Oil Inventory Index

Note /

Gasoline stocks vs implied demand: what the weekly mismatch means

Aerial view of a rural inland tank farm with storage tanks, pipework, and surrounding fields.

Every Wednesday the same argument breaks out on the desk. Gasoline stocks draw half a million barrels, but product supplied comes in flat or even down on the week, and someone asks how both things can be true at once. The short answer: they're not measuring the same thing, and the gap between them is where most of the bad takes live.

What "implied demand" is

Product supplied is EIA's plug figure for what left the primary supply chain that week: production plus imports plus the stock change, backed into a number. It's a residual, not a metered count of cars filling up at the pump. That means every input error, every reporting lag from a terminal, every blending component counted differently than usual shows up as noise in product supplied before it shows up anywhere else.

Add the weekly refiner/blender net input adjustment, which exists specifically because EIA knows the raw balance doesn't tie out, and you've got a number that's built to absorb discrepancies rather than deliver a clean demand read. That's fine for a quarterly average. It's a liability for a Wednesday morning call on a four-week change.

Then there's the calendar. Holiday weeks, the first week of a month after a reporting cutoff, weeks with a pipeline outage rerouting barrels between PADDs: all of these distort the single-week print without moving the underlying barrel count much at all. A trader who reacts to one week's implied demand print the way they'd react to a crude stock change is going to get whipsawed.

Reading the draw against the demand signal

The gasoline draw itself, the change in stocks, is a cleaner number. Refiners and terminals report inventory levels with less ambiguity than EIA has to use to back into consumption. So when stocks draw hard and implied demand looks soft, the first question isn't "is demand collapsing," it's "where did the barrels go." Exports pulling supply out of the domestic balance, a terminal doing maintenance and drawing down working inventory ahead of a turnaround, blending components getting reclassified: all of these move stocks without showing up cleanly in the demand proxy.

The four-week average smooths a lot of this, which is why desks lean on it more than the single print, but it still lags. If you're trying to get ahead of the Wednesday release rather than explain it after the fact, the four-week average doesn't help, because it's built from the same lagging weekly reports.

This is the case for an inventory read that doesn't route through the refiner/blender adjustment at all. A floating-roof read on tankage at the terminals and tank farms in your coverage area tells you the stock side of the equation independent of how EIA chooses to balance the week. It won't tell you product supplied. But when the two diverge, having an independent stocks number lets you tell draw-driven noise from demand-driven noise instead of guessing which one explains the headline. Oil Inventory Index builds exactly that series, daily, from satellite passes over covered tank farms, so the stocks side of the comparison doesn't have to wait for Wednesday to tell you something.

The mismatch worth watching

Not every divergence matters the same amount. A one-week gap between a stock draw and a soft demand print, especially around a holiday or a known pipeline event, is usually explainable noise. What's worth flagging to a desk is a mismatch that persists for three or four consecutive weeks: stocks drawing consistently while implied demand stays flat, or the reverse, stocks building while implied demand holds up. That pattern is harder to wave off as reporting noise, and it's usually either an export story, a regional imbalance between PADDs, or a genuine shift in underlying consumption that the weekly proxy is slow to catch.

Treat product supplied as a noisy read on demand and the stock change as the firmer fact on the ground. When you need a stocks number that doesn't wait on Wednesday's balancing act to confirm what the market is already pricing, that's the gap worth closing first.

Get started

← Back to the blog