It's a perennial question attached to a conspiracy theory. Does the United States reap some great advantage owing to the fact that oil is priced and traded in dollars?

The naive conspiracy theory is concisely described by Dave Chiang (commenting on Brad Setser's blog):

World trade is now a game in which the US produces dollars and the rest of the world produces things that dollars can buy. The world's interlinked economies no longer trade to capture a comparative advantage; they compete in exports to capture needed dollars to service dollar-denominated foreign debts and to accumulate dollar reserves to sustain the exchange value of their domestic currencies.

This phenomenon is known as US dollar hegemony, which is created by the geopolitically constructed peculiarity that critical commodities, most notably oil, are denominated in dollars. Everyone accepts dollars because dollars can buy oil. The recycling of petro-dollars is the price the US has extracted from oil-producing countries for US tolerance of the oil-exporting cartel since 1973. By definition, dollar reserves must be invested in US assets, creating a capital-accounts surplus for the US economy.

The standard rejoinder, described recently by Steven Kyle (and with which I mostly agree), is that it absolutely doesn't matter what currency a commodity is priced in, so long as there are liquid FX markets. If the US prints dollars, that doesn't result in creating new buying power for oil as the dollar hegemonists assert. The newly printed greenbacks simply reduce the value of a dollar in terms of oil and other currencies, driving up the dollar price of oil, but not its effective price in euro or yen. Rational third-parties with a need to purchase oil or commodities would not accumulate dollars under these circumstances. They would accumulate assets expected to hold their value, and purchase oil by converting to dollars transiently on an as-needed basis.

Brad Setser recently took up the question. While not disagreeing with the standard argument, his take is that it might, in fact, matter a bit that oil is priced in dollars, if one considers that oil exporting nations peg their currencies to the dollar. It is dollars that are wired into oil exporting countries; policymakers would have to proactively trade to convert those dollars into some other currency or commodity. They could sell those dollars for euros, for example. But that would weaken the dollar against the euro, and by extension the buying power of their own currencies for the European goods they consume. So they don't, and hold more dollars than they might have, if they hadn't been paid in dollars.

Brad's argument is subtle and interesting, though it's worth pointing out that it is fundamentally a political or behavioral-finance explanation rather than one based on rational actors. Dollar-pegged oil exporters who wish to maintain the buying power of the dollar against the euro ought to be indifferent to the currency for which their oil is initially sold. Receiving euros and selling hem for dollars would be equivalent to receiving dollars in the first place and just holding them. Brad's argument hinges on status quo bias — that oil importers are willing to enhance the buying power of their currencies by not acting in ways that they would not if the same policy required affirmative market interventions.

There is another subtle reason why I think it does matter, some, that oil is priced in dollars: Firms who hedge oil price risk in public markets are required, by the fact that the contracts are written in dollars, to take on USD currency risk, which they may hedge by accumulating dollars. Consider, as an example, a medium-sized European firm in an energy-intensive manufacturing industry. The firm wishes to take on no speculative position with respect to the future price of oil, but it does wish to plan for profitable operations over the next year. To avoid the risk that a spike in the price of oil would send it into the red, this sort of firm is likely to purchase oil forward, using public futures markets, effectively locking in a known price today for the coming year's energy needs. Since the futures contracts are USD denominated, though, the firm has locked in a price in dollars, although its accounts and expenses are in euro. The firm has exchanged oil price risk for USD price risk. It must now hedge the latter, either by purchasing and holding sufficient dollars to cover the USD prices it has locked in, or by purchasing USD/EUR futures (which only shifts the USD/EUR price risk to some other party, who will need to hedge by holding dollars).

So, in fact, I think the dollar hegemonists have a bit (but just a bit) of a point with respect the the denomination of oil and other commodity contracts. To the degree that commodities are purchased forward by actors primarily concerned with hedging risk (rather than maximizing return via speculation), the currency in which futures contracts are traded determines the currency they will have to accumulate to fund their purchases.

But I don't think this argument holds much water as an explanation for central bank USD reserve growth. It only makes sense to the degree that entities have binding forward obligations in the currencies they are accumulating. Even if one considers nations as consolidated entities, and let central banks implicitly hedge the risks of private commodity consumers, I'm pretty sure that the scale of emerging-market reserve accumulation dramatically outstrips any plausible estimate of forward purchases. (If anyone has decent data, country-by-country foreign purchases of USD-denominated commodities, it would be fun to run some regressions, though.)