Thursday 02 Apr 2020 | 23:20 | SYDNEY
Thursday 02 Apr 2020 | 23:20 | SYDNEY

Oil: Prediction is hard, especially about the future


Mark Thirlwell

5 December 2008 14:51

Remember when oil was at US$147/barrel? It’s not hard to do – it was as recently as July. Yesterday, the oil price had fallen to below US$44/barrel, more than US$100 below the July peak and its lowest level in almost four years.

With price swings like that, it’s not a surprise that forecasts have also oscillated widely: in May, Goldman Sachs was telling the world to prepare for an ‘oil superspike’ that could take the price up to US$200/barrel. Yesterday, Merrill Lynch produced a somewhat different warning, suggesting that US$25/barrel is a possibility for next year — which would be really bad news for the ‘axis of diesel’

This very rapid shift in both prices and price expectations serves as another reminder of the perils of forecasting. It might also tell us something interesting about the nature of the oil market. Paul Krugman has argued that one way to think about the latter is in terms of a backward bending supply curve and multiple equilibria, such that events can quickly ‘tip’ the market from one equilibrium to another, making it vulnerable to sudden price spikes or crashes. He sketches out a simple model here

If the hypothesis of a backward bending supply curve is correct, then the current fall in price should bring an increase in oil production, rather than the fall in oil output implied by a conventional supply curve.

Photo by Flickr user nelgallan, used under a Creative Commons license.