The Fed, the dollar and oil prices

With the Fed signalling a tightening bias, a variety of arguments for Fed tightening are gaining credence. One hot story is that the Fed will hike rates in an attempt to push down oil prices. The argument is that easy Fed policy has caused a weak dollar and because oil producers are paid in dollars a weak dollar puts upward pressure on the price of oil. On close inspection, this argument is being greatly exaggerated.
Is Fed easing the cause of the weak dollar?
Fed easing has played some role in the drop in the dollar, but it is not the main story. The dollar has been falling since 2002 and most of this trend decline is due to a chronic problem: the unsustainably high trade and current account deficits.
Fed policy has created cycles around this declining trend: tightening cycles have tended to slow the dollar decline and easing cycles have speeded it up. Recent aggressive Fed easing appears to have accelerated the drop, but even here it is important to distinguish between the proximate cause of the dollar weakness and the underlying cause.
Absent aggressive Fed action the credit crunch would likely have gotten a lot worse, driving the economy into a major recession. This is hardly a dollar-friendly scenario: indeed we believe if the Fed had not eased, the dollar would have probably fallen even more. Ironically, this is one of the dangers of effective policy: by forestalling disaster, it makes the policy look unnecessary to policy critics.
Is the weak dollar driving up oil prices?
The op ed page of the Wall Street Journal (WSJ) argues for a strong link between the dollar and oil prices on almost a weekly basis. Consider their piece on May 23rd, “Oil and the Fed.” The piece correctly points out that “the price of oil has risen far more rapidly in dollars than it has in euros since 2002.” It then argues that because oil is denominated in dollars, when the dollar falls, oil sellers will demand more for their product. They conclude that if “the Fed merely kept the dollar stable against the euro, the dollar price of oil would be closer to $80 than to $131 a barrel.”
This sounds plausible: in equilibrium oil producers should be compensated for the declining value of the dollars they are earning. However, the WSJ math is only correct if the oil producers buy all of their imports from the euro area. Taking the other extreme, if they buy all their products from dollar block countries, then changes in the dollar do not affect their purchasing power.
More realistically, the equilibrium price of oil is determined not by one currency, but by a weighted average exchange rate of the oil producer’s trading partners. Since the euro is only part of this basket and it is one of the strongest currencies in the world over this period, using the WSJ methodology (and the correct exchange rate) implies that even with a flat dollar-euro exchange rate the price of oil would be much closer to $131 than $80.
Using the corrected WSJ methodology the recent run-up in oil prices is even harder to explain. How is it that the 5% or so rise in the Euro this year is causing a 33% jump in the dollar price of oil? At most, the dollar weakness should explain a few dollars’ increase in oil prices. Perhaps the weak dollar has become an excuse to bid up oil prices, causing a huge multiplier effect on the oil market? And couldn’t the surge in oil prices boost expected inflation, adding to demand for oil as an inflation hedge? This kind of thinking may be driving the market, but both arguments sound like a bubble in the making rather than a rational market.
Is there a direct link between Fed policy and oil prices?
What if we cut out the middleman—the dollar—and look at the direct link between easy Fed policy and oil prices? Figure 1 reproduces a chart from another recent WSJ op ed, (“That Stagflation Show”, June 9, 2008) showing a strong negative relationship between the funds rate and oil prices. The correlation between the two series over this period is roughly -90%. Of course, it is dangerous to infer causation from such a short time frame. For example, the same period has seen a steady drop in payroll employment, a sharp drop in consumer confidence and an accelerating decline in home prices, but that does not mean these variables are causing higher oil prices.
If this is a meaningful relationship, it should hold in other periods. Figure 2 widens the view to cover the last 25 years. The correlation coefficient drops to about -25%. Perhaps the relationship is episodic? In the last easing cycle oil prices fell during the first 475 bp of Fed easing, but rose with the last 75 bp of easing (and have continued to rise almost continuously ever since).
We have estimated rolling 24 month correlations and find several episodes of strong negative correlation—1996-97 and 2003—but these are very different periods in terms of the ease or tightness of monetary policy. We also test to see if periods of unusually easy monetary policy cause surges in oil prices. Specifically we correlate residuals from a Taylor Rule model of Fed policy with the growth in oil prices. We find a -15% correlation over the 1987-2008 period. In other words, unusually big Fed easings or tightenings do seem to affect oil prices.
The Fed is not the root of all evil
Recent developments have put the Bernanke-led Fed in a very rough spot. Not only is the Fed battling a credit and housing crisis, but now it is dealing with a growing commodity shock. Central banks are poorly equipped to deal with commodity shocks: they can’t simultaneously ease to fight the downside shock to growth and tighten to fight the upside shock to inflation. Moreover, as we have seen, the Fed should neither be blamed for the surge in oil prices nor expected to stop the surge by hiking rates.