Cause and effect?
Martin Feldstein perhaps explains the unexpected decline of oil inventories today (the situation is a little absurd because although crude inventories fell, refined product inventories surprisingly rose and overall demand continues to fall):
when will an owner of oil reduce production or increase inventories instead of selling his oil and converting the proceeds into investible cash? A simplified answer is that he will keep the oil in the ground if its price is expected to rise faster than the interest rate that could be earned on the money obtained from selling the oil. The actual price of oil may rise faster or slower than is expected, but the decision to sell (or hold) the oil depends on the expected price rise.
There are of course considerations of risk, and of the impact of price changes on long-term consumer behavior, that complicate the oil owner’s decision – and therefore the behavior of prices. The Organization of Petroleum Exporting Countries (the OPEC cartel), with its strong pricing power, still plays a role. But the fundamental insight is that owners of oil will adjust their production and inventories until the price of oil is expected to rise at the rate of interest, appropriately adjusted for risk. If the price of oil is expected to rise faster, they’ll keep the oil in the ground.
The NYT’s story on today’s stock market decline is consistent with Feldstein’s explanation: “Oil surged to new records above $144 a barrel as the government reported a bigger-than-expected drop in U.S. supplies and as investors worried about tensions in the Middle East.”
In the next sentence, the Times said something that implies that the current trend can’t last forever: “Worries that GM could go so far as to declare bankruptcy only added to investors’ unease. The stock fell $1.77, or 15 percent, to $9.98 — the first close below $10 since September 1954 when Dwight Eisenhower was president.” At some point, rising unemployment, sparse sales of large autos, and changes in consumption will have their effect. Some observers predict a moment like 1987 when those expectations turn — it can’t happen soon enough.