Dissecting the Oil-Dollar Affair

Don’t we all have a penchant for drawing inferences from patterns! Put that together with a fascination with gossip, and you get all sorts of (wild) stories—great for entertainment’s sake but potentially dangerous if they lead to, say, a suicide attempt or… the wrong investment idea!

Take for example the rumor that Tom Cruise has a crush on… Will Smith . Apparently he has been following him “everywhere” recently... and, let’s face it, Smith is such a charming guy! A pretty harmless gossip, you might think, though not if it were to drive Cruise’s official other-half, Katie Holmes, to a suicide attempt after spending sleepless nights reassessing her skills as a wife.

So here is the latest in gossipville.. Dollar and Oil are having an affair! Yeap, the two have been flirting for years now, but they say this time it looks serious. They have been spotted together way too many times… and their (cor)relation seems much, much stronger than before! Each time the dollar weakens, oil strengthens—and vice versa.

What’s going on?

I decided to do a bit of a detective’s work to find out for myself. No, I’m not jealous, this is not personal! It’s about setting the record straight.

Who is after who? Generally, you would expect the dollar to respond negatively to an increase in the price of oil. All else equal, higher oil prices make America’s imports more expensive. So the trade balance worsens, aggregate demand falls, interest rates are cut to kickstart the economy, returns on dollar assets fall, dollar loses its relative appeal and, long story short, it depreciates. That’s how a higher oil price might bring about a weaker dollar.

Another version looks at the other side of the coin: The oil exporters. Since the barrels of oil they export are priced in US dollars, every time the dollar depreciates the price of their exports goes down, all else equal: The dollars they receive for their barrels of oil will buy fewer Gucci bags, since the latter are priced in euros. So, what can they do?

Well, they can either cut their oil supply, in which case prices will go up, re-establishing their buying power; or they can just raise the price of a barrel of oil outright. If the dollar falls, say, 20% against the euro, they could set the price of a barrel by about as much in dollars, to make sure they can buy as many Gucci bags as before. So that’s how a weaker dollar might bring about a higher oil price.

Both stories are sound in principle (subject to caveats), yet what disturbs me is the following: Both are as much true now as they have been for the past few decades! So why is everyone talking about a stronger dollar-oil correlation NOW? Has anything changed?

So they say. So let’s examine the recriminating evidence.

Gucci or Krispy Kreme? One argument has to do with the fact that America’s exports are not as sought after as before among oil exporters: Back in the early 1980s, for example, the US accounted for around 15 percent of the total imports of oil exporters. Whereas now this has dropped to barely 9 percent. What’s this got to do with the dollar-oil affair?

The idea is the following: As the oil price increases, oil exporters become richer since they receive more money for their barrels. Yet, they don’t seem to be using their new wealth to buy more Heinz ketchup and Krispy Kreme’s. Instead, they’re going after more and more Chinese textiles, Korean semi-conductors and some of those Gucci bags—among other things.

The result would seem to be that, as the price of oil goes up, America’s trade balance deteriorates faster than before: US (oil) imports are more and more expensive while US exports remain stagnant (I repeat, all else equal).

Indeed, this deterioration is much worse that in most other advanced economies: America’s deficit with oil exporters rose by almost $60 billion between 2003 and 2007. This compares with just $8 billion for the Eurozone. And that’s why the dollar should be falling (e.g. versus the euro) these days faster than previous years.

Perhaps… but not necessarily. As we have already seen, oil exporters are becoming a lot richer—as in "530 billion dollars richer" in export revenues in 2007 compared to 2003. Sure enough, they’ve been spending more on imports. But even after you take these into account, you’re still left with an increase in their trade surplus of around $140 billion in 2007 compared to four years before. Right, that’s (a whopping) $140 billion of pocket money! What do they do with it?

Citigroup or Ferrari? Here is the catch. You see, America doesn’t just export movies, ketchup and donuts. It also “exports” investment assets: Like government debt securities (quite a lot of those); shares or bonds of American companies (like Citigroup); debt of government-sponsored enterprises (such as Fannie and Freddie); real estate (like condos in Miami); what have you.

And you might be surprised to hear that oil exporters still fancy America’s financial assets quite a bit, despite their growing intentions to diversify…and despite the shameless collapse of stocks like Citigroup, Lehmans & co. Indeed, according to the latest quarterly report by the Bank of International Settlements (BIS), also summarized here by RGE, oil exporters increased the dollar assets they hold in BIS-reporting banks by record amounts during the first three months of 2008.

What does this mean? It means that whatever deficit America had with the oil exporters continues to be more than offset by flows from these countries back into the US. Ergo, the argument of higher oil-->higher US trade deficit-->higher dollar-oil correlation lacks sufficient corroborating evidence.

Keeping up with Gucci: What about the purchasing power argument? Let's see... If oil exporters had as big a bargaining power as to raise oil prices every time the dollar fell and their (non-dollar denominated) imports became more expensive, why didn’t they do it before? Like, when a barrel of oil fetched just 20 bucks?

More generally, you would have seen oil prices moving up every time oil exporters faced an adverse price shock on their imports, so as to maintain their purchasing power. In other words, you would have seen fairly constant terms of trade (ratio of export to import prices) in these countries. But you don’t see that; oil exporters’ terms of trade have been fluctuating a lot, and very much in line with… oil prices.

The “photos”: Best bit last.. Actually, there is not even a “correlation” to gossip about. Take a look at some numbers: The dollar appreciated during both April and May (versus the euro), yet oil prices continue to rise unabated. This is the opposite direction to the one all those gossipmongers are talking about (which is weak dollar-->strong oil). Sure, perhaps if you look at weekly data, or maybe daily or hourly data, you might start seeing "correlations" that last for… a few days. But, I'd rather call that a "fling," if not "data mining" (the economists' equivalent of "gossip fishing").

Where does this leave us? Yes, dollar and oil have a “thing” going on, for years now. Yes, that “thing” tends to be in the direction of strong-oil-weak-dollar (and vice versa). Yes, had the dollar not weakened (because of the abysmal state of the US economy), oil prices might have been somewhat lower (1/).

But no, there is no compelling reason to believe the correlation has recently increased. Their "status" is at best a fling, and certainly nothing I would bet my money on.

Glossary: correlation, terms of trade, all else equal, data mining, fling.

1/ I feel I should reference here studies by the IMF (in its April WEO report, Chapter 1, Box 1.4) and the Dallas Fed, which have tried to estimate the impact of the dollar’s depreciation on oil prices... Even though I actually find their methodologies lacking in a number of ways that probably deserve a whole other article.