Adjudicating the Adjustment
Understanding the wonky statistical debate about US crude data that’s roiling the oil market
Fair warning: this post dips into wonky territory on the hot topic of the EIA’s US oil industry data. I promise that this particular debate does indeed require additional context and contributes to our understanding of fundamental pressures on the oil market.
If you’re already subscribed and/or appreciate the free chart and summary, hitting the LIKE button is one of the best ways to support my ongoing research.
The topic du jour for many analysts is the accuracy of Energy Information Administration (EIA)-reported data, especially as it concerns the infamous crude “adjustment” factor used in the agency’s Weekly Petroleum Status Report (WPSR) and Petroleum Supply Monthly (PSM) reports.
Of course, virtually all rigorously estimated government data includes residual adjustments like these “unaccounted for” barrels; however, in the case of the EIA adjustment, these residuals are growing quickly and overwhelmingly point in one direction, which means that material information is increasingly missing from the model.
Broadly, there are two reasons for the outsized and growing adjustments to EIA’s crude data: the blending of other products into crude (e.g., non-crude liquids blended with crude during the upstream-through-midstream process) and under-reported crude production (e.g., some light hydrocarbons essentially missed entirely).
The current debate around these adjustments is further muddled by the common fixation on the weekly data rather than much higher fidelity but lagged monthly data—essentially compounding the question with the evergreen inaccuracies and volatility inherent in the weekly data.
We don’t know exactly how much crude oil is currently sloshing around the United States and the problem is only getting worse. While tangible and blockbuster-esque factors like Russian sanctions, Chinese lockdowns/reopening, and recessionary risks have dominated headlines over the past year, the topic du jour for many analysts is far more esoteric: the accuracy of Energy Information Administration (EIA)-reported crude oil data, especially as it concerns the now-infamous “Adjustment” factor used in the agency’s Weekly Petroleum Status Report (WPSR) and Petroleum Supply Monthly (PSM) reports. In response to the mounting concerns, the EIA Administrator recently outlined the results of a 90-day review study on these Adjustments in a Twitter thread and promised a more detailed report on March 22 on the This Week in Petroleum blog.
You see, the United States has the largest, most varied, and also best-tracked energy market in the world thanks, in large part, to the efforts of the EIA, the statistical branch of the US Department of Energy, which publishes literally millions of energy-related data series for free. And, while I have often written about the fact that the oil market is much bigger than the US alone, US weekly (WPSR) data is extremely high frequency relative to data we get from the rest of the world, which is typically 2-3 months lagged. The availability, quality, and frequency of US data makes it especially important to oil market sentiment and crude price discovery.
This post aims to provide more context around two key themes: why the crude adjustment is more important than your run-of-the-mill statistical debate as well as natural discrepancies between weekly and monthly EIA estimates. The goal is to provide the reader with critical context in this ongoing debate in the hopes of avoiding common misconceptions and analytical pitfalls.
To get it out of the way, this growing statistical anomaly reflects the evolving nature of the US oil and gas industry—and not at all some kind of nefarious data manipulation; the EIA is one of the most important pillars of open-source energy market analysis, in addition to being the most authoritative final source on US petroleum balances. Yet, at the same time, it’s important to note just how material these adjustments have become: in December 2022 (i.e., latest more-acute monthly data available), the crude oil adjustment was sitting at roughly 1 MMbpd, which is more or less the same flow of oil that was drawing from the SPR at its peak this past summer. Weekly data adjustments have blown out even further to more than 2 MMbpd in the months following. That’s a lot of unaccounted for crude and gives rise to additional uncertainty at a time of more generalized market tumult.
So, what does the adjustment mean?