By John Kemp, Reuters

“No other industry begins to offer the data problems that are presented by petroleum,” wrote John Blair, a former U.S. government official exactly 40 years ago ("The Control of Oil", Blair, 1976).

The production of basic statistics and forecasts about oil reserves, production, consumption and stocks ought to be a matter of routine.

But it has at times sparked fierce debate and even political controversy when statisticians and forecasters have been accused of making significant errors ("The Politics of Mistrust," Wildavsky and Tenenbaum, 1981).

Doubts about the reliability of energy statistics were a major part of the “energy crisis” that erupted during the 1970s.

As late as 1968, the United States reportedly had 4 million barrels per day of spare production capacity and thousands of wells across Texas and Louisiana were being operated for fewer than 10 days per month.

But by March 1972, spare capacity had dropped to zero, every well was at maximum production, domestic output was falling, and politicians began to speak of an energy crisis.

The oil embargo, announced in October 1973, intensified the sense that something had gone badly wrong, leaving the country unprepared. ("Energy Policy in America since 1945," Vietor, 1984).

Politicians and the media blamed a conspiracy between domestic producers and the Organization of the Petroleum Exporting Countries for engineering the crisis to drive up prices and profits.

Congress held hearings amid a sense the statistics and forecasts prepared by oil and gas producers and the U.S. Department of the Interior had been either inaccurate or deliberately manipulated.

One outcome of the crisis was the creation of a new U.S. Department of Energy and within it a new Energy Information Administration (EIA) in 1977 to produce more accurate and independent data.

Another was the creation of the International Energy Agency (IEA) in 1974 to gather better statistics and bring greater transparency to the international energy markets.

Prior to the energy crisis, most data and forecasters were confidential and under the control of oil and gas producers themselves.

“It is all well and good for Exxon and other firms to say ‘well, we have been providing the statistics and the figures.’ The trouble is there have been no independent statistics,” one congressman complained during a hearing in 1974.

After the energy crisis, data collection and forecasting would be led by impartial civil servants at national and international levels.

The EIA and IEA, statistics twins, have improved the comprehensiveness, quality and availability of data on the oil market significantly ("IEA: The First 20 Years," volume 2, Scott, 1994).

Improvements in data collection and forecasting in the United States, led by the EIA, have largely quelled controversy about domestic oil production, consumption and stocks.

But information on international markets remains much less comprehensive and accurate, mostly as a result of data collection problems in emerging markets and the deliberate secrecy of the oil producers.

And data collection and forecasting can still generate controversy when analysts believe the agencies have made errors affecting market views of the supply/demand/stocks balance and hence prices.

Missing Barrels In 1998-1999

The most notorious controversy in the post-crisis era erupted in the late 1990s when the IEA data showed a large discrepancy between the amount of oil reportedly produced but not reported in the consumption numbers or visible in the recorded stocks.

“It is just over a year since the saga of the ‘missing barrels’ began,” the IEA wrote in June 1999. “Oil prices tumbled as a result of the Asian financial crisis, a mild winter, increasing production from Iraq and an OPEC quota increase in November 1997.”

The agency went on to wonder: “Millions of barrels of unneeded oil were placed in storage around the world; many of them have yet to reappear ... Did the barrels reported as produced, but that have not shown up in OECD stocks, ever really exist?”

Fast-forward six months and the agency was confident it had found the answer. Its December 1999 monthly report entitled “an un-fond farewell to the missing barrels” explained:

“During the first half of 1998 a large amount of the excess supply in the oil market was unaccounted for ... There was strong disagreement at the time as to whether these ‘missing barrels’ were the result of statistical errors, or whether they represented a large increase in oil stored in non-OECD areas.

“Almost two years later ... the weight of evidence is that the missing barrels did exist and that they have now returned to the market.”

But the missing barrels in the IEA’s statistical system brought a withering response from some analysts, notably Matthew Simmons, who wrote a blistering critique faulting the agency.

“The longer the IEA arrogantly refuses to admit that its supply and demand analysis has some flaws, and the longer the agency holds to a belief that massive amounts of petroleum are still hiding somewhere, the greater the risk becomes that oil markets might crash again on another perception of oversupply.”

“Even worse, this misguided information could accidentally create a situation where oil supply falls too much and a bona fide physical shortage is created,” he grumbled.

Simmons claimed the IEA had over-estimated the amount of oil being exported from OPEC countries; the missing barrels never existed ("There Are No Missing Barrels: Our Oil Markets Are Tightening" 1999).

Statistics can be surprisingly political.

Notional Barrels In 2015-2016

The missing barrels controversy is now stirring again because of the huge stock builds implied by the IEA’s estimates for crude supply and demand between 2014 and 2016.

“A notional 1 billion barrels was added to global inventories over 2014/15 and our latest supply and demand balances suggest builds will persist with up to 285 million barrels expected to be added to stocks over the course of 2016,” the IEA wrote in its latest monthly assessment.

In fourth-quarter 2015, global oil stocks swelled by a notional 1.8 million barrels per day, according to the IEA ("Oil Market Report", January 2016).

The problem is that only part of the stock build is visible. Some analysts, particularly those with a bullish view on oil prices, are once again accusing the IEA of getting its estimates wrong.

This time, the view seems to be the IEA is underestimating demand. Consumption is notoriously difficult to estimate and forecast even in the advanced economies; in most emerging markets it involves a large amount of guesswork because their domestic reporting and statistical systems are so weak.

In the statistical framework, non-OPEC production and OECD stocks are generally the easiest to track, while OPEC output, emerging market demand and non-OECD stocks are the source of most of the errors.

The IEA carefully prepares forecasts for oil consumption and supply for the year ahead, and estimates of actual outturn, which are revised for several years afterward.

Between 1994 and 2015, the forecasting error between the initial forecast and the final estimate for non-OPEC supply was zero, with a standard deviation of just 90,000 barrels per day (

The implication is that the IEA has been able to predict non-OPEC supply with a high degree of accuracy and with no systematic bias in its forecasts.

The forecasting error for global consumption was much larger, at an average of 400,000 barrels per day, with a standard deviation of as much as 1.1 million barrels per day (

Oil demand has proved much trickier to forecast and measure accurately, mostly because its large cyclical component and the problems posed by emerging markets, where demand has been growing rapidly but is measured badly.

The IEA has already raised oil consumption in 2015 from a forecast 93.3 million barrels at the time of the January 2015 Oil Market Report to an estimated 94.5 million barrels in the January 2016 report.

Two things follow from this analysis, one more bullish for the outlook for oil prices, and one more bearish.

Based on past performance, it is possible the IEA underestimated oil consumption in 2015, and is forecasting too little oil demand in 2016.

If that proves to be true, the degree of excess supply, and the buildup in stocks, both in 2015 and 2016, might be somewhat smaller than the IEA has estimated and predicted, which would be mildly bullish for prices.

But if the errors are mostly concentrated on the demand side, then anything which causes economic growth and fuel demand to slow in 2016 would be even more bearish.