Is there such a thing as a currency-neutral measure of prices?

November 7, 2007 10:03 AM

Is there such a thing as a currency-neutral measure of the price of a commodity?

The price of oil has gone way up in dollars. But the dollar is also way down against other world currencies. I'm looking for a graph that shows the price of oil in a way that's currency neutral. Ie, a graph that filters out the fluctuations in the US dollar. I'm not just interested in oil, really what I want is to translate dollars to "world currency" for a variety of things.

Naively I think you could do this by creating a virtual currency, an average of various world currencies weighted by the currency's economic influence. Ie: "30% USD, 20% EUR, 20% CNY, ..." But I'm not an economist and I'm not sure what I'm asking for even makes sense.

I did find a set of graphs of oil prices in various currencies, including vs. gold. The comments include a discussion about exactly what I'm asking for, with inconclusive results. The most interesting was a suggestion to look at oil vs. the CRB index. I don't really understand the CRB index, but what I do understand suggests it's not the right thing to normalize against.

posted by Nelson to work & money (15 comments total)

Various people have done that; the usual way is to use "constant dollars" e.g. (constant 1970 dollars).
posted by Steven C. Den Beste at 10:08 AM on November 7, 2007

Big Mac Index?
posted by Comrade_robot at 10:35 AM on November 7, 2007

I think a constant dollar measurement would work best. The Big Mac Index is a rough guide to purchasing power parity across countries and only a year or two old, though it's nice to see The Economist break it our every couple of months.
posted by yerfatma at 10:44 AM on November 7, 2007

There are several ways to do this that come to mind. One would be to use constant dollars, in the way Steven suggests (you basically use a particular year's dollar, so you control for variations in its buying power over the same period).

You could also look at one commodity compared to another commodity that's held in high regard; e.g. oil versus gold. However that ignores fluctuations in the buying power of an ounce of gold, which certainly isn't always constant (although I think it tends to be pretty stable). What might be better is to use a basket of commodities, say the ones used to establish the CPI.

What I'm not sure of is, because the CPI is used to index inflation (at least, I think it is?), would you get the same results by using the CPI as a benchmark directly, that you'd get if you used 'constant dollars'? Not sure if one way or the other would be advantageous.
posted by Kadin2048 at 10:45 AM on November 7, 2007

The iPod index ?
posted by iviken at 10:51 AM on November 7, 2007

Thanks for the suggestion of constant dollars. That normalizes for inflation, but does it control for fluctuations of one currency against another's?

The iPod and Big Mac indices are cute, but I'm looking for something more academic.

I'm realizing the underlying question I'm asking here is the core thorny economic issue of what "value" is. I don't want to answer that hard question, I just want something simple that's not entirely US dollar biased.
posted by Nelson at 11:11 AM on November 7, 2007

Using "constant dollars" doesn't achieve your goal: it only compensates for inflation, not changes in the value of USD versus other currencies (not to say there aren't links between inflation rates and exchange rates, but they aren't going to cancel each other out).

The simplest way of doing this would probably be to compare against a basket of major currencies (weighted by GDP perhaps). Comparing against the CRB isn't unreasonable either (though it would be better take take the energy components out).
posted by ssg at 11:16 AM on November 7, 2007

look for "trade-weighted" exchange rates. This section of the Federal Reserve website shows a few indexes that show the USD exchange rate against a basket of other currencies, with each currency's weight in the basket being determined by its trading volume/value with the USA. This is a price-adjusted trade-weighted USD exchange rate index.
posted by patricio at 11:34 AM on November 7, 2007

Using "constant dollars" doesn't achieve your goal: it only compensates for inflation, not changes in the value of USD versus other currencies

I'm not disagreeing, just asking because I'm clearly missing something: given a "constant dollar" is a snapshot of a currency at an instant in time, doesn't that make relative fluctuations irrelevant? As long as you stop time for the price of the currency, you should have what the poster is looking for, as far as I can see.
posted by yerfatma at 12:24 PM on November 7, 2007

yerfatma: You can't actually buy or sell a barrel of oil in "constant dollars", because there isn't a unit of currency called "constant dollars" (hence the scare quotes). You can adjust the price paid in current USD for inflation back to 1970 or whenever you like, but that doesn't account for changes in the value of the dollar relative to other currencies.

For example, say for whatever reason the value of the USD goes down 10% relative to every other major currency over a year. If we assume that the price of oil stays constant in all those currencies (it would actually go down slightly in the short term), then oil would be 11% more expensive in USD. In the same year, say that inflation was 5% in the USA. Even if you adjust the value of the current dollar to the previous year's dollar, you are still going to find that oil costs ~6% more in "constant dollars". However, in all those other major currencies, the price of oil has stayed the same. Using "constant dollars" leads you to the conclusion that oil is more valuable than it was a year ago, even though that is not the case.

Of course, inflation is going to increase as the dollar goes down in value because the cost of imported goods will increase, but unless there are absolutely no barriers to trade, inflation and exchange rate are not going to compensate exactly for each other.
posted by ssg at 1:03 PM on November 7, 2007

I think what you want is Special Drawing Rights, a potential claim on the freely usable currencies of International Monetary Fund members. The SDR is made up of a composite of the values of major currencies.
posted by happyturtle at 1:14 PM on November 7, 2007

You can't actually buy or sell a barrel of oil in "constant dollars", because there isn't a unit of currency called "constant dollars" (hence the scare quotes).

I follow you. The poster was asking for "a graph that shows the price of oil in a way that's currency neutral". I guess I took that to mean, "How would I see fluctuations in the price of oil without the noise of inflation, etc." But yeah, to make it truly currency-neutral, you'd need some sort of basket of goods, but most any good would be affected by the price of oil, so it becomes a snake that eats itself.
posted by yerfatma at 1:33 PM on November 7, 2007

Today's post on Econbrowser ('Does Dollar Weakness 'Cause' High Oil Prices?') is highly relevant.
posted by Aloysius Bear at 1:36 PM on November 7, 2007

Naively I think you could do this by creating a virtual currency, an average of various world currencies weighted by the currency's economic influence. Ie: "30% USD, 20% EUR, 20% CNY, ..." But I'm not an economist and I'm not sure what I'm asking for even makes sense.

This is exactly what the International Monetary Fund does for Special Drawing Rights. The current configuration is USD 44%, EUR 34%, JPY 11%, GBP 11%. This is next due for re-evaluation in 2011.
posted by happyturtle at 2:00 PM on November 7, 2007

Thanks for the pointers. Patricio's link to trade-weighted exchange rates was terrific. There's an accompanying paper that describes in detail the construction of the kind of index I'm looking for. You can see the relative weighting in this table. It's US-centric in that it's based on US trade; for instance, Mexico figures quite highly.

Special Drawing Rights are also helpful, and significantly simpler. It's not quite as precise and seems to exist to solve a legislative problem as much as to satisfy economics inquiries. Compared to the trade-weighted index, it's missing China and Mexico entirely as well as a significant contribution from Pacific Rim countries.
posted by Nelson at 10:49 AM on November 8, 2007