlation between movements in the dollar/euro cross and the dollar price of oil ?

I'm curious -- what is the source for that -0.7 number? I assume that is not a simple correlation (which would suffer from serious omitted variable bias), but would like to make sure.

Regardless, as I conceded, the relevant relationship is between the ratio of dollar/euro prices of oil (not the actual dollar price) and the dollar/euro exchange rate. That would certainly be consistent with the correlation you mention -- for example, if the dollar depreciates by 20% against the euro, the dollar price of oil goes up by less than 20%, and the euro price of oil falls to make up the difference. None of this changes my main point -- that there is little to be gained by pricing oil in another highly liquid currency.

As Jan asked -- are you seriously arguing that exchange rate movements generate large unexploited profit opportunities by buying oil in one currency and selling it in another?

Guy

[Chas]

I'm curious -- what is the source for that -0.7 number? I assume that is not a simple correlation (which would suffer from serious omitted variable bias), but would like to make sure.

Don't have the primary source , so can't vouch for the methodology , but the correlation is mentioned here and here . Whatever the numerical value of the correlation actually is , the important point is that it is not - 1.0 , which is what your initial assertion , "a 20% depreciation of the dollar against the euro will cause the price of oil to go up by 20% in dollars and remain unchanged in euros " implied .

Regardless, as I conceded, the relevant relationship is between the ratio of dollar/euro prices of oil (not the actual dollar price) and the dollar/euro exchange rate. That would certainly be consistent with the correlation you mention -- for example, if the dollar depreciates by 20% against the euro, the dollar price of oil goes up by less than 20%, and the euro price of oil falls to make up the difference. None of this changes my main point -- that there is little to be gained by pricing oil in another highly liquid currency.

As I said , given present conditions I agree with your point , but not with your reasoning in support of it - reasoning which if it were sound , would rule out any contingency in your statement .

As Jan asked -- are you seriously arguing that exchange rate movements generate large unexploited profit opportunities by buying oil in one currency and selling it in another?

Of course not ; nothing I wrote supports such an imputation in my view .

[J Larsen]

I would certainly expect the coefficient of correlation between the dollar/euro exchange rate to be less than unity in abs value - this is just a reflection of the fact that exchange rates are not the only factor in determining the dollar price of oil. This observation, which I frankly consider to be trivial, does nothing to refute the claim that the fortunes of oil exporting countries is completly indepenant of the currency in which oil is priced. If I wanted to get really nitpicky, I'd go so far as to assert that if anything, a weakned dollar leads to (very slightly) increased efficiency in oil markets, because the measurement unit is more fine.

[Jazzmoose]

So. Uh. How 'bout those Mets?

[Guy Berger]

I'm curious -- what is the source for that -0.7 number? I assume that is not a simple correlation (which would suffer from serious omitted variable bias), but would like to make sure.

Don't have the primary source , so can't vouch for the methodology , but the correlation is mentioned here and here . Whatever the numerical value of the correlation actually is , the important point is that it is not - 1.0 , which is what your initial assertion , "a 20% depreciation of the dollar against the euro will cause the price of oil to go up by 20% in dollars and remain unchanged in euros " implied .

Regardless, as I conceded, the relevant relationship is between the ratio of dollar/euro prices of oil (not the actual dollar price) and the dollar/euro exchange rate. That would certainly be consistent with the correlation you mention -- for example, if the dollar depreciates by 20% against the euro, the dollar price of oil goes up by less than 20%, and the euro price of oil falls to make up the difference. None of this changes my main point -- that there is little to be gained by pricing oil in another highly liquid currency.

As I said , given present conditions I agree with your point , but not with your reasoning in support of it - reasoning which if it were sound , would rule out any contingency in your statement .

What conditions would have to change to make my point invalid? You say my argument is contingent -- contingent on what? (Let's leave out the following: exchange rate or commodity markets collapse; either the dollar or the euro becomes a highly illiquid currency).

As Jan asked -- are you seriously arguing that exchange rate movements generate large unexploited profit opportunities by buying oil in one currency and selling it in another?

Of course not ; nothing I wrote supports such an imputation in my view .

OK. "Exchange rate movements don't generate large unexploited profit opportunities by buying oil in one currency and selling it another" implies "it doesn't make a difference to oil producers what currency oil is priced in".

[J Larsen]

Let's leave out the following: exchange rate or commodity markets collapse...

Why? It seems to me that both of those scenarios will have equal implications for oil producers regardless of the currency used for pricing. The only exception I see is if the worth of the dollar collapsed to zero, in which case it could no longer be used to measure anything. Also, don't we only have to consider cases where the liquidity of the dollar and/or euro decreases seperately if we are assuming that the transactions are actually done in the currency in which oil is priced?

[Chas]

What conditions would have to change to make my point invalid? You say my argument is contingent -- contingent on what? (Let's leave out the following: exchange rate or commodity markets collapse; either the dollar or the euro becomes a highly illiquid currency).

OK. "Exchange rate movements don't generate large unexploited profit opportunities by buying oil in one currency and selling it another" implies "it doesn't make a difference to oil producers what currency oil is priced in".

How do you reckon that the absence of arbitrage opportunities implies that " it doesn't make a difference to oil producers what currency oil is priced in " ? Assume that at some time in the future supply and demand for oil are in long-term balance and that the price of oil is thus in long-term equilibrium . Assume also that the dollar continues its secular decline against the euro . Given the increasing propensity of the oil exporters to spend their oil receipts on imports ( rather than adding to their already enormous dollar reserves or further unbalancing their investment portfolios ) , and given that the eurozone ( and other non-dollar trading partners ) is the source of most of those imports , wouldn't the oil exporters want to maintain their purchasing power by compensatory increases in the dollar price of oil , perhaps through production cuts ? In addition , oil exporters who end their currencies' dollar-peg would see rising production costs that could exceed growth in oil revenues , adding additional impetus to any attempt to engineer a rise in the dollar price of oil . It is in this sense that , " it doesn't make a difference to oil producers what currency oil is priced in " , is contingent . Your reasoning in support of that conclusion made it sound like an analytic truth , and that is the gravamen of my objection .

[J Larsen]

Assume that at some time in the future supply and demand for oil are in long-term balance and that the price of oil is thus in long-term equilibrium . Assume also that the dollar continues its secular decline against the euro . Given the increasing propensity of the oil exporters to spend their oil receipts on imports ( rather than adding to their already enormous dollar reserves or further unbalancing their investment portfolios ) , and given that the eurozone ( and other non-dollar trading partners ) is the source of most of those imports , wouldn't the oil exporters want to maintain their purchasing power by compensatory increases in the dollar price of oil , perhaps through production cuts ?

If the dollar is falling against the euro, and oil exporters price oil such as to maintain their purchasing power in countries using the euro, then the dollar price of oil would be expected to rise regardless of whether oil is priced in euros or dollars.

[Chas]

If the dollar is falling against the euro, and oil exporters price oil such as to maintain their purchasing power in countries using the euro, then the dollar price of oil would be expected to rise regardless of whether oil is priced in euros or dollars.

Yes , but in the counterfactual I outlined , the rise in the dollar price of oil is engineered by the oil exporters , whereas if oil were priced in euros , the rise in the dollar price of oil would be a direct consequence of the depreciation of the dollar against the euro . The imprecise nature of the former method of adjustment is the reason why Guy's assertion that a " a 20% depreciation of the dollar against the euro will cause the price of oil to go up by 20% in dollars and remain unchanged in euros" would only be true if oil were priced in euros , which it isn't . The main point is that there exist scenarios wherein oil exporters seeking to maintain their purchasing power and diversify their assets have reason to prefer oil priced in euros to oil priced in dollars given the additional efforts needed to achieve those ends when oil is priced in depreciating dollars . The fact that two-thirds of foreign reserves are in dollars , and that any move away from the dollar as a medium of international exchange would undermine its function as a store of value , makes this truly a 'thought experiment' more than anything .

[J Larsen]

So if I understand you, you are only disputing the use of the word "cause". You are not arguing that the economic consequencies would be any different under a different pricing convention, you are merely arguing that the underlying processes that lead to the observed prices would be different. Am I understanding you?

BTW, even if I understand, I'm not completely sure I agree.

Edited December 3, 2007 by J Larsen