Reading oil supply and demand through a balanced ledger

Draw the boundary around the balance

Decide whether the question concerns crude oil, finished gasoline, distillate, or all petroleum liquids. A refinery consumes crude as an input while producing several outputs; combining its input and output volumes without a defined boundary can double-count activity. Specify the country or region, reporting period, storage coverage, and whether strategic stocks are included. Keep crude inventories separate from refined-product inventories until the analytical question calls for aggregation.

EIA's petroleum definitions provide the components used in its balances. Read the relevant table notes rather than applying a simplified formula to every product. Imports, exports, transfers, processing gains, and adjustments can enter differently across tables. A global liquids balance and a United States crude balance also have different coverage, so their residuals should not be expected to match.

Reconcile rates and stock changes in compatible units

Production and trade often appear as barrels per day, while inventories appear as total barrels. Multiply an average daily flow by the actual number of days in the observation period before comparing it with a stock movement. Check whether a monthly figure is an average rate or an accumulated volume. Otherwise, calendar length can create a false change in the apparent supply-demand balance.

A balancing adjustment is not automatically a hidden source of consumption or production. It can reconcile separately estimated components, timing differences, and measurement error. Treat a large or persistent residual as an investigation item. Determine whether a source changed its estimation method, revised production, or reclassified a flow before assigning the unexplained volume to a preferred economic thesis.

Worked hypothetical example: explain a crude inventory draw

Consider an invented seven-day domestic crude balance with production of 12.0 million barrels per day, imports of 6.0, exports of 4.0, refinery inputs of 15.0, and zero net other items. Available supply is 18.0 and measured disposition is 19.0. The resulting stock change is minus 1.0 million barrels per day, or a seven-million-barrel draw over the week.

Now reduce imports to 5.0 while leaving everything else unchanged. The hypothetical draw becomes fourteen million barrels. No increase in refinery inputs or final fuel use was needed to produce the larger draw. This example demonstrates why the inventory headline alone cannot identify demand strength. A real analysis would retain all required balance terms and source notes rather than assume the omitted components are zero.

Interpret product supplied carefully

EIA explains product supplied using production, trade, storage changes, and adjustments. It is a measure derived from the supply and disposition system, not a direct count of every consumer's fuel purchase. Distribution patterns, exports, and inventory movements can affect the estimate. Use the source's terminology and describe the inference when interpreting it as an indicator of consumption.

Weekly and monthly estimates are not interchangeable confirmations. EIA notes differences in survey coverage and timeliness between its weekly report and Petroleum Supply Monthly. Compare compatible periods and consider multiweek averages when the research question permits. Smoothing can reduce timing noise, but it also delays turning points and does not repair a systematic measurement problem. Retain the unsmoothed components for diagnosis.

Connect physical constraints to prices without overclaiming

EIA's price explanation discusses supply, demand, inventories, and disruptions, including the limited short-run responsiveness of production and consumption. Those mechanisms explain why equal volume changes need not have equal price effects in every setting. A loss of immediately usable supply can matter differently when substitute grades, transport capacity, and accessible inventories are abundant than when they are constrained.

Investigate geography and product mix before making a broad conclusion. Rising crude inventories alongside falling product inventories can accompany refinery constraints rather than uniformly weak demand. A local crude discount can reflect logistics instead of a worldwide surplus. Futures maturity spreads provide additional evidence about delivery timing, but they do not reveal a unique decomposition into demand expectations, inventory value, and risk compensation.

Build a source-aware release comparison

For each release, record the reference period, publication time, revision notes, and changes to the prior observation. Compare a release with an expectation only when that expectation was documented beforehand and measures the same variable. Without a reliable contemporaneous expectation, describe the published change rather than invent a surprise. Market prices may already reflect information outside the statistical release.

A useful research sheet groups findings into measured components, inferred mechanisms, and unresolved explanations. Ask whether an inventory movement survives adjustment for seasonal refinery maintenance, shipment timing, or a calendar mismatch. Where those adjustments are estimates, show both raw and adjusted findings. Do not choose the seasonal baseline after inspecting which version best supports a bullish or bearish story.

Finish with falsifiable explanations

Write competing interpretations and identify the observations that would distinguish them. A shipment-timing explanation predicts a different follow-up pattern from a sustained production outage. A refinery explanation should be checked against throughput and product balances. This creates a research agenda without making a current price forecast. Define the conditions under which the interpretation would be revised before the next release arrives.

Common failures include mixing crude and total liquids, comparing daily rates with weekly stock volumes, treating product supplied as directly observed retail demand, and extrapolating a domestic report to the world. A balanced conclusion can acknowledge a clear stock movement while leaving its economic cause uncertain. Accounting discipline narrows the plausible explanations; it does not turn a single release into a trading instruction.

What to retain

• Inventory changes must be explained through a product-specific balance.
• Convert rates and volumes consistently and investigate balancing adjustments.
• Product supplied, direct consumption, and global demand are different measurements.

Source notes

1. EIA: Petroleum table definitions, sources, and explanatory notes
2. EIA: How do we calculate product supplied?
3. EIA: Differences in product supplied among publications
4. EIA: Oil prices and outlook

Related research