SPR Temptations

Today's Wall Street Journal includes an op-ed advocating the sale of oil from the US Strategic Petroleum Reserve, based on a clever twist on the usual argument about the need to drive down global oil prices. Rather than worrying about the economic burden on low-income Americans, the author sees an opportunity for the federal government to earn an arbitrage profit on the SPR inventory, possibly creating an attractive way to plug the budget gap that will be created by reforming the Alternative Minimum Tax. The problem is not with the author's math, which seems generally correct, but with his assumptions about the nature of the futures market and how it would respond to such a scheme. Nor is his idea of depleting the SPR and ceding its function entirely to the market prudent, given the kind of world we in which we live. This is a classic half-baked idea: it contains the seeds of something interesting, but in its present form it would likely prove disastrous.

Mr. Henderson's idea depends on the shape of the "forward curve", the relationship between the futures market's price for oil delivered promptly, compared with the price for delivery in subsequent months. The market is currently in steep "backwardation," with yesterday's contract for delivery in December 2007 closing at a price $8.59/barrel higher than that for delivery one year later. Mr. Henderson looks at all that oil in the SPR and sees a chance to sell now and buy back later, earning the "front-to-back spread" on every barrel. If you look at the open interest and the daily volume in the Dec'08 contract, you might conclude that a million barrels per day (MBD) would disappear into that vast pool with scarcely a ripple. But with relatively few of those futures contracts ultimately resulting in a physical delivery, an extra MBD or two would change the entire market, not just via arbitrage, but by altering the expectations that set its current shape. In fact, a large portion of the arbitrage opportunity would probably disappear the moment the government announced its decision to sell SPR oil, and before the first SPR barrel was sold. The front-to-back spread would shrink quickly, and the total profit captured by the government might only be a few tens of millions of dollars.

The key to Mr. Henderson's strategy is how large a difference in supply or demand is necessary to flip the market from backwardation to "contango," in which oil for later delivery is worth more than prompt supply, and what would happen next. As the first SPR deliveries eased the current competition for prompt barrels, the market would more towards oversupply, and the basis of his whole proposition would be stood on its head. As long as the government continued to sell, the market would shift towards contango, and Mr. Henderson's front-to-back play would turn negative, with the Dec'08 repurchase costing more than the revenue from Dec'07 sales. The moment the SPR sale stopped, the market would revert to its former shape, though not quite as far, because participants would expect the government to intervene again.

The result of this scheme would be a game driven by expectations of future government intent, and that seems like a very undesirable sort of meddling in a complex market that underpins so much economic activity, globally. Nor is it clear that driving down the global price of oil by this means--the author's larger goal--would create more than a short-lived price holiday, during which demand growth here and in developing countries might accelerate. That would compress the gap between demand and actual global production capacity still further, rendering the market more volatile once the SPR ran out, and leaving us no way to replace the lost inventory without driving prices even higher.

I have long regarded the SPR as an outmoded holdover from a highly-regulated era. Its existence deters companies from holding larger inventories, and it offers minimal protection west of the Rocky Mountains. But simply abolishing the SPR without providing a practical alternative would be irresponsible, given the geopolitical risks we face; an unregulated market won't perform this function without a mandate or carefully-targeted incentives. The goal of any prudent proposal to privatize these stocks must be to position them closer to where they would be needed in an emergency, and to put them in more responsive, market-savvy hands, rather than using them all up in an unsustainable binge.