Hormuz logjam splits Asian refiners: who bears the costlier-crude squeeze?

Two true headlines about the same crisis refuse to reconcile. Nearly six months after the Strait of Hormuz was shut, a headline refining margin has climbed back past $70 a barrel, a level that seemed outlandish two years ago. And Asia's import-reliant refiners are scrambling to buy American crude at premiums they have never paid, because the Gulf cargoes that once fed their plants cannot get out. The market has resolved the paradox by selling first and asking later: Korean and Japanese shares led Asia's sell-off when the war began, and Asian bourses still dip whenever crude rises. The template says Hormuz means higher oil means Asian refiner pain. The margin data tells a different story, and it identifies who really pays.

Start with the mechanism. A refiner does not bet on crude; it bets on the gap between the price of a barrel going in and the value of the gasoline, diesel and jet fuel coming out — the "crack". Hormuz distorts the two sides of that gap unevenly. Crude is reroutable at a price: you replace the lost barrel with Mars or West Texas Intermediate and pay freight, insurance and a premium on top. Refined products are not replaceable. The Gulf refineries that exported them are damaged or throttled, Persian Gulf product exports have fallen by roughly a million barrels a day, Russian processing sits at a 21-year low, and the IEA estimates that global refinery throughput fell by about 4.5m barrels a day in the second quarter. Someone still has to make the world's jet fuel, and it is not the Gulf. Hence product cracks at records — the benchmark 3-2-1 crack near $69 in mid-August against a long-run mid-cycle range of $15-25, with diesel's crack at a record $93.84 — while crude sits far below its wartime peak. Singapore's complex margin, the regional benchmark for Asia, went from roughly $6 before the war to almost $30 in early March, toward $60 at the spring peak, back to $19 by early June, and by mid-August a headline cracking margin was once more above $70.

The first half of the squeeze is universal. Korea routes about 70% of its crude through the strait; Japan draws about 95% of its oil from the Middle East. The disclosed bills show the tariff on substitution: South Korea's GS Caltex paid $13-14 a barrel over Dubai to secure two million barrels of Mars from Shell for November arrival; Eneos paid more than $10 a barrel over WTI; Taiwan's CPC paid $8-9 over Brent. Every Asian barrel now carries freight and a premium, and the most exposed firms have chosen volume over margin: GS Caltex cut daily crude runs from 800,000 to 675,000 barrels rather than feed the cost through. This is the costlier-crude squeeze, and it is real.

But the same companies sit on the other ledger, and that is where the dispersion appears. Product cracks are global, and the complex Korean and Japanese refiners — whose crackers turn heavy crude into middle distillates — have the export vents to sell into the shortage. Korea's jet-fuel exports hit a nine-month high in May. SK Innovation swung to a group operating profit of 3.5trn won ($2.4bn) in the second quarter, nearly three times the street forecast and against a year-earlier loss; S-Oil kept second-quarter operating income at 965bn won after a blowout first quarter. The arithmetic is the point: a crack near $70 minus a $14 delivery premium still clears mid-cycle by a wide margin. Korea is at once the most import-exposed refiner in Asia and the one banking the windfall, because it can price its output at the shortage while paying up to stay fed.

Two complications discipline that picture. The first is accounting. A good share of the profit has been inventory and "lagging" gains on crude bought before prices fall — the June truce knocked SK Energy's refining segment down by 632bn won quarter on quarter, most of its second-quarter profit being inventory-related — and such gains reverse. Cash margins, not book profits, are the durable signal. The second complication is that the dispersion runs within countries as well as between them. Refiners whose barrels exit as naphtha and petrochemical feedstock are losing: naphtha jumped from $776 a tonne to over $1,000 within a week in March, while petrochemical margins were already thin, so Korea's LG Chem shut a cracker rather than feed it. And procurement flexibility separates the winners from the trapped. India routes only about 40% of its crude through the strait, bought tens of millions of barrels of prompt Russian crude on waiver, is Asia's second-biggest buyer of American crude and pushed exports up by about half from May to July. China refused "war-priced" barrels, ran down its own stocks and cut June imports by 41% year on year, to the lowest since 2016 — it can wait the premium out. Korea and Japan cannot. Even those who win are politically naked: prosecutors indicted all four Korean refiners in July for colluding to lift domestic fuel prices, a reminder that a windfall is easier to earn than to keep.

Dispersion is therefore the default. But it is two events away from collapse. The first is a crack that decays toward its pre-war range while the replacement premium holds. The premium is the floor: at a $70 crack a $14 premium is a rounding error; at $30 it is half the margin; below $25 the most exposed payers go underwater and the squeeze turns uniform. The warnings are visible — OPEC+ is raising output, and the IEA now forecasts global oil demand falling by 1.6m barrels a day this year on persistently high fuel prices. The second, nearer kill switch is reopening. The strait has been "reopened" twice, in April's ceasefire and in June's memorandum, and re-closed twice; every rumour of a deal knocks crude sharply lower. A durable reopening would be worse for refiners than a stalled one, because it removes the crude premium and the product windfall at the same stroke. The refiner who paid $14 over Dubai for November Mars is holding the bag.

None of this has fully reached the price, and that is where Asia is being mispriced. The market's uniform short keeps trading because it models Asian refiners as crude buyers, when their economics are those of product sellers. The contradiction is the margin data itself: a record crack, a near-triple earnings surprise, record export volumes. In America the re-rating has already happened — independent refiner shares are up 89% to 175% so far this year, Valero earned $12.54 a share in the second quarter against a street estimate near $10, and fund-flow data show the VanEck Oil Refiners ETF has taken in more than $340m of creation flows this year. In Asia it lags, because the fear runs strongest exactly where the exposure is largest. The lesson is not a tip in either direction. It is that in a chokepoint crisis the crude price is the noise and the crack is the signal — and the two cut in opposite directions for the same firm, which is why a market that watches only the crude price keeps answering the wrong question.