There is an army of oil analysts out there, constantly predicting that a day of reckoning is coming, but wisely omitting details about which day that will be. There are the broken-clock forecasters, always predicting an oil price collapse, and anxious to take credit when right – once a decade or so.
And there are the investment banks and government agencies that invest millions to peer into the future.
Do we really need this confusing array of oil price forecasts?
I recently took a careful look at the crude oil futures market as a forecasting tool. Futures forecasts are widely cited, and just as often disparaged. The facts tell us that futures prices are not very accurate forecasters, barely beating out a random walk that just says tomorrow’s oil price will be the same as today.
But futures (and usually the random walk) are substantially better than competing statistical forecasts or expert surveys. Superiority of the futures market is expected, as futures are so efficient in digesting publicly available information that they should not be beaten by routine statistical forecasts or expert surveys
What about the lack of accuracy? Look at the futures strip this morning, write down the quote for the 12-month contract, and one year from today compare the forecast to today’s spot price. Why might they differ? Perhaps it is poor analysis. Or it may be that our forecast was completely predicated on now-stale information from a year ago.
Unanticipated events have repeatedly plunged oil markets into chaos: the Arab oil embargo, the fall of the Shah or the invasion of Kuwait. An efficient market scrapes together all available data and uses it to look forward, but no one should pretend it can somehow divine the future.
And it doesn’t take big headlines to upset the forecast. The global crude oil market depends on the volatile politics of producing countries, various economic cycles, and a vast infrastructure of pipes, ships and refineries. Even if we account for the big issues correctly, we could list 1,000 or more low-probability events that could push our forecast off course.
Suppose that each event has a probability of one in a 1,000 over the next 12 months. There is no reason to incorporate any of these possibilities into our forecast or even list them as a risk. But if these events are independent of each other, the chance that at least one will unexpectedly affect the oil market within a year is 1-(.999)1000 or 63.2%. If the events are not independent, it is potentially worse. Think of those YouTube videos with 1,000 set mousetraps, and what happens when you toss in a ping pong ball.
Futures stay ahead of publicly available data and routine statistical models, but what about the gifted analyst, proprietary models, or deep pools of information accumulated through millions invested in forecasting models? Let me express my doubts.
If what makes the futures market inaccurate is the passage of time, and the unfolding of a series of low-probability events, it seems unlikely that the most gifted analyst or well-funded modeler could do much better. We might just as well set them to forecasting the winning lottery number to be picked next Saturday night. Or the cards you will draw in your next hand of poker.
I am forced back to the futures market. Will the price of the 12-month contract in this morning’s paper be an accurate prediction of next year’s spot price? Probably not, because the facts underlying the forecast will constantly change. However, futures may be the best guess we can make based on the information available today. The forecast will change every day, providing a regular update in response to emerging facts.
For now, I will stick to the futures strip for my oil price forecast – available daily, tested against the experts, constantly updated, and all for the price of a newspaper.
Bill Gilmer, Director, Institute for Regional Forecasting, C. T. Bauer College of Business at the University of Houston.
Recommended for you
