Financial Market Tools for Traders, Investors & Finance Professionals
Macro Research Workbench — Energy & Crude Series 07
EIA crude oil inventories in the Weekly Petroleum Status Report are the net result of one balance—crude production, imports, exports, refinery inputs, product stocks and utilization. This lesson shows how to read the composition with a simplified balance formula, unit conversion and a waterfall breakdown, rather than judging price direction from the headline stock change alone. We separate the reference week from the release date, commercial stocks from the Strategic Petroleum Reserve, and crude from product stocks, using one consistent set of fictional educational data and a weekly balance-decomposition worksheet.
Key takeaways
Answer
The first thing to settle about how to read EIA crude oil inventories is clear: the crude stocks in the EIA Weekly Petroleum Status Report are not an indicator that sets price direction on their own. The level and its change are the net result of several flows—crude production, imports, exports, refinery inputs (the crude processed) and a statistical adjustment. So instead of pulling out the headline “stocks rose” or “stocks fell,” read production, net imports, refinery activity and product stocks together as one supply-demand balance.
This lesson reads the Weekly Petroleum Status Report through the lens of “decomposing the balance.” It covers separating the reference week from the release date, converting between a daily rate and a weekly volume, the simplified balance formula, a waterfall breakdown of the headline stock change, and how to use refinery utilization and the four-week average—all with one consistent set of fictional educational data. Every figure, chart, table and worksheet default shown here is an illustrative example, not a real market value, forecast or trading recommendation. Correlation does not prove causation, and inventory changes do not determine the direction of crude prices.
This article is one part of a wider macro-research workflow. The full picture that connects COT, rates, real yields and energy is set out in the Macro Analysis Guide, within which this lesson owns the structure of the Weekly Petroleum Status Report and how to read the balance. Testing the lead-lag relationship between stocks and price is left to Lead-Lag Analysis, and comparing historical regimes to Macro Regime Analysis.
Terminology
Before reading the balance, separate the main terms used in the EIA Weekly Petroleum Status Report. Not mixing units and categories is the precondition for reading the later decomposition correctly.
The key point here is not to confuse flows (production, imports, exports, inputs) with the stock (the inventory level). Flows are reported as a daily rate; the stock level is reported in million barrels. The weekly conversion (daily × 7) is what links them. A tidy way to keep the categories apart is set out in the order-of-checks table below.
Dates and units
With weekly statistics, not confusing several dates is the starting point for timestamp integrity. Treat at least these three as separate fields.
Because release times, categories and the schedule can change, always confirm the actual release date, time zone and holiday exceptions on the official EIA schedule. The timeline below is a concept diagram of how the reference week, the normal release, the holiday shift and the retrieval date relate.
Fix the unit conversion up front too. Flows such as production, imports, exports and refinery inputs are reported as a daily rate (kb/d), while the stock level is reported in million barrels. To turn a daily rate into a weekly contribution, multiply by 7. For example, 1,000 kb/d × 7 days = 7,000 kb = 7.0 million barrels/week. Never add or subtract a daily rate directly against a stock level, or mix weekly volumes with daily rates: align the whole balance to the same unit (here, weekly million barrels) before summing.
The balance idea
Why crude stocks rise or fall can be understood with a simplified supply-demand balance. Barrels entering the country (production + imports), minus barrels leaving or being consumed (exports + refinery inputs), plus a statistical adjustment, approximates the change in stocks.
In symbolic form (align everything to the same unit):
ΔStock ≈ Prod + Imp − Exp − Runs ± AdjNetImp = Imp − Exp gives ΔStock ≈ Prod + NetImp − Runs ± AdjThe supply side (production + imports) lifts stocks; the draw side (exports + refinery inputs) lowers them. The flow diagram below shows this relationship in the direction “supply → stocks → draw.”
The point of the formula is that the adjustment (Adj) is a residual filling the gap between the “reported stock change” and the “estimate built from production, net imports and inputs.” Include the adjustment and the identity holds; but looking at the gap between the estimate excluding the adjustment and the reported value reveals a magnitude that a simple balance alone cannot explain. The next section decomposes this with concrete fictional data.
Coherent fictional example
Here we define the fictional educational data used throughout the article. All later prose, figures, tables, worksheet defaults and FAQ examples use these same values (not real market values, forecasts or recommendations). First, the reference-week flows and stock levels.
| Item | Symbol | Value | Weekly conversion (×7) |
|---|---|---|---|
| Crude production | Prod | 13,200 kb/d | +92.4 million bbl |
| Imports | Imp | 6,300 kb/d | +44.1 million bbl |
| Exports | Exp | 4,100 kb/d | −28.7 million bbl |
| Net imports (Imp−Exp) | NetImp | 2,200 kb/d | +15.4 million bbl |
| Refinery inputs | Runs | 16,000 kb/d | −112.0 million bbl |
| Adjustment | Adj | +300 kb/d | +2.1 million bbl |
| Prior-week commercial crude stocks | S₀ | 425.0 million bbl | — |
| Current commercial crude stocks | S₁ | 422.9 million bbl | — |
Substitute the values into the formula. First, the daily-rate stock change is 13,200 + 6,300 − 4,100 − 16,000 + 300 = −300 kb/d. Converted to weekly, −300 × 7 = −2,100 kb = −2.1 million barrels—a draw. The reported stock change is S₁ − S₀ = 422.9 − 425.0 = −2.1 million barrels, which matches the estimate including the adjustment.
Next, look at the estimate excluding the adjustment: Prod + NetImp − Runs = 13,200 + 2,200 − 16,000 = −600 kb/d → weekly −4.2 million barrels. The residual versus the reported value (−2.1) is −2.1 − (−4.2) = +2.1 million barrels, which equals the adjustment’s contribution. In other words, the simple balance alone looks like a 4.2-million-barrel draw, but the statistical adjustment adds back +2.1 million barrels, leaving the reported change at −2.1 million barrels. The waterfall below stacks each component’s weekly contribution to reach the headline stock change.
What the waterfall shows is that behind the small headline of −2.1 million barrels lie far larger offsetting flows—production (+92.4) and refinery inputs (−112.0). The single-week stock change is the remainder of the difference between these large flows, and even a small shift in production, inputs or exports can flip its sign. That is exactly why you need to check the component flows, not just whether stocks rose or fell.
Order of checks
When reading the Weekly Petroleum Status Report, you reduce misreads by not starting from the crude headline but working through the balance factors, product stocks and utilization in order. The table below sets out the series to check, what to watch, the units and this article’s fictional values. Note that crude and gasoline/distillate are separate series.
| Order | Series | What to watch | Unit | Fictional value (this week) | Change vs prior week |
|---|---|---|---|---|---|
| 1 | Commercial crude stocks | Result of the balance; do not add the SPR | million bbl | 422.9 | −2.1 |
| 2 | Production / net imports | Supply-side lift | kb/d | 13,200 / 2,200 | +100 / −150 |
| 3 | Refinery inputs | Largest draw factor | kb/d | 16,000 | +200 |
| 4 | Refinery utilization | Governs the size of inputs | % | 88.4 | +0.6pt |
| 5 | Gasoline stocks | One read on product demand; separate from crude | million bbl | 232.0 | +1.5 |
| 6 | Distillate stocks | Diesel / heating oil; separate from crude | million bbl | 118.0 | −0.8 |
| 7 | Four-week average (inputs) | Trend behind single-week noise | kb/d | 15,800 | +120 |
| 8 | Strategic Petroleum Reserve (SPR) | Policy factor; separate from commercial stocks | million bbl | 355.0 | 0.0 |
Read in this order and internal contradictions come into view. For example, a headline of “crude stocks fell −2.1 million barrels” may be supported by higher refinery inputs (+200 kb/d) and a rise in utilization (+0.6pt), while gasoline stocks instead built by +1.5 million barrels. When a crude draw comes with a product build, you cannot simply call it “tight.” To cross-check against positioning data such as COT for one-sided bets, see How to Read the COT Report and COT Percentile and Z-Score.
The crude stocks, refinery inputs, utilization, four-week averages and product stocks covered so far can first be viewed in the Free Oil x EIA Inventory template. When you need to compare these continuously against long history, percentiles and z-scores, place several energy series on one board, and save a layout to export to CSV or PNG, check the coverage of the Pro-level EIA Energy Board on the current plans page.
Utilization and inputs
The largest draw on crude stocks is refinery inputs (the crude processed). The size of those inputs is governed by refinery utilization, which is calculated as the share of operable capacity being run.
Utilization = Runs ÷ Capacity × 10016,000 ÷ 18,100 × 100 ≈ 88.4% (operable capacity of 18,100 kb/d is a fictional example)Utilization and inputs have a seasonal pattern: they fall during spring and autumn maintenance seasons and rise heading into the driving season and the heating-demand season. So whether crude stocks fell because of demand strength or simply because refineries were running near full cannot be told without looking at utilization and inputs. Conversely, whether a build reflects higher imports or a maintenance-driven drop in inputs only becomes clear once you read the composition.
On top of this, weather (hurricanes halting production or refineries), logistics constraints (pipelines and ports), statistical adjustments and rounding all shake the single-week numbers. To separate these one-off factors from the trend, we next use a four-week moving average. In Table 2’s fictional values, this week’s inputs of 16,000 kb/d sit above the four-week average of 15,800 kb/d, so the single week is running somewhat above trend. Rather than interpreting from the single-week headline alone, it is practical to read both the trend smoothed by the four-week average and the single-week deviation. If you suspect a time-shifted lead or lag relationship, confirm the lagged correlation with Lead-Lag Analysis.
Mini tool
The worksheet below is a small educational tool that computes the implied stock change, the difference versus the reported change, each component’s weekly contribution and the unit conversion from production, imports, exports, refinery inputs, adjustment and the prior/current stock levels. It calculates entirely in your browser; inputs are neither sent nor saved. It does not judge price direction and does not turn results into bullish/bearish or buy/sell output.
First, so it is readable even with JavaScript disabled, here are the defaults and a static worked example (the same fictional data as Table 1).
| Item | Input / calculation | Result |
|---|---|---|
| Net imports (daily) | 6,300 − 4,100 | = 2,200 kb/d |
| Implied stock change (daily, incl. adjustment) | 13,200+6,300−4,100−16,000+300 | = −300 kb/d |
| Implied stock change (weekly) | −300 × 7 | = −2.1 million bbl |
| Reported stock change (weekly) | 422.9 − 425.0 | = −2.1 million bbl |
| Difference (reported − implied) | −2.1 − (−2.1) | = 0.0 million bbl |
| Adjustment’s weekly contribution | +300 × 7 | = +2.1 million bbl |
Each component’s weekly contribution is production = +92.4, net imports = +15.4, refinery inputs = −112.0 and adjustment = +2.1, summing to −2.1 million barrels (rounding may shift the last digit by ±0.1).
Limits
Even after decomposing the balance, you cannot determine the direction of crude prices from inventory data. Always state the following limits alongside your reading.
A healthy approach does not leap from one statistic to price direction; it carries a counter-condition (for example, “a draw led by exports is not demand strength”) together with the additional data to check (production, net imports, utilization, products and the gap versus expectations).
Operations
Before reading the EIA Weekly Petroleum Status Report, confirming the following reduces misreads.
Using the service
The steps above can be checked using the public macro data in the Macro Research Workbench. A rough guide to the graduated tiers is below (feature names, storage behavior and coverage can change, so treat the current plans page and the workbench display as the only source of truth).
| Tier | Main use | Representative capabilities |
|---|---|---|
| Free | Review public macro data | Public views such as the Oil x EIA Inventory template; COT for major currencies and energy; 52-week and 3-year percentiles; source and share |
| Pro | Continuous comparison, saving, export | EIA Energy Board (crude, gasoline, distillate, inputs, utilization); 5-year-to-all-history percentiles and z-scores; 1/4/13/26-week change rankings; multi-market heatmaps; local saving; CSV/PNG and similar exports |
| Premium | Advanced integration and scenarios | Inventory-regime analysis; in-browser CSV joins; point-in-time; lead-lag; scenario builder; reporting |
In practice, first open the Oil x EIA Inventory template on Free and check commercial crude stocks alongside refinery inputs and utilization. Next, when you want to compare crude, gasoline, distillate and the four-week average continuously against long history and percentiles, consider the Pro-level EIA Energy Board. Then, when you reach the stage of reproducing past inventory regimes, joining your own CSV and turning it into a report, Premium’s inventory-regime analysis and data joins apply. This workbench mechanically organizes and visualizes public macro data and data loaded locally on your device; it does not provide lot, margin, trading-cost, trading-signal or individual investment advice. Lot and trading-cost calculations are handled by other tools and articles—the FX and CFD Lot-Size Calculation Guide and the Trading Cost Calculation Guide—while strategy validation is covered by the TradingView Backtesting and Robustness Guide. For designing how the whole macro picture becomes a report, see How to Build a Macro Scenario Analysis.
FAQ
Summary
What to take away about how to read EIA crude oil inventories is that crude stocks are the net result of several flows—production, net imports, refinery inputs and adjustment—and the change alone cannot tell you price direction. Separate the reference week from the release date, align the units of daily rate and million barrels, and decompose the composition with “stock change ≈ production + imports − exports − refinery inputs ± adjustment.” Then keep commercial stocks and the SPR, and crude and product stocks, apart, and use refinery utilization and the four-week average to separate single-week noise from the trend. Follow this order and you avoid the error of judging supply and demand from a single-week headline.
To read next, an article that tests the stock-versus-price relationship from the angle of time-shifting will deepen your understanding. Whether stocks lead, price leads, or the two move together is a question that connects to lagged-correlation thinking.
Read next
MR08: Lead-Lag Analysis Explained — Test Time-Shifted Relationships Without Confusing Causality — check the stock-versus-price time shift through the sign of the correlation and the lag direction.
You can also browse related lessons from the English financial learning articles. For the design of the whole macro picture, continue to Macro Regime Analysis for the full view.
References
Confirm the definitions of crude stocks, production, trade, refinery inputs and utilization, and the release schedule, in the primary sources below. The figures in this article are fictional educational data, not the actual values published by these institutions.
Disclaimer