Oil price forecasts – well below expectations

Oil. The most well known commodity in the world. Every country needs it. Most industries use it in some form. Sheikhs, Sultans and Oligarchs all rely on it for political influence. Vast resources are spent analysing and attempting to predict it. Everyone has an opinion on it. Over the last 20 years, its price has been as low as $10 per barrel and as high as $140. Institutions as reputable as The Economist, Goldman Sachs and Merrill Lynch have all made price forecasts that, with the benefit of hindsight, now look ridiculous.

oil price

How can the price of oil have been so volatile? And why are we so unable to forecast it?  A detailed answer is not possible in one blog post, given the multiplicity of factors affecting the price of oil. Instead, I want to focus on one aspect of the oil market which I find particularly useful in understanding and interpreting the movements in price – price elasticity. In this post, I’ll look at how price elasticity can explain some of the big peaks and troughs in the oil price over the past 20 years. A future post will focus on the oil market today.

Economists use the term “elasticity” to describe the extent to which quantity demanded or quantity supplied of a good changes in response to a given change in price. “Inelastic” describes a situation where the percentage change in demand or supply is less than the percentage change in price, while “elastic” is the reverse – the percentage change in demand or supply is greater than the percentage change in price.

What is striking about the oil market is how inelastic both demand and supply are over short periods of time. On the supply side, the reason this happens is sunk costs. Much oil production involves upfront costs which, once spent, are not recoverable. The decision whether to produce oil from an existing field is just a case of working out the marginal cost of pulling out the extra barrel from the ground. This can be very low for existing oil fields – in some cases as low as $5-10 a barrel. And on the demand side there is also evidence that changes in the price do not cause demand to change materially in the short term. This is because it takes time for the users of oil – consumers and firms – to fully factor in a lower price into their behaviour. This is a key reason why, in the short term, if the oil market is over (under) supplied the price can fall (rise) very quickly.

However over the long-term – more than 1-2 years – things are different. Low prices discourage investments in new capacity and new technologies. Instead the industry runs down its existing fields and supply growth tails off. And, over time, the users of oil also change their behaviour in response to higher or lower prices.

I believe understanding this dynamic of short run versus long run elasticity is key to making sense of the wild swings in the oil price that we have seen happen so often. It may also partly explain why forecasters get it so wrong. Let’s revisit the historical episodes referenced in the chart above.

In early 1999, the oil price hit a low of $10, down 50% from its level two years previously. Around this time, the Economist ran a special report suggesting that the price could half again to $5, with “[little] chance of prices rebounding”. The article cited falling demand due to the economic crisis in Russia and Asia, the demise of OPEC, and new more efficient production technologies (sounds familiar..?)1.

The Economist article missed some important points. The expectation that demand would not pick up in response to much lower prices was likely an overly simplistic extrapolation of recent short term trends out into the future. Demand did rebound, both in Asia and in the US, while China emerged as a huge source of new demand. Subsequent history has also shown that the cost of new sources of supply has ended up being higher than the article suggested. To be fair to The Economist, some of these factors would have been much less obvious at the time than they are now, but the situation was certainly not as clear-cut as the article made it out to be.

The Economist’s $5 prediction pretty much timed the bottom of the oil price, which subsequently rose over 10-fold to a peak of $145 a barrel in mid-2008, at which point Goldman Sachs issued their infamous $200 oil forecast, as “growth in supply fails to keep pace with increased demand from developing nations.”2 This was the era of the “peak oil” theory – ever rising demand in the face of geological limits to supply.

The main reason oil rose so much up to 2008 was demand growth from non OECD countries being consistently greater than expected, in particular from China. The oil industry was unable to respond with new supply, in part because it was not expecting this demand increase, but also because the low prices of the 1990s had discouraged investment in new capacity. But by the time of the Goldman Sachs $200 forecast, the industry had begun investing very heavily in all manner of new higher cost production techniques that, due to higher oil prices, had become economically viable  – for example oil sands, shale oil, deep water etc. These investments effectively underpinned future sources of new supply, making ever higher prices an increasingly unlikely scenario.

It was barely six months later, with the global financial crisis in full swing and oil prices at $40 – down nearly 75% from their peak – when Merrill Lynch, another investment bank, suggested the price could fall further to $25 due to weak demand from the US, Europe, Japan and China. Even their central scenario suggested a price of only $503. However, once again the price had simply overshot on the downside due to a temporary fall in demand and the inelastic nature of short term supply. By the end of 2009, the price was back at the $75-80 level.

One obvious conclusion from these experiences is that we should expect the oil price to be much more volatile over the short term than the underlying long-term fundamentals might suggest. Both theory and practical experience show that it takes some time for supply and demand to respond to price signals. I find this useful to keep in mind at times – such as now – when the oil market appears in complete disarray and panic. For long-term investors, understanding the longer-term fundamentals of the oil market is much more important than worrying – or trying to predict – how low or high the price might go in the next 6 months.

Matthew Tillett


1. The Economist, March 1999, “The Next Shock”. See also, “Drowning in oil
2. Bloomberg, May 2008, http://www.bloomberg.com/apps/news?pid=21070001&sid=ambnKPJvPLDg
3. Financial Times, December 2008, http://www.ft.com/cms/s/0/3beefae2-c1f7-11dd-a350-000077b07658.html#axzz3M9e5tnlI


Author - Matthew Tillett

Matthew Tillett

Articles by this author

Golden nuggets
July 12, 2017

In my previous post, I outlined my framework for investing in gold equities, highlighting the critical importance of ensuring a…

Gold – it’s marmite for investors
June 27, 2017

At one extreme, I know investors who have their entire portfolios invested in a combination of gold, gold equities and…

Mr Market is smiling on consumer staples. For now.
April 11, 2017

Warren Buffet famously described companies that “buy commodities and sell brands” as the sort of business model in which he…

A sign of the times…
March 6, 2017

Would you pay a premium for financial leverage? For most of my career – barring a short period before the…