Stocks and oil prices

Jan. 31, 2000
An easily overlooked number in this week's Forecast & Review special report raises important questions for the oil and gas industry (p. 44).

An easily overlooked number in this week's Forecast & Review special report raises important questions for the oil and gas industry (p. 44). It is the average 1 million b/d of supply assumed to be available to the global market from inventory-or stocks-over the course of 2000.

A warning that appeared in this space last week about the world's need for more oil alluded to that heavy stock draw, which follows an even larger withdrawal, estimated at 1.4 million b/d, in 1999 (OGJ, Jan. 24, 2000, p. 17). But the point then was that the Organization of Petroleum Exporting Countries needs to raise production, by quota adjustment or simple cheating, soon. The subject of oil from storage deserves attention in its own right.

No assurance

It deserves attention because there happens to be no assurance that storage can supply 1 million b/d for all of the year. The incentive to pull oil out of inventory will have force as long as recent price patterns last, with values higher for prompt than for future delivery. But assertion of the incentive begs two large questions: How soon will inventories reach levels too low to support the withdrawal rate? And what does "too low" mean?

The answer to the second question is changing. Stocks traditionally have been considered too low when they fall below some "minimum operating level" calibrated against consumption. A rule of thumb for industry stocks sets the low point at 50 days of forward supply.

By that standard, US industry stocks are already at minimum operating levels, which is partly why OGJ's forecast assumes no net change in US inventories this year. And stocks held in other members of the Organization for Economic Cooperation and Development are nearly that low.

If stocks are to supply 1 million b/d this year, then, most of the oil will have to come from outside the OECD. To the extent that doesn't happen, the industry must operate with stocks below traditional comfort levels.

The industry seems prepared to test new inventory minimums. At a Houston workshop last week hosted by the University of Houston's Energy Institute and US Department of Energy, David H. Knapp, senior analyst at the International Energy Agency, said oil companies have told OPEC members not to worry about plummeting inventories. They say they can comfortably operate with stocks well below the old floor.

So the remaining question is the volume of oil available from stocks outside the OECD. And the answer is that no one really knows.

The deficiency received much attention at the Houston workshop, the purpose of which was to study the role of oil-market data in recent price volatility. The extraordinary stock build of 1998 and equally extraordinary stock draw last year focused unusual attention on the quality of inventory data. Even by the oil market's coarse standards of measurement, it's an area of towering imprecision. While the problem led to calls at the workshop for improvement to stock reporting and analysis, a contrary view emerged.

"What is it that we'd do with this information if we had it?" asked W.L. Littlejohn, president of Littlejohn Associates of Houston. The important quality is demand for stocks, he said. And that's reflected in the price structure-the relationship between prompt and forward prices.

Returns to storage

Philip K. Verleger Jr., senior advisor to the Brattle Group, elaborated the point with charts plotting futures prices by delivery month. On this "forward price curve," futures values drop sharply with time from the present. Verleger pointed out that the distant-month values generate low "returns to storage," indicating a "very tight" market.

A nuts-and-bolts industry might not find economic curves totally persuasive as indicators of shortage. And OPEC seems content for now to simply bank the evidence of a market swing. But prices are in fact telling producers everything they need to know about market conditions, whatever the precision of data about physical values. They always do.