One-Third of U.S. Diesel Leaves as the Pump Nears $6/Gal

The soybean oil curve is sending a stronger warning than yesterday. October versus January is around -0.55 cents/lb after trading above +2 cents/lb during June, a reversal of roughly 2.5 cents/lb. October BOGO is around +$148/mt after breaking the +$160 area, versus more than +$650/mt earlier this year. On the weekly chart, the market has moved back toward levels seen before the 2025 rally. My expectation is for +$100/mt to be tested next. A break below +$100 puts the previous low area around +$50/mt back in view.

BOGO

June U.S. feedstock consumption makes this forward weakness more interesting. Soybean oil use reached roughly 706,000 mt versus 474,000 mt in June 2025, up about 49%. Yellow grease reached about 313,000 mt versus 222,000 mt, up 41%. Canola increased from roughly 108,000 mt to 145,000 mt, up 34%. Corn oil increased from about 156,000 mt to 166,000 mt, up 6%. Tallow was essentially unchanged at roughly 342,000 mt. Across these six major lipid feedstocks, consumption increased from about 1.32 million mt to 1.69 million mt, up roughly 28%. Soybean oil supplied about 42% of this pool versus 36% a year earlier.

UCO is central to how I read those numbers. EIA includes UCO inside its yellow-grease category, so imported Chinese UCO appears in U.S. feedstock consumption once plants process it. China shipped roughly 166,000 mt of UCO toward the U.S. in June and about 75,000 mt in July. June Chinese shipments alone were equivalent to roughly 53% of June U.S. yellow-grease consumption of about 313,000 mt, although sailing time, storage and consumption timing prevent a direct month-for-month reconciliation. The important comparison is numerical. Soybean oil consumption rose 49% while yellow grease rose 41%. June was not a month where waste oils disappeared and soybean oil filled the gap. Both were being consumed heavily.

The estimated quarterly U.S. UCO import profile reinforces this point. Imports fell sharply at the start of 2026 and recovered strongly through Q2, with our estimate pointing toward elevated Q3 and Q4 volumes. Chinese UCO remains importable and usable by U.S. renewable fuel producers. Under current 2026 45Z guidance, Chinese-origin feedstock does not receive the feedstock-origin benefit available to qualifying North American material. Yet Chinese UCO continues moving into the U.S. because RIN value, state clean-fuel credits and delivered feedstock economics still support its use. For soybean oil, this means another feedstock competes for the marginal renewable diesel barrel even after accounting for the tax disadvantage. BOGO at +$148 and Oct/Jan soybean oil at -0.55 cents/lb fit this interpretation.

The soybean harvest adds another source of pressure. July U.S. soybean crush reached a record 222 million bushels, up 8.3% from July 2025, while soybean oil stocks still stood at 1.963 billion lb after falling 6.4% from June. Strong crush means more soybean oil supply entering Q4. I still expect farmers to move beans aggressively into processors because crush margins remain attractive, but I do not expect Q4 soybean oil biofuel use to exceed the June record. Imported UCO, higher crush, the Oct/Jan carry and winter feedstock requirements all point away from another soybean oil usage high.

Diesel is moving in the opposite direction. Arizona diesel is now $5.931/gal versus $5.776 a week ago and $5.593 a month ago. Prices have therefore risen 15.5 cents in one week and 33.8 cents in one month. A year ago Arizona diesel was $3.627, leaving today’s price about $2.30/gal higher, an increase of roughly 64%. The national average is around $5.783/gal. Arizona is less than seven cents from $6, and some local markets are already above $6.

The refinery numbers make the diesel shortage harder to dismiss. U.S. refineries processed roughly 17.5 million b/d of crude at 98% utilization, around 627,000 b/d more crude than a year ago. Yet distillate output was only 5.13 million b/d, 2.4% below last year. Distillate production represented roughly 29.3% of crude input versus about 31.1% a year earlier. Refineries are processing more crude while producing less distillate per barrel. Higher crude runs alone are therefore failing to solve the diesel shortage.

The inventory geography is equally important. U.S. distillate stocks increased 0.8 million barrels to 104.2 million barrels, against expectations for a 1.3 million barrel decline. Gulf Coast inventories increased about 3.1 million barrels while East Coast stocks fell another 1.7 million barrels to 19.3 million barrels, a record low. National inventories remain roughly 14% below the five-year average. The national build therefore masks a worsening East Coast position.

Exports create the most uncomfortable political number. U.S. distillate exports are running around 1.74 million b/d against domestic production of 5.13 million b/d. Exports therefore equal roughly 34% of current production and stand about 29% above last year. At the same time East Coast stocks are at 19.3 million barrels and Arizona diesel is $5.93/gal. Gulf Coast refiners have strong economic incentives to export, but one barrel out of every three leaving the country becomes increasingly difficult to defend politically as retail diesel approaches $6. I expect pressure on product exports to rise sharply if the national average reaches $6. Political intervention is now the largest risk to my bullish diesel view.

The heat crack eased today toward roughly $103/bbl after recently reaching about $106 to $107/bbl. A 3% to 4% retreat from the peak does little to alter the physical picture while refinery utilization sits at 98%, distillate production is 2.4% below last year, East Coast stocks are at 19.3 million barrels and exports remain near 1.74 million b/d. I read today’s decline as a correction from extreme levels, not evidence of a repaired diesel balance.

Europe Bio window remains active. Today’s paper market traded 67.5 kt of HVO Class II, 42 kt of RME, 36 kt of RME/FAME and 27 kt of UCOME/FAME. Dutch September soybean oil was around €1,210/mt against rapeseed oil near €1,225/mt, leaving only a €15/mt premium for rapeseed oil. As temperatures fall, rapeseed gains relative value because of winter-grade performance. Rhine logistics have also deteriorated again. Kaub has fallen to 48 cm from 77 cm on August 29, a decline of 29 cm, with indicative tanker payloads back around 15% to 25% of design capacity.

Asia is providing limited support to soybean oil. Q4 CPO is around $1,272/mt while Q4 POGO has fallen to roughly +$23/mt. The latest Malaysia poll has August production at 1.82 million mt versus 1.796 million in July, up 1.5%. Exports fell to 1.32 million mt from 1.392 million, down 5.2%, while stocks increased to 2.76 million mt from 2.628 million, up 4.9%. Brazil-to-Asia October grain freight is near $57/mt, around $17/mt above last year, while Santos-China March 2027 freight around $47/mt stands roughly $14/mt above its three-year average.

U.S. harvest logistics are tightening as well. St. Louis grain barge rates are about 33% above their 90-day average and Cairo-Memphis rates about 42% above their 90-day average. St. Louis river levels around 2.4 feet sit only about 1.4 feet above the NOAA reduced-tonnage trigger. A large soybean harvest therefore enters the system with strong crush capacity but rising inland transportation costs.

My call is straightforward. BOGO around +$148 has broken the +$160 level, and I expect +$100 to be tested next. The Oct/Jan soybean oil spread at -0.55 cents/lb reinforces the bearish Q4 soybean-oil signal despite June consumption running 49% above last year. Imported UCO, record crush and winter feedstock requirements add pressure. Diesel gives the opposite signal. Refinery utilization sits at 98%, distillate production is down 2.4% year over year, East Coast stocks are at 19.3 million barrels and exports equal roughly 34% of production. I still expect diesel to move higher. If $6 becomes a national diesel price rather than an Arizona price, I expect Washington to become the dominant risk to the trade.