Earlier this week I did a post on oil prices in which I failed to take account of futures markets. That’s a no-no. I stated in a comment that I would likely do another post. This is it.

When I graphed oil supply and demand, I looked simply at the current spot market. But, of course, expectations about the future drive oil supply and demand in the spot market.

One good, though insufficient, measure of expectations in the future is the futures market price. I say “the,” but, of course, there are numerous futures prices. They go out more than a year and sometimes substantially more than a year. In Thursday’s Wall Street Journal, the furthest out futures market price reported for crude oil is the price for December 2017 futures.

Over the last few years, I have had Bob Murphy write, for the Econlib Feature Article, a couple of very good articles on oil prices and futures markets. They are “Oil Prices” and “Oil Speculators: Bad or Good.”

One of the important bottom lines in the literature, which Bob points out in “Oil Prices,” is that in the absence of backwardation (which happens when futures prices are below the spot price) there is a tight relationship between spot market prices and futures prices. Specifically, futures prices should exceed spot prices by just enough to cover interest, insurance, and storage costs. If that were not true, there would quickly be arbitrage to make it true. That’s why I find the differential between the March 2016 futures price (which is effectively the spot price) and the April 2016 futures price to be puzzlingly large. The March 2016 futures price for Thursday was $30.66 per barrel; the April 2016 futures price that same day was $32.98. A differential that high seems out of line with reasonable estimates of interest (almost zero), insurance, and storage costs.

It’s dangerous to talk about the spot market without thinking through the futures market. Say, for example, that enough speculators expect that the future demand will fall. That will drive down futures prices. People in the spot market, noticing the lower futures prices, will reduce inventories, which will increase current supply and drive down spot prices. (This assumes that we start with contango, that is, a situation in which futures prices exceed spot prices.) So the supply increase that I discussed in my earlier analysis could be due to an inventory drawdown. Fortunately, this one wasn’t. I noted in my earlier post that inventories were building. So my original analysis may not have been far off. But I should have explicitly brought in futures markets.