Thursday, April 10, 2008

Market Memory

Sometimes I wonder if the greatest flaw in the global oil market is that it has such a precise memory. Anyone with an internet connection can see what the price of oil was in New York fifteen minutes ago, and then view another website and track its movements all the way back to a Wednesday in March twenty-five years ago when the NYMEX West Texas Crude Oil (WTI) contract began trading. The problem with this is that, because the market can't forget, and because each moment's price is set with reference to the previous price, there is little opportunity for the market to catch its breath, weigh all of the fundamentals, and arrive at today's price from scratch. As a result, any distortions that creep in take a while to be expunged, if they ever truly are.

Most days I am grateful that I am no longer an oil trader, with one eye always glued to a screen displaying the gyrations of the global energy commodity exchanges. It's not good for one's health, even if you have a calm disposition. That's especially true these days, when a single news item can send the entire market up or down by as much as $4 per barrel from a starting point over $100/bbl. The process was different when I traded crude oil on the West Coast in the late 1980s. Although the NYMEX contract was becoming a bigger factor in the pricing of physical grades of crude oil, such as the cargoes of Alaskan North Slope crude I bought for Texaco's Los Angeles and Anacortes, WA refineries, it was still quite common to buy and sell pipeline quantities of oil at a premium or discount to posted prices--periodically-updated fixed prices at which refiners or traders solicited producers to sell them oil--or as often as not, at a fixed price unique for that deal on that day, e.g., $17.21/bbl for 2,000 bbl/day of Buena Vista Light during April 1987. (It's getting hard to believe oil was ever that cheap.)

Because of the way such deals were struck, a trader and the refinery for which he was buying had to have a very clear sense of the intrinsic value of that oil, either in terms of the products into which it could be refined, or the price at which it could be resold before delivery, if requirements changed. This process involved significant risks that could not easily be hedged, then. The stakes were high enough that, unless the counterparty was a long-term, reliable supplier or customer, deals could fall apart over a difference of 5 cents per bbl. Without suggesting that traders today are any less diligent or astute, I believe a market in which most prices are set with reference to WTI, Brent, or some other ultra-transparent exchange-traded marker entails less accountability for ensuring that the price involved is reasonable, rather than defensible--i.e., "Well, I paid the market price for it."

In a popular film of a few years ago, "Memento," the protagonist had an unusual form of amnesia and woke up each morning without any clear recollection of the previous day. He had to rely on notes he had previously scribbled to himself. What if the oil market worked that way, and each day, traders had to re-establish the price of oil from scratch, relying only on the fundamentals of supply, demand and inventory? A classical economist might suggest that the result would be no different, because the market price is merely the level at which supply and demand are balanced, every day. I'm less sure things are that simple, and I suspect that the practitioners of behavioral economics might be skeptical, as well. For example, yesterday's increase in the price of WTI to $111/bbl in response to an unanticipated 3 million bbl drop in US crude inventories only makes sense if you accept that $108/bbl accurately reflected all of the market factors before that news. However, a similar overall configuration of global inventory and spare production capacity in 2005 yielded prices in the mid-$50s to mid-$60s. $111 makes more sense as the net sum of three years of individual price movements, than as the bottom-up evaluation of all the factors in today's market.

I recently received a copy of an academic paper by a Ph.D. candidate at the University of Michigan and his professor. It tackles the question of whether the oil futures market provides a reliable forecast of future oil prices and finds that it does not. Even my view that it is a good indicator of current expectations of future prices seems shaky, in their analysis. Taken together with my concern that the market may be influenced more by its own price history than it ought to be, I conclude that decision makers, policy makers, and consumers would be well-served to take a somewhat more jaundiced view of that daily WTI settlement price that the media has grown so fond of displaying. Its impact on fuel prices and on our trade deficit is certainly real and tangible, but it might not be telling us as much about the world and the future as we have come to believe.

No comments: