naked capitalism

We have noted before that oil supply and demand data is pretty poor. Let’s face it, this is a product that is traded internationally with no central mechanism for purchases and sales. And certain important actors, namely OPEC members, are less than fully cooperative. But investors and commentators, hungry for anything that passes for hard information, treat fuzzy data with more respect than it deserves (note I do not object to them using this information, particularly since there are no good alternatives; I just wish there was more frequent admission of how soft it is).

Confirmation of our view comes via an article in the Oil & Gas Journal (hat tip reader Michael) which summarizes an analysis by Deutsche Bank of the accuracy of forecasts by the leading suspects and finds them to be not too good on the demand side. All exaggerated demand and supplies, but interestingly, the demand overestimates were considerably greater than ones for supply. OPEC has done best on estimating demand, the DOE on non-OPEC supply (and the forecasts there have been pretty reasonably accurate, this is not as helpful as it might seem, since OPEC is very close-mouthed about its production plans, and with such big producers more than a bit opaque, the overall supply picture is murkier than the article suggests).

The article also indicates that it expects 2008 and 2009 demand to fall markedly short of forecasts and OPEC supplies to rebound, which by implication means less pressure on oil prices.

From the Oil & Gas Journal:

The discrepancies between actual world oil supply and demand and the forecasts provided by leading agencies—the International Energy Agency, the Organization of Petroleum Exporting Countries, and the US Department of Energy—show an unreliable track record since 2001, indicated a study recently carried out by Deutsche Bank analyst Michael Lewis.

In his assessment of global oil demand and non-OPEC supply growth, Lewis found that, with the exception of 2003-04, all three agencies have been too optimistic about the strength of oil demand, with OPEC the most cautious. All three agencies have been too optimistic about non-OPEC production growth, with DOE the least bullish and therefore the most accurate in its forecasts.

Concerning global oil demand growth since 2001, Lewis found that forecasts have been “quite similar” in their under or overestimation, with OPEC the most accurate over the decade. Lewis worked out that in absolute percentage terms, since 2002 OPEC has averaged a forecasting error of 53.5%, compared with IEA’s and DOE’s error rates hovering at 70-75%. OPEC’s greater demand forecast caution “may help to explain its reluctance to increase oil quotas and production for fear of bearish implications for the oil price,” said Lewis.

However, in 2003-04, “all three agencies underestimated the surge in global oil demand and specifically the strength in Chinese and US oil demand during those years,” he noted. The discrepancy was large in 2004 as DOE forecast demand growth of some 1.6 million b/d whereas the actual increase exceeded 2.5 million b/d. In contrast, in 2005 DOE overstepped market demand accuracy by forecasting a 2 million b/d growth compared with the actual 1.3 million b/d.

Regarding oil production, all three agencies’ forecasts were relatively close to actual non-OPEC growth, which was averaging 1 million b/d for each year during 2001-04. But 2005 was the worst forecasting year as non-OPEC oil production fell to roughly 0.4 million b/d and the three agencies failed to change their growth assumptions, forecasting a non-OPEC production rise of 0.93 million b/d. Since then, non-OPEC growth levels have failed to recover the early decade levels.

DOE has tended to be the most accurate production forecaster: Since 2001 DOE has posted a forecasting error of 51% compared with 63% for IEA and 72% for OPEC.

Looking into 2009 and taking into account the removal of fuel subsidies in many parts of the developing world, Lewis sees DOE’s 1.4 million b/d oil demand forecast growth as “likely to be too optimistic and even the 0.9 million b/d assumed by the IEA and OPEC may prove too ambitious if recent history is a guide.” He said he expects the recent non-OPEC oil production rebound since 2005 set to continue in 2009.

With OPEC the most bullish in expecting a strong production rebound, this may “provide another clue as to why the cartel [is] likely to be on guard to defend against any weakness in the oil price,” concluded Lewis.