Oil may fall further, if global demand slows
Aug 6 2008 by Andrew Miller, The Journal
OIL prices soared in the first half of the year, reaching a peak of $145 a barrel at the start of July, causing misery for motorists and road hauliers, to name but two.
Since then, the price of crude has slipped back to close to $125 a barrel and at the time of writing has dropped to about $120.
The key issues for investors – and inflation expectations in particular – is whether this move is likely to prove temporary, or whether it represents the start of a trend towards lower prices. For consumers and motorists, and of course equity markets, a move towards softer and more stable oil prices would most definitely be welcome.
Markets loathe uncertainty and the robustness of the oil price has been a major drag on stock market performance in recent months.
A study by the European Central Bank early this year looked beyond straightforward supply and demand issues to help explain, statistically speaking, trends in oil prices. It considered utilisation rates at refineries, the non-linear effect of changes in capacity utilisation among Opec producers and conditions in the financial futures market. The study found that these factors appeared to work well in explaining price changes.
When we tried replicating the ECB’s research, using the same run of data that it had used, we found no difficulty in getting much the same sort of results “in sample” – ie using data from 1986 through to the end of 2006.
But the ECB model did a very poor job at explaining the recent run-up in oil prices. Indeed, thanks to the New York Mercantile Exchange futures terms having returned to backwardation (with prices for future delivery of oil being lower than spot prices of oil), the model actually predicted that the oil price would fall to under $30 a barrel in recent quarters! This suggests some fundamental problems in the ECB’s model, or a regime change in the oil market.
So, we tried making a number of changes to the ECB model. We broadened its definition of oil stocks to include OECD stocks relative to global (rather than just OECD) demand. We looked beyond the spikes in the capacity utilisation measure, to find a fairly smooth underlying trend. And, finally, we permitted the impact of the futures market to vary, according to whether the market was in contango or backwardation (that is, prices for future deliveries of oil being higher than the spot price, or vice versa).
Whatever the precise mechanisms at play, the new oil price model does a much better job than the ECB’s in fitting recent years’ development. It reckons that the long-run “fair value” for the oil price is about $80 a barrel, much higher than the ECB figure but still some way below recent prices.
The new model also explains changes in the past two decades, although the “residuals” (ie errors made by the model) have grown in the past few years. The fact we have had to work so hard to rescue this model, econometrically speaking, suggests that we should take our equation with a very large pinch of salt. Nonetheless, we tentatively conclude that the oil price may have further to fall if global demand continues to slow.
Andrew Miller is regional office head, Barclays Wealth, Newcastle
