RFS Island Turns into Diesel Island
The move toward a 90-day U.S. diesel export ban follows the direction I had anticipated, although implementation remains unconfirmed. October heating oil fell 16.57 cents to $4.7764/gal, while December lost only 3.18 cents to $4.5171/gal, compressing October–December backwardation from 39.32 cents to 25.93 cents/gal, a 34.1% contraction. Europe moved in the opposite direction: October ICE gasoil gained $56.25 to $1,485.25/mt, while January gained $27 to $1,286/mt. October–January backwardation consequently widened from $170 to $199.25/mt, an increase of 17.2%. The calendar intervals differ, but the regional divergence is measurable: 13.39 cents/gal less nearby premium in America and $29.25/mt more in Europe. The domestic supply argument also has numerical support: EIA reported a roughly 400,000-barrel distillate inventory draw despite 94% refinery utilization. At an assumed export rate of 2 million barrels/day, a 90-day restriction would retain up to 180 million barrels before changes in production or trade flows.
For U.S. biofuels, those 180 million barrels equal 7.56 billion gallons. If all become additional obligated domestic diesel, EPA’s 2026 biomass-based diesel standard of 5.24% implies approximately 396 million additional D4 RINs. At an average 1.6 RINs per gallon, that represents approximately 248 million gallons of biomass-based diesel, or 825,000 metric tons using a biodiesel-equivalent density of 0.88 kg/litre. Supplying the entire volume from soybean oil would require approximately 842,000 metric tons of oil at an illustrative consumption of 7.5 lb per gallon. This is a gross compliance scenario, not an announced mandate increase or an immediate physical purchase requirement. Compliance timing and available banked RINs could defer new production demand, while lower refinery runs, displaced imports and competing feedstocks could reduce the soybean-oil requirement. I also expect the administration to consider offsetting relief, although no adjustment is confirmed. The soybean-oil implication is therefore more supportive of deferred prices than a forecast of 842,000 tons of prompt buying. December 2027 D4 gained approximately 6 cents to $2.190/RIN, compared with a 2-cent gain to $2.110 for December 2026; that larger deferred increase is consistent with, but does not prove, this interpretation. Oct/Mar Soyoil spread did not budge with a -1.01 carry.
Northwest Europe’s biofuel trading showed that the $56.25/mt October gasoil rally did not translate into uniform increases in biofuel premiums. October RME/FAME traded at $105/mt, up $5, while October UCOME premium trades finished at $377/mt, down $18, and Q4 HVO II at $1,600/mt, down $35: changes of approximately plus 5%, minus 4.6% and minus 2.1%, respectively. In the physical window, RME traded at premiums of +$454 and +$464/mt, and UCOME at +$495/mt; no FAME 0 or HVO II transactions were reported in that window. Physical and paper indications have different delivery and pricing bases and should not be compared directly. Feedstock offers moved less than prompt gasoil: Dutch October rapeseed oil rose €2 to €1,245/mt, or 0.16%, while Dutch soybean oil gained €10 to €1,240/mt, or 0.81%. Inland, Kaub fell to 9 cm, compared with 10 cm at the corresponding time on September 22 and a 77 cm equivalent low-water reference. The 68 cm shortfall against that reference adds a delivery constraint to the $199.25/mt October–January gasoil backwardation; the gauge reading itself is not navigable channel depth or a vessel-payload estimate.
Asia’s feedstock numbers contrasted with Europe’s diesel rally. September 23 dollar CPO indications put October at $1,180.50/mt, down $10.75, and December at $1,219.75/mt, down $6.75, declines of approximately 0.90% and 0.55%. December retained a $39.25/mt premium to October. On the quoted palm–gasoil curve, October palm’s discount widened by $104.94 to $261.69/mt, while the Q4 discount widened by $85.95 to $171.32/mt. Those differentials use their own contract bases and observation times and should not be reconstructed from the separate afternoon ICE marks. They also exclude processing, yield and freight costs, so the $261.69/mt discount is not a finished biodiesel margin. Chinese January palm olein gained 41 yuan to 9,781 yuan/mt, or 0.42%, moving in the opposite direction to the dollar CPO indications. For waste feedstocks, the latest available public Chinese FOB ranges remained September 18 levels of $1,120–1,140/mt for standard RED bulk UCO and $1,170–1,180/mt for premium material. Those five-day-old indications cannot establish September 23 replacement costs or confirm an immediate feedstock response to the proposed 90-day US diesel ban restriction.
My further prediction is that restrictions could extend beyond diesel to other refined fuels and ultimately crude oil, with protecting U.S. refining margins becoming part of the policy rationale. Those extensions remain forecasts, not announced measures. Today’s displayed diesel-crack calculation explains the concern: heating oil’s 16.57-cent/gal decline removed approximately $6.96/bbl of product value, while the WTI benchmark in that calculation rose $1.64/bbl, compressing the screen crack by approximately $8.60/bbl to $108.45/bbl. This is a diesel-versus-crude indicator, not a refinery’s total realized margin. Restricting more product exports would not automatically reverse that compression; on a simple barrel-for-barrel calculation, restoring the lost $8.60/bbl would require crude to fall $8.60/bbl relative to diesel, all else equal. That is the numerical basis for my crude-restriction scenario, rather than evidence that it has been decided. For renewable diesel producers, even a 2-cent D4 increase contributes only 3.2 cents/gal at 1.6 RINs per gallon, against the 16.57-cent fall in October heating oil. Today therefore combines a 13.37-cent/gal reduction in that illustrative diesel-plus-RIN revenue measure with potential deferred soybean-oil demand equivalent to 842,000 metric tons under the full-retention scenario. The immediate margin loss is observable; the deferred feedstock benefit remains conditional on compliance, production and policy decisions.
Comments