≡ Menu

Why oil (and lots of other stuff) gets cheaper

Arnold Kling argues that oil prices should rise roughly at the rate of interest. Oil is an asset. If you think oil prices over the next year will rise by 1% when interest rates are only 5%, you’re better off selling your oil, taking the money, and putting it in the bank to earn interest. That urge to sell today lowers price today and steepens the growth rate in prices over the next year. If you think oil prices are going to rise by 10% when interest rates are only 5%, you’re better off keeping the oil in the ground rather than selling it. That pushes up prices today and lowers the growth in prices. Only when prices rise by roughly the rate of interest are people content with their decision to leave some oil in the ground and to sell some.

Tyler Cowen wonders how to reconcile this result, sometimes called the Hotelling Rule, with the finding of Julian Simon that most, if not all natural resources get cheaper over time in real terms. Tyler cites some work that suggests that the standard explanations for reconciling these claims don’t work:

This literature suggests that most of the standards “outs,” such as changing costs of production, or new discoveries, don’t square the circle for anything other than the short-run. Holding oil still ought to yield (risk-adjusted) market rates of return, unless again investors are fooled into underestimating how much prices will fall.

I have a semi-standard out. The Hotelling Rule assumes that the cost of extracting the oil (the cost of production) is zero. Selling your oil is just like taking money out of a savings account. But getting oil out of the ground takes real resources. (So does investing in bonds, but the cost of extraction is presumably much higher than the transaction fee on an interest-bearing asset.) When there is a cost of extraction, the arbitrage condition is no longer that the price rise at the rate of interest but that net profit rise at the rate of interest. That still implies that prices rise over time.

In the real world, there is an extraction cost, but over time, it doesn’t stay constant. And it doesn’t just change. It falls steadily over time, driven by technology. This allows the real price of oil to fall over time. To see the intuition, if oil prices are expected to rise at 1% when interest rates are 5%, that would normally encourage me to sell my oil today, driving down the price today and pushing the rate of growth in oil prices up toward 5%. But if I expect it to be a lot cheaper to extract oil next year because of technological improvements in extraction, my net return can actually be higher keeping the oil in the ground and harvesting it next year when extraction costs will be lower than they are today.

One more piece of intuition. You’d think that if oil prices were going to fall steadily over time, it would never be worthwhile to hold on to oil. Get rid of it. Sell it. Get the money out. Put the money into an asset that grows over time rather than one that is going to be worth less year after year. But if extraction costs are expected to fall steadily over time and fast enough to overcome the falling prices of oil, than it’s still rational and profitable to keep some oil in the ground for the future rather than selling all today. The implication of this claim is it’s not enough for extraction costs to fall over time—they have to fall faster than the expected fall in oil prices.

The forces of creativity and innovation that drive down extraction and production costs for oil are the same forces that drive down the price of most goods even when one component of costs, labor costs, is rising.


Next post:

Previous post: