On September 23, the U.S. Energy Information Administration reported that commercial crude oil stocks rose 2.969 million barrels during the week ended September 18. In the same weekly balance sheet, gasoline stocks fell 1.686 million barrels.
If you own oil stocks, which number should you trade?
The gasoline draw is the more useful warning, but it needs an explanation. The EIA's September 23 estimates show refineries processed less crude that week. The gasoline demand proxy rose only a little. Fuel can leave storage because fewer barrels reach it.
That puts refiners and integrated oil companies on different sides of the same question: will a smaller fuel cushion lift the value of gasoline and diesel faster than it cuts the volume refineries can sell?
The Crude Build Is Real
Let's start with the measure behind the headline. Commercial crude stocks, which exclude the Strategic Petroleum Reserve, rose from 423.429 million to 426.398 million barrels in the week ended September 18, according to EIA's September 23 balance sheet.
You may also see a crude build of 2.564 million barrels for that week. That is the same EIA table's total including the reserve. The reserve fell 405,000 barrels, which explains why the two crude figures differ. Neither figure describes the amount of gasoline available to a driver.
Crude and products offered different five-year comparisons in the September 18 week:
Gasoline inventories decreased 1.7 million barrels, 6% below the five-year average. Distillate inventories decreased 0.4 million barrels, 12% below the five-year average.
The EIA's September 23 summary gave those product comparisons and put commercial crude 2% above its own five-year average. We would keep the products on a separate line from crude in our notebook. A refinery needs time and operating capacity to turn the barrel in storage into gasoline or diesel. An extra crude barrel does not instantly replace a gallon already drawn from a product tank.
The location of the September 18 crude build also gives us a clue. The EIA's regional stock table, released September 23, shows Midwest commercial crude stocks rising 4.407 million barrels while Gulf Coast stocks fell 2.093 million.
In that same September 18 week, Midwest refinery crude inputs fell from 4.241 million to 3.802 million barrels a day, according to EIA's September 23 regional estimates. Less crude going into Midwest plants is consistent with more crude remaining in storage there. The table does not tell us whether maintenance, an unplanned outage or another operating decision reduced those runs.
A crude build alongside falling refinery runs may say more about processing than about a sudden flood of new oil.
The Gasoline Draw Has Two Sides
National refinery crude inputs fell 519,000 barrels a day to 16.811 million in the week ended September 18, according to the EIA estimates released September 23. Utilization fell to 94.0% from 96.8%.
Gasoline output from refiners and blenders slipped from 9.644 million to 9.590 million barrels a day in those September 23 estimates. The drop was 54,000 barrels a day. That alone cannot account for the stock draw, because imports, exports, blending and deliveries also move the balance.
The EIA's September 23 trade table shows that imports supplied another piece of the draw, with gasoline imports at 401,000 barrels a day for the September 18 week, down from 537,000 a week earlier. The table reports a 135,000-barrel-a-day decline in that source of supply.
EIA's September 23 balance sheet puts finished motor gasoline product supplied at 8.847 million barrels a day, up from 8.798 million the prior week. The table reports an increase of 50,000 barrels a day.
Product supplied is the EIA's demand proxy; it does not count fuel burned at every pump. One weekly rise of that size cannot establish a new demand surge when gasoline output and imports moved down by more.
The broader comparison is even less friendly to the demand boom claim. Over the four weeks through September 18, gasoline product supplied averaged 8.8 million barrels a day, 0.8% below the year-earlier period, according to the September 23 EIA summary.
So what emptied the gasoline tank? The weekly evidence points more strongly to a reduction in available supply than to a sudden jump in demand. We cannot assign every barrel of the 1.686 million-barrel September 18 draw to one cause from these headline rows, and EIA's September 23 balance sheet should keep us from pretending we can.
The distillate line widens the concern: distillate fuel oil stocks also fell in the September 18 week, to 107.431 million barrels from 107.859 million, according to the EIA's September 23 stocks table. The summary puts that fuel 12% below its five-year average as of September 18.
Low stocks alone cannot forecast a shortage. They mean the next refinery disruption or import change starts with less stored fuel available to absorb it.
Which Oil Stock Gets Paid?
We asked who owns the barrels in Who Owns America's Oil. This week the investment question moves to the plant between the crude tank and the fuel tank.
For Valero owners, ticker VLO, the company's second-quarter 2026 results are the baseline for the refining business. Marathon Petroleum owners, ticker MPC, have the same starting point in its second-quarter 2026 results.
If gasoline and diesel remain scarce relative to crude, a wider difference between product prices and crude costs can help those refining businesses. The risk is visible in the same EIA release: a refinery that runs fewer barrels can lose sales volume even if the value of each processed barrel improves. We need both margins and throughput before calling the weekly draw an earnings gain for VLO or MPC.
Exxon Mobil owners, ticker XOM, have a broader oil exposure. Its second-quarter 2026 results give the company-level baseline. A change in refinery economics has to compete with changes elsewhere in the business, including what Exxon earns from producing crude. That makes XOM a less direct expression of a gasoline stock draw than VLO or MPC.
We have no dated crack-spread price in this report to prove that the stock draw has already improved anyone's margin. Nor does the EIA's national balance say which company's plants caused the fall in runs. Treating the draw as booked profit would put an estimate where an earnings figure belongs.
For your portfolio, the upside case for VLO and MPC is straightforward: fuel stocks stay low, product prices hold up against crude, and plants maintain enough throughput to capture the margin. The risk is that demand softens, imports recover, or further run cuts leave the companies processing fewer barrels.
An XOM position gives you the broader crude and refining mix if the oil price itself becomes the stronger part of the trade. It gives you less concentration on the product squeeze we can actually see in the September 18 EIA tables.
The Next Report Has to Confirm It
The EIA lists September 30 as the next release date on its Weekly Petroleum Status Report page. We will compare that report with the week ended September 18 on three lines: refinery inputs and utilization, gasoline and distillate stocks, and the four-week product-supplied averages.
If product stocks rebuild while runs recover, the immediate fuel squeeze eases. If stocks keep falling even as the demand proxy stays soft, the refiner supply side deserves more attention. A stronger demand proxy alongside low inventories would be a different, more forceful signal.
In October, we will watch the weekly reports for evidence that lower refinery runs persist. We will look at the next VLO and MPC refining results for margins and throughput before turning a national inventory draw into a company earnings call.
For now, we are watching VLO and MPC through September 30 for fuel refining exposure. XOM remains the wider oil decision. We would need sustained product tightness and company results before making a larger claim from one week's draw.
The barrel in storage is not the gallon at the pump. The refinery stands between them.
Seek the truth and be prepared,
Equedia
Sources
- U.S. Energy Information Administration, Weekly Petroleum Status Report, September 23, 2026 balance sheet
- U.S. Energy Information Administration, Weekly Petroleum Status Report, September 23, 2026 stocks table
- U.S. Energy Information Administration, Weekly Petroleum Status Report, September 23, 2026 regional estimates
- U.S. Energy Information Administration, Weekly Petroleum Status Report, September 23, 2026 trade table
- U.S. Energy Information Administration, Weekly Petroleum Status Report text summary, September 23, 2026
- U.S. Energy Information Administration, Weekly Petroleum Status Report release schedule
- Valero Energy, Second Quarter 2026 Results
- Marathon Petroleum, Second Quarter 2026 Results
- Exxon Mobil, Second Quarter 2026 Results
- The Equedia Letter, Who Owns America's Oil
Disclaimer: This letter is for informational and educational purposes only and does not constitute investment advice. Past predictions and performance are not indicative of future results. Please see our full terms of use and disclaimer at equedia.com.
