March 21, 2019

facebook twitter linkedin tumblr google pinterest

Forecasting The Oil Price For Fun And Profit (But Not Accuracy)

Forecasting The Oil Price For Fun And Profit (But Not Accuracy)
Spread the love

Over twenty years ago, while in Norway, I showed a group of economists their government’s track record in forecasting national oil production: they consistently predicted a near-term peak and decline and had done so nearly since the beginning of production. At the time of my talk, production had grown regularly and did not appear to be slowing, suggesting that the forecasts needed to be revised. One of my listeners explained to me, in simple terms, that a pessimistic forecast of oil production meant a pessimistic forecast of revenue, which tended to restrain politicians’ spending. I admit that I was stunned to realize that there were logical reasons to create an inaccurate forecast, but in truth this is actually common.

Before the first Oil Crisis in 1973, few bothered to try to predict the price of oil.  Indeed, most assumed it would simply track inflation. (A noted exception was the report The Oil Import Question prepared for President Nixon, which assumed prices would decline in the United States, but that was due to the expected removal of import restrictions.) After 1973, a cottage industry grew up populated by economists (academic and private sector) who sought to cash in on public interest in the price of oil, I mean, answer the pressing question of future oil price direction.

The problem was that nearly everyone got it wrong, which was made especially clear after 1985, when the oil price crashed despite the near-total consensus that it was too low. (See my book The Peak Oil Scare, chapters 2 and 3.) Oil companies that had published annual long-term forecasts ceased to do so, and economists moved into more lucrative areas like electricity deregulation and later climate change. Of course, when prices rose sharply after 2002, the stampede was reversed.

But has the correct lesson been learned? What went wrong in the 1970s was the mistaken belief that depletion or resource scarcity was driving up the price of oil, not short-term disruptions of supply like the Iranian Revolution. Yet when prices rose in the mid-2000s, instead of looking to the problems in Iraq and Venezuela (later Libya, Nigeria, Syria, Yemen, Iran, South Sudan….), many embraced the idea that peak oil was responsible. As the figure below shows, Google searches for peak oil shot up with the price of oil in 2008. (In 2005, the first peak was presumably due to publication of the popular book Twilight in the Desert, which predicted an imminent collapse of Saudi oil production. Oopsie)

Google searches for “peak oil” and the oil priceThe author from Google data

And, just as the 1986 price collapse was unforeseen by the experts despite abundant indicators, the 2014 oil price collapse caught many insisting that $100 was the new floor price, a cliché that had minimal theory or evidence behind it. My own predictions that the price would drop to $50-60 for the long-term were assumed to be a joke, raised the question of whether I was on drugs, and even saw one oil company CEO suggesting I might be an idiot. Yet, just as the tide disobeyed King Canute, so the oil price ignored the consensus of a $100 floor.

So, why do people forecast the price of oil if, as George Bernard Shaw said of marriage, it represents the triumph of hope over experience? First and foremost, prices are important to many industries, whether they are producers, consumers or competitors of the stuff. Budgets for fuel need to be created by airlines and trucking companies, while oil producers base their investments on expected cash flow.

Unfortunately, given the huge uncertainties involved in long-term oil and energy markets, it becomes very easy to produce nearly any forecast desired, which means that bias is frequently a major factor in the process. And since, as mentioned, there are many reasons to produce a forecast that is pessimistic or optimistic, regardless of actual expectations, it should not be assumed that most forecasts represent objective views of likely future prices.

Bias of course takes on many forms. Groups like the International Energy Agency, which was charged with responding to the oil crisis, might be expected to produce alarmist forecasts. OPEC could be pulled in both directions, with some members wanting a high price forecast to bolster their spending plans, and others preferring a low price forecast to discourage conservation and competition. As the Figure below shows, the IEA had the highest forecast in 2016 while OPEC was well below that. (OPEC has since ceased projecting the long-term price of oil.)

2016 Oil Price ForecastsThe author

Companies have their own reason for bias. For a major oil consumer, if your price forecast is too low, your profits will be lower than projected; if you overestimate future oil costs, your profits will come in above expectations. Of course, that only applies to the short-term; an auto company that bases its product line on very high prices, as Chrysler did in the 1980s, can find itself at a loss when prices are below expectations. Chrysler stop producing its largest vehicle, the New Yorker, but managed to recover when prices dropped by creating the minivan.

A company that produces oil has a preference for high price projections, because it implies higher revenue, allowing more investment and potentially better profits and a higher share price. But this can backfire if prices are below expectations, as high-cost projects struggle and international gas sales have a harder time competing with cheaper fuels.

Which helps explain why some oil industry executives prefer more bearish oil price forecasts. Managers who think ever-rising prices will bail them out tend to focus on completing projects quickly, sacrificing efficiency. Telling them to expect low oil prices means they will hold costs down; if the forecast is too bearish, the company reaps higher profits, which is hardly detrimental.

Investors naturally like the idea that oil prices will rise and take equity prices up with them, but at the same time, they want companies to avoid irrational exuberance, making high-cost investments in anticipation of higher prices. Sometimes. Wall Street vacillates between pushing companies to maintain production volumes and return on equity, creating difficulties for the industry. The service industry—those renting rigs for example—need to not only predict oil prices but Wall Street psychology to project their future business.  Good luck with that.

All too many think the lesson is that forecasting can’t be done. This is reminiscent of Mark Twain’s description of a cat which jumps on a hot stove:  it will never jump on a hot stove again, he noted, but it will also not jump on a cold stove. The lesson isn’t that forecasting is bad, but that bad forecasts are bad, and if you remove the bad theories and biases, you can make a better forecast. Not precise, but at least a lot better than the abysmal track record of the punditry-industrial complex.

Source link

More from my site

About The Author

Related posts

Leave a Reply