Crude Oil Positioning vs Inventories: Reading COT and EIA Together

Q: What is the WTI COT Index?

A 0-100 rank of Managed Money's current net position in WTI futures against its own 3-year range. 80 or above means crowded long; 20 or below means crowded short. It standardizes raw contract counts so they are comparable across time.

Q: Why compare EIA inventories to a 5-year seasonal band instead of the raw level?

U.S. crude inventories follow a predictable seasonal pattern tied to refinery maintenance and demand cycles. The raw stock number moves with that pattern regardless of whether the market is tight or loose. Comparing the current level to where stocks sat in the same calendar week over the last five years removes the seasonal noise and isolates the part of the move reflecting real supply-demand balance.

Q: Does a counter-seasonal build mean WTI is likely to fall?

No. We tested that assumption against the record rather than asserting it. Of fourteen candidate relationships between an inventory sitting above or below its seasonal band and the price that followed, one cleared our bar for sample size and significance: WTI stocks above the band, all seasons, independent n = 32, t = +1.49, average forward return +3.45% over the following 8 weeks. That is a higher price following a build, the opposite of what a build-is-bearish assumption predicts.

Q: Does crowded positioning meeting a counter-seasonal inventory move carry extra risk?

We tested that crossed grid directly, positioning against inventory band position together, not inventory alone. Every corner cell fell short of 30 independent observations across 12 to 16 years of weekly data, in all four markets we checked, so no direction is published for any of them. The one relationship we do publish a direction for, WTI stocks above the band, is measured on inventory alone and does not include positioning.

Q: Does combining COT positioning and EIA inventories predict where WTI will trade?

No. Neither lens, alone or combined, forecasts price. They describe how crowded speculative positioning is and whether the physical inventory picture is behaving normally for the season. We tested the crossed positioning-and-inventory grid directly and it does not clear our publication bar in any cell, in any of the four markets checked, so no direction is reported for the combination. The one measured base rate we do publish is inventory alone, without positioning.

Q: Where can I see this for WTI without building it myself?

Energy Radar (Ultimate) computes the WTI COT Index against the EIA seasonal-band read automatically every week, alongside the wider crude and refined-product balance. The TradingView indicator puts the positioning half on your chart, and the free weekly COT email covers positioning across major markets as soon as new CFTC data is processed.

30-second answer: Positioning tells you how crowded a trade is. Inventories tell you what the physical barrel picture is doing. WTI speculative positioning (the COT Index) and U.S. crude stocks measured against their 5-year seasonal band are two separate lenses on the same market, and reading them together shows whether the crowd and the physical market agree or pull in opposite directions. This is descriptive context about crowding and physical tightness, not a price forecast, and not a signal to trade on by itself.

Introduction

Most crude oil commentary picks one lens and stays there. Positioning traders watch Managed Money in the futures market. Fundamentals traders watch the weekly inventory print. The two rarely talk to each other, which is odd, since the questions they answer are complementary, not competing: positioning tells you who is exposed and how much room is left for that exposure to grow, inventories tell you whether the physical barrels support that exposure or contradict it.

This article lays out both lenses, then sets them side by side, and closes with what we actually measured when we tested whether a seasonal-band position has meant anything for the price that followed. None of this predicts where WTI trades next. It describes two independent facts, how crowded the speculative side is and where inventories sit against their season, and reports a direction only where the historical record clears a stated bar for sample size and significance.

The positioning lens: the WTI COT Index

The CFTC's Commitment of Traders report breaks out Managed Money, the hedge funds, CTAs, and commodity pools that make up the speculative crowd in WTI futures. Their raw net position is a contract count, not directly comparable across time as open interest and participation change. To fix that, COTInsight converts it into the COT Index, a 0-100 rank of the current net position against its own 3-year range.

COT Index >= 80: positioning is crowded long. Most funds likely to add to longs already have.
COT Index <= 20: positioning is crowded short, the mirror case.
Anywhere in between: positioning is neutral, with room to build in either direction.

"Crowded" is a statement about marginal buyers and sellers, not about direction being wrong. A crowded long means fewer funds are left to add fresh buying, and more of the existing length is available to be sold if sentiment turns. It measures how much of the speculative capacity is already used, not a verdict on the position. See the COT z-score explained for the mechanics, and our COT report for crude oil traders guide for how positioning fits the rest of the WTI and Brent picture.

The fundamentals lens: EIA inventories against the seasonal band

The other side comes from the U.S. Energy Information Administration's Weekly Petroleum Status Report, which publishes commercial crude inventory levels every Wednesday. The number alone tells you almost nothing. U.S. crude stocks move in a predictable seasonal rhythm: refinery maintenance, summer driving demand, and the shift into winter distillate production push the raw level up and down on a schedule unrelated to whether the market is tight or loose right now.

That is why the number that matters is the level compared against its 5-year seasonal band for the same calendar week: the range of where inventories have historically sat in that specific week across the last five years. That strips the seasonal noise out and leaves the part reflecting actual supply-demand balance.

Building above the band: stocks rising faster than normal for the season, a counter-seasonal build.
Drawing below the band: stocks falling faster than normal for the season, a counter-seasonal draw.
Inside the band: inventories moving roughly as expected for the time of year.

A build in April, when refinery maintenance normally pulls crude out of the system and stocks are seasonally supposed to be flat to lower, is a very different signal than the same build in September. The seasonal band is what makes the comparison fair.

Crossing the two lenses, and what we tested

Put the two reads side by side and there are nine combinations: three positioning states against three inventory states. It is tempting to read the corners, crowded long meeting a build, crowded short meeting a draw, as automatically the more fragile cells. We tested that crossed grid directly, positioning against inventory band position together. Every corner cell came up short: fewer than 30 independent observations across 12 to 16 years of weekly data, in all four markets we checked (WTI, gasoline, distillate, natural gas). That is not enough evidence to support a direction, so the crossed grid publishes none; the cell counts are shown, nothing more.

Separately, we tested a narrower question: whether an inventory sitting above or below its own seasonal band, without crossing it against positioning at all, has meant anything for the price that followed. That test covered fourteen candidate relationships between an inventory's band position and the price over the next 8 weeks, across WTI (as the control), gasoline, distillate and natural gas. A relationship counts as measured only if it clears two bars: at least 30 independent, non-overlapping observations, and a t-statistic of at least 1 in absolute value. Weekly-sampled 8-week windows overlap roughly 8 to 1, so a raw weekly count alone overstates how much distinct evidence exists.

One of the fourteen cleared both bars. WTI crude stocks sitting above the top of their 5-year seasonal band, all seasons, drawn from 259 raw weekly observations (32 independent), produced an average forward return of +3.45% over the following 8 weeks, t = +1.49. That is a higher price following a build, not the flush that the old "elevated long-flush risk" label implied. It is one measured relationship with a modest sample and a t-statistic just above our own bar, not a strong edge, and it describes history rather than a rule this week is bound to follow. It is an inventory-only measurement; positioning plays no part in it.

The other thirteen inventory-only candidates, including WTI stocks below the band and every gasoline, distillate and natural gas relationship tested, did not clear the bar, either too few independent observations or a t-statistic too small to treat as anything but noise. We do not attach a direction to any of those, or to any cell of the crossed positioning-by-fundamentals grid above; all of it is reported as counts and current positions, and nothing more.

What agreement and divergence actually tell you

Positioning and the seasonal inventory read can point the same way or opposite ways in a given week. Whether that agreement or divergence has meant anything for the price that followed is the question the crossed-grid test above was built to answer, and it came up short on sample size in every cell, in all four markets. Reading the two lenses together still tells you two real things: how much of the speculative capacity is already used, and whether the physical barrels are behaving normally for the season. It does not tell you what happens next when the two agree or disagree; the one inventory-only relationship that does clear our bar, WTI stocks above the band, does not involve positioning at all.

How COTInsight brings the two together

Assembling this by hand means pulling the CFTC Disaggregated report, pulling the EIA Weekly Petroleum Status Report, building your own 5-year seasonal band from historical stock levels, and cross-referencing it against the current COT Index, every week, before the next print lands and you do it again.

Energy Radar (Ultimate) computes this automatically for WTI, and covers the wider crude and refined-product balance besides: the current COT Index, where U.S. crude and product stocks sit against their 5-year seasonal bands, and the one measured base rate above shown with its own sample size, refreshed weekly as CFTC and EIA data update. You see where positioning and the physical picture currently sit directly instead of rebuilding the spreadsheet.

For the chart itself, the COTInsight TradingView indicator (Ultimate) puts the same z-score and COT Index panels on your WTI chart, so the positioning side lines up visually with price action instead of living in a separate tab.

If you would rather not check either source manually, COTInsight also sends a free weekly COT email with the week's positioning extremes as soon as new CFTC data is processed. Get the free weekly COT email →

Frequently Asked Questions

What is the WTI COT Index?

A 0-100 rank of Managed Money's current net position in WTI futures against its own 3-year range. 80 or above means crowded long; 20 or below means crowded short. It standardizes raw contract counts so they are comparable across time.

Why compare EIA inventories to a 5-year seasonal band instead of the raw level?

U.S. crude inventories follow a predictable seasonal pattern tied to refinery maintenance and demand cycles. The raw stock number moves with that pattern regardless of whether the market is tight or loose. Comparing the current level to where stocks sat in the same calendar week over the last five years removes the seasonal noise and isolates the part of the move reflecting real supply-demand balance.

Does a counter-seasonal build mean WTI is likely to fall?

No. We tested that assumption against the record rather than asserting it. Of fourteen candidate relationships between an inventory sitting above or below its seasonal band and the price that followed, one cleared our bar for sample size and significance: WTI stocks above the band, all seasons, independent n = 32, t = +1.49, average forward return +3.45% over the following 8 weeks. That is a higher price following a build, the opposite of what a build-is-bearish assumption predicts.

Does crowded positioning meeting a counter-seasonal inventory move carry extra risk?

We tested that crossed grid directly, positioning against inventory band position together, not inventory alone. Every corner cell fell short of 30 independent observations across 12 to 16 years of weekly data, in all four markets we checked, so no direction is published for any of them. The one relationship we do publish a direction for, WTI stocks above the band, is measured on inventory alone and does not include positioning.

Does combining COT positioning and EIA inventories predict where WTI will trade?

No. Neither lens, alone or combined, forecasts price. They describe how crowded speculative positioning is and whether the physical inventory picture is behaving normally for the season. We tested the crossed positioning-and-inventory grid directly and it does not clear our publication bar in any cell, in any of the four markets checked, so no direction is reported for the combination. The one measured base rate we do publish is inventory alone, without positioning.

Where can I see this for WTI without building it myself?

Energy Radar (Ultimate) computes the WTI COT Index against the EIA seasonal-band read automatically every week, alongside the wider crude and refined-product balance. The TradingView indicator puts the positioning half on your chart, and the free weekly COT email covers positioning across major markets as soon as new CFTC data is processed.

Summary

Positioning (the WTI COT Index) measures how crowded speculative longs or shorts are against a 3-year range: >=80 crowded long, <=20 crowded short.
Fundamentals (EIA weekly crude stocks) only mean something once compared against the 5-year seasonal band for that same week; the raw level alone is not informative.
We tested the crossed positioning-by-fundamentals grid directly: every corner cell had fewer than 30 independent observations across 12 to 16 years of weekly data, in all four markets, so no direction is published for it. Separately, we tested fourteen inventory-only relationships between a seasonal-band position and the price that followed. One cleared our publication bar: WTI stocks above the band, independent n = 32, t = +1.49, average forward return +3.45% over 8 weeks, the opposite of the old assumption that a build is bearish. The other thirteen are not shown as a measured direction either.
Energy Radar (Ultimate) computes this overlay automatically for WTI every week; the TradingView indicator (Ultimate) puts the positioning side on your chart.

Reading positioning and inventories apart gives you half the picture each. Reading them together tells you how much of the speculative capacity is used and whether the physical barrels are behaving normally for the season. It does not tell you where WTI goes next, beyond the one measured relationship above.

For the broader WTI and Brent positioning picture, including Managed Money vs producer hedging and the forward curve, see our COT report for crude oil traders guide. For a timely read on positioning and inventories during a specific stretch of 2026, see Oil in June 2026: the Hormuz MOU and the COT data.

Data sourced from the CFTC Commitments of Traders report (cftc.gov) and the EIA Weekly Petroleum Status Report (eia.gov). This article describes positioning and inventory relationships; it does not forecast price and is not investment advice. Futures trading involves substantial risk of loss.