America Loaned Out Its Emergency Oil at 20 Percent. The Borrowers May Have Made More.

On February 28, 2026, conflict with Iran choked the Strait of Hormuz to less than 10 percent of its normal flow. About a quarter of the world's seaborne oil moves through that strait. Eleven days later, IEA member countries announced the largest coordinated oil release in history. The final tally: 426 million barrels from 34 countries.

The United States contributed 172.2 million barrels, the largest single share. But it did not sell them. It structured the release as an exchange: oil companies borrow the crude now and repay it in kind between late 2026 and 2028, plus a premium of 18 to 22 percent, paid in barrels. Lend 172 million, get roughly 200 million back. The Department of Energy called it a release at no cost to taxpayers. The first tranche: 45.2 million barrels lent, 55 million due back.

The borrowers are named: Shell took the largest share at roughly 16 to 18 million barrels, alongside Vitol, Trafigura, Gunvor, Mercuria, and Marathon.

Almost nobody else did it this way. Japan sold reserves outright and took in about 540 billion yen. Germany lowered its industry stockholding mandate and let firms sell at market price. Canada, which holds no strategic reserve at all, counted a production increase. Of the 426 million barrels released, roughly one in three was never owned by any government. The United States could lend because it is the outlier: it owns its barrels outright, in salt caverns. The Vault Report published the full country-by-country breakdown.

A 20 percent return, paid in oil, sounds like the taxpayer won. The futures curve at the time tells a more interesting story.

When the exchange was announced, Brent had crossed 100 dollars for the first time in four years and touched the mid 120s at the peak. The quarter closed at 118, up from 61 at the start of the year, after the largest monthly price jump on record. But the futures curve was in steep backwardation: front-month WTI traded near 99 dollars while contracts for late 2026 delivery priced in the mid 70s. The market was pricing the disruption as temporary.

Look at the deal from the borrower's seat. You borrow a barrel and sell it into a crisis market above 110 dollars. You owe about 1.2 barrels back in 2027 or 2028, and the futures market will sell you those barrels today in the 70s. Sell high now, lock in cheap repayment now, keep the difference. On rough deal-time arithmetic: sell one borrowed barrel at 110, lock in the 1.22 barrels you owe for about 92, keep roughly 18 dollars per borrowed barrel. Across 172 million barrels, that is on the order of 3 billion dollars, and a borrower could fix that spread with futures the day the loan was signed.

Now run the same arithmetic from the government's seat. It handed over barrels worth about 110 dollars each at the time, and it will receive barrels the same market was pricing near 77. The roughly 200 million barrels coming back were worth about 15 billion dollars on the deal-time curve; the 172 million going out were worth about 19 billion at spot. Measured in barrels, the reserve grows by 28 million and the premium is real. Measured in deal-time dollars, the exchange may have left several billion on the table versus selling outright into the spike.

There is a defensible reason to prefer barrels anyway. The reserve's job is oil, not mark-to-market profit. Selling outright means refilling later at unknown prices, and refill is capped by cavern geology at roughly 3 million barrels per month. The exchange guarantees restock without the government ever competing for barrels in the open market. Repayments run from November 2026 through September 2028, and realized outcomes depend on prices at delivery; the arithmetic above is locked only for borrowers who hedged.

Both sides of this deal can call it a win. But only one side got to hedge. The premium is the headline. The curve is the story.