Oil Price Speculation

Robert Waldmann

I almost feel I might disagree with Paul Krugman about international economics with imperfect competition. In particular Krugman confidently asserts that the wild swings in the the price of petroleum are the result of simple supply and demand curves (which are both extremely price inelastic so almost vertical) and not of speculation. He argues that speculation can only affect the spot market price via hoarding. Given supply and demand curves, the only way to affect the spot price is to take product off the market. This didn’t happen during the period of increasing petroleum prices so it was peak oil not speculation.

Given supply and demand curves.

So what do we have to assume to get supply and demand curves ? After the jump I wonder.

update: Well that was quick. Rule number 1 never debate Paul Krugman. Rule number 2 when Krugman seems to be wrong, look at rule number 1.

Krugman has a new post on oil prices and speculation. He notes that this time the price increase is clearly being caused by speculation (which might be perfectly rational) as is shown by the build up of inventories of stored oil above the ground. This means that my theory (after the jump) about how speculation can affect oil prices without causing large inventories of oil above the ground is false unless I can explain what is different now compared to a year ago. This is possible (spare oil tankers not required for shipping are being used for storage) but does sound like special pleading.

To have supply and demand curves it is sufficient (but not necessary) that there be perfect competition. The fact is that some economists tend to accidentally assume perfect competition without thinking. Paul Krugman is not one of those economists and, besides, competition in the market for petroleum is very famously not perfect (remember OPEC ?).

It is also possible to derive a supply curve for a monopolist or for a Cournot oligopoly. To get a supply curve from a cartel one has to consider game theory, but I’d say OPEC can be considered to be Saudi Arabia plus free riders.

However, and crucially, the derivation assumes that the suppliers maximize the present value of revenues. Perfect competition is not needed, but rationality is.

Clearly the fact that Saudi Arabia has oil and wants other things can’t imply anything including a supply curve. If they are total crazy fools they might do anything.

Also, importantly, under imperfect competition the cost of a small deviation from the optimal present value of revenue maximizing price is small squared. The cost of huge pricing errors is huge squared, but after decades of near price stability a boundedly rational non fool might have little idea about pricing as the costs of mistakes are tiny even calculated with the benefit of hind sight.

So how about this model. Saudi Arabia charges a spot price which is a simple function of the futures price and pumps enough to meet demand. They follow this rule unless they are pumping at capacity, in which case they charge a higher price, or less than half of capacity, in which case they charge a lower price.

The futures price would then feed right into the spot price so long as Saudi Arabia is pumping between half and full capacity.

The hoarding of oil is all under the Saudi desert.

The only difference between my model and Krugman’s is that Krugman believes that Saudi Arabia has long been pumping all it can. True Saudi pumping capacity is, in fact, an unobservable variable. One can’t tell if futures market speculation controls the spot price of petroleum without knowing how much oil Saudi Arabia can pump. They are telling us a number, but they are presumably lying so we just don’t know.