Global Ethanol Market 2025-2033

Market definition, perimeter and taxonomy

Ethanol is sold as three different products at three different prices, and any published total for this market is really a position on how much weight each of them carries. As a blendstock it trades as a commodity, with a corn-and-cane cost floor and a gasoline-driven ceiling. As an industrial solvent and chemical intermediate it trades at a premium to fuel, set by specification and logistics rather than by energy content. As beverage-grade spirit it trades on excise, provenance and brand, none of which appear in a fuel price assessment.

The perimeter used here. Ethanol sold as ethanol, at the point of production or first sale: denatured fuel ethanol for blending; industrial and chemical grades covering solvents, coatings, cleaning, intermediates, de-icing and electronics; pharmaceutical and personal-care grades; and beverage-grade neutral spirit supplied to the drinks industry as a purchased input.

Four exclusions matter, because each of them is where a larger number usually comes from. First, the value of the finished drink is excluded — a case of whisky is a branded consumer product, not a quantity of ethanol, and sizing drinks revenue as an ethanol market inflates the category by an order of magnitude. Second, alcohol a brewery or winery ferments for its own use is excluded, because it never becomes a purchased product and never carries a merchant price. Third, synthetic ethanol from petrochemical routes is excluded where it competes directly, although its price sets the practical ceiling on industrial grades. Fourth, ethanol consumed inside ETBE and other derivatives is counted once, at the molecule, not again at the blendstock.

Taxonomy. By grade: fuel, industrial and specification, pharmaceutical, beverage. By feedstock: corn, sugarcane and molasses, wheat and other grains, and lignocellulosic biomass. By process: dry mill, wet mill, sugarcane fermentation, and cellulosic hydrolysis or corn-kernel-fibre co-production. By buyer: obligated parties under a blending mandate, refiners and blenders, industrial and pharmaceutical consumers, and the drinks industry. By delivery model: merchant producer selling into an obligation, integrated producer selling into its own blending or distillery operation, and toll or contract processor.

The distinction that carries the most analytical weight is the last one. A merchant producer is exposed to the spread between feedstock and product. An integrated producer is exposed to the spread between ethanol and whatever the ethanol becomes. The two carry different margins, different risk, and in a bad year, different survival odds.

2025 in volume: 122.9 billion litres of fuel, 18.2 billion of everything else

World fuel ethanol production reached 122.9 billion litres in 2025, up from 119.2 billion in 2024, with 127.7 billion forecast for 2026 — a 3.91% step up. Consumption ran at 122.7 billion litres, so the pool cleared. Those estimates reconcile to the country level: the United States at 62.5 billion litres, Brazil 33.2, India 10.4, the European Union near 7.2 and China near 4.5, with roughly 5.3 billion across the rest of the world.

Three lines in that list carry the market. The United States produced 62.5 billion litres — 50.9% of the world — and exported 8.3 billion, both records. Brazil produced 33.2 billion litres, 27.0% of the world. India produced a record 10.4 billion litres, 8.5%. Together they are 86.4% of world output, and that concentration is the most useful single fact about the supply side: this is not a globally diversified industry, it is a three-country industry with a long tail.

Non-fuel ethanol is a separate and much smaller pool — roughly 11.0 billion litres of industrial and chemical grade, 2.6 billion of pharmaceutical and personal-care grade, and 4.6 billion of beverage-grade neutral spirit sold as a merchant product. Added to fuel, 2025 volume is 141.1 billion litres.

The three use-splits behind that total are analysed separately in this series: fuel blending, industrial ethanol, and ethanol in the beverage industry.

The ratio between the two is the point. Fuel is 87.1% of volume but only 75.5% of value, because a litre of specification solvent, or a litre of USP ethanol, is worth two to three times a litre of blendstock. A model built on fuel alone gets the volume right and the revenue wrong.

The price spine: four assessments and one differential

Four published price points set the model, and they do not agree with one another — which is the most important thing about them.

The American plant gate averaged about USD 0.42 a litre; Rotterdam ran roughly 70% higher. That gap is not a quality premium — the molecule is identical — it is a mandate premium. Europe's blend obligation, its import dependence and its failure to grow domestic capacity all push the clearing price above the American one, and the American producer captures as much of it as trade access allows.

The export number is the one that repays attention. US ethanol was worth USD 0.56 a litre FAS at the border against USD 0.42 at an Iowa plant gate — 1.32 times as much, before the cost of moving it to the coast. For an industry whose average 2025 net margin was USD 0.21 per gallon, about USD 0.055 a litre, against a long-run average of USD 0.13, that differential is larger than the whole profit pool. Exports took 13.4% of American output and a far larger share of its margin. The trade chapter follows from that arithmetic.

India is the fourth type of price and the odd one out. Procurement under the Ethanol Blended Petrol programme is set administratively, not indexed to either the American or the Brazilian market; it clears above the US plant gate and does not move with corn. The model carries it as a premium to the US assessment, and that is flagged here as an assumption rather than an observation.

Cost. On the American model corn is roughly 80% of the cost of production. A representative Iowa dry-mill plant converted 2.95 gallons of ethanol per bushel in 2025, alongside 15.25 pounds of distillers grains and 0.95 pounds of corn oil, and earned USD 0.21 a gallon net across the year. But the range was minus USD 0.05 to plus USD 0.56, and the peak arrived in October. A business with a sixty-cent annual range around a twenty-one-cent mean does not have a stable cost curve. It has a harvest calendar.

Forecast 2025 to 2033

USD 81.8 billion in 2025 rising to USD 111.8 billion in 2033, a compound annual growth rate of 3.99%. The build is volume times realised price, by region and by grade, with every price banded to one of the four assessments above. It is not an extrapolation of a published total.

Volume contributes 2.44 points of the 3.99%, price about 1.5. That split is the whole argument of this report, and it is worth stating plainly: the next eight years are a volume story, not a price story. The 2010s were the reverse — American output doubled on a blend mandate that moved from roughly 4% to 10% of the gasoline pool in a decade, and the real price went nowhere. There is no comparable mandate step left to take, in any market of size.

The conservative case holds fuel volume to 136.5 billion litres by 2033 — no further mandate step in India or Brazil, American output flat, Chinese consumption continuing to fall — and lets industrial grades grow below their historical rate. The optimistic case takes fuel to 158.0 billion litres on an Indian E30 and a Brazilian E35, and adds a genuine aviation-fuel draw from ethanol. The spread between them, USD 30.4 billion of 2033 value, is almost entirely a bet on two governments.

North America is the largest region and the slowest, which is unusual enough to be worth saying twice: the biggest part of this market is not where the growth is. Of the USD 30.0 billion of absolute growth to 2033, Latin America supplies 28.8%, Asia Pacific 26.7% and North America 25.0%, with Europe at 17.2%. North America's share of value falls from 41.5% to 37.1% without any volume falling — it is simply that nothing there grows faster than 2.5%.

The two cuts — regional and by grade — are built independently and land within 0.06 billion of each other on 2033, a rounding gap of about 0.05%, which is stated rather than smoothed away. Industrial and chemical grade is the fastest-growing line in the table at 6.90% on a small base, and it is where the revenue quality sits: it earns roughly twice the price per litre of fuel, on contracts rather than obligations, with no government setting the term.

The year-by-year path, and where the USD 0.06 billion actually sits

The three lines of this series are additive, and this page is their sum — not an independent model reconciled to them at the endpoints. That distinction is testable, and this is the test: built as the sum of its three children, the parent reconciles in every year of the window, not only on the two dates the headline quotes.

The non-fuel grades do the growing. They are USD 20.00 billion of the USD 81.76 billion base and they supply 40.1% of the absolute growth to 2033. That is the whole strategic argument of this report expressed as a single ratio, and the annual build is where it becomes visible: the first three years compound at 3.94% and the last five at 4.01%, because the faster lines gain weight inside the total even though neither is large enough to move it much.

The USD 0.06 billion gap now has an address. Every previous version of this report carried a note that the regional cut lands at USD 111.81 billion and the grade cut at USD 111.75, described as the gap between two models built independently. Adding the three children's regional cuts locates it. Asia Pacific is where the two cuts diverge: fuel carries USD 15.35 billion there in 2033, industrial USD 7.28 and beverage USD 1.72, which sums to USD 24.35 billion against the parent's USD 24.41. The USD 0.06 billion is Asia Pacific's, it is 0.25% of that regional block, and it is not distributed across the three children.

That is worth more than the size of the number suggests. A reconciliation that cannot say where a gap sits is a claim; one that can is a check. The gap is in the fastest-moving non-fuel region — Asia Pacific industrial demand — which is exactly where an independently built model should be expected to drift, and it is worth watching rather than smoothing.

Two totals for 2033, both stated. This table totals USD 111.74 billion, because it is the grade cut summed. The headline above says USD 111.8 billion, which is the regional cut of USD 111.81 billion rounded. Volume is unaffected — both cuts agree at 171.1 billion litres.

What the table cannot tell you. Each line carries its published rate smoothly, so the table is a trustworthy statement of composition and a silent one about timing. Two events would break it and neither is modelled: an Indian E30 or a Brazilian E35 would steepen the fuel line inside a single year, and the expiry of the United States production credit for fuel sold after 31 December 2029 is the one date in this series with a legislated cliff in it.

What decides the number

Brazil moved the blend ladder, not the market. The anhydrous blend rose from 27.5% to 30% on 1 August 2025, adding more than a billion litres a year of demand into a cane crop that was simultaneously shrinking. Centre-South crush fell to 611.15 million tonnes in 2025/26 from 621.93 million, ethanol output fell 3.56% to 33.72 billion litres, and hydrous volumes fell 7.82%. Anhydrous rose 4.22% to 12.89 billion. That is a mandate doing exactly what a mandate does: forcing the mix toward the blended grade while the total shrinks around it. Corn ethanol carried the balance, at 9.19 billion litres and 27.28% of Centre-South output.

India finished its first mandate and started writing the second. The public-sector oil marketing companies reached a 19.24% average blend across ESY 2024-25 and effectively delivered E20 in 2025, five years ahead of the original 2030 target, on 10.4 billion litres of supply and against distillation capacity that has reached 19.9 billion litres a year as of February 2026. Projected demand for ESY 2025-26 is above 12 billion litres. A roadmap beyond 2026 is in formulation. India is the only large market where the mandate is still moving up.

China stopped. The national average blend rate is 2.1%. Fifteen provinces carry an E10 mandate and none has been added since 2019. Fuel ethanol consumption is forecast at 4.3 billion litres in 2025, down 11%, because the gasoline pool itself is shrinking under electric-vehicle adoption. An April 2025 negative list subjects all fuel ethanol investment to provincial verification. China is not a growth market for this product and should not be modelled as one.

The United States is flat, and its credit has an end date. Production capacity stood at 18.477 billion gallons across 191 plants on 1 January 2025, with the largest producer holding 17% of it. Output of 62.5 billion litres implies roughly 89% utilisation. Growth in the fuel pool is over; the remaining levers are export access, the aviation-fuel pathway and the cellulosic co-product credit, all of which are covered in the chapters below.

Europe is smaller than its policy noise suggests and grew anyway. Bioethanol consumption rose 4.5% to 7.21 billion litres in 2025 on gasoline-pool growth and E10 expansion, while domestic production reached a record — and the gap between the two widened, which is why imports rose. In December 2025 the Commission proposed cutting the 2035 fleet target from a 100% to a 90% carbon reduction, with the residual compensable through sustainable renewable fuel credits. That proposal, if adopted, is worth more to this market after 2030 than any blend-rate change now being debated.

Strategic outlook

Three conclusions follow, and they are not the ones the volume numbers suggest.

First, own the export channel or own nothing. The 1.32-times export-to-domestic price ratio is the only structural margin in the fuel business, and trade access is negotiated, not earned. The US-UK arrangement that opened a tariff-free quota from 30 June 2025 — and which was followed within months by the closure of the two plants that produced virtually all British ethanol — is the template, and it shows both how quickly access can be won and how completely it can remove a competitor.

Second, the value is in the non-fuel grades. They are 12.9% of volume and 24.5% of value, they grow at roughly twice the rate, and they do not depend on a legislature renewing anything. A producer with a specification-grade or pharmaceutical qualification has a business; a producer with only a fuel obligation has a position on corn.

Third, treat aviation fuel as optionality, not as a plan. The first commercial-scale ethanol-to-jet plant came online in November 2025, which is a genuine milestone and also a facility designed for roughly 10 million gallons a year. It proves the chemistry works. It does not yet move the volume.

Methodology

Volume anchors are official or intergovernmental: world and country fuel ethanol production and consumption from the International Sugar Organization's February 2026 outlook; American plant count and capacity from the Energy Information Administration as of 1 January 2025; market concentration from the Federal Trade Commission's 2025 ethanol report; Centre-South Brazilian crush, sugar and ethanol output from UNICA's close of the 2025/26 harvest; European consumption from the USDA Foreign Agricultural Service's 2025 biofuel annual; Indian blending rates, supply and distillation capacity from the IEA Advanced Motor Fuels country report. Prices are the four published assessments in the table above. Non-fuel volumes and the grade-level price premiums are Pheonix builds, banded to those assessments and to the observed fuel-to-specification differential; they are estimates and are labelled as such wherever they carry a conclusion. All rates and shares were computed, not asserted.

Table of Contents

1. Market Definition, Perimeter and Taxonomy
   • The four exclusions that keep the number honest
   • Taxonomy by grade, feedstock, process, buyer and delivery model
2. 2025 in Volume
   • 122.9 billion litres of fuel ethanol, 18.2 billion litres of everything else
   • Three countries, 86.4% of world output
   • Why 87.1% of volume is only 75.5% of value
3. The Price Spine
   • Four assessments that disagree
   • The 1.32-times export premium
   • Corn at 80% of cost, and a sixty-cent annual range
4. Market Forecast 2025 to 2033
   • USD 81.8 billion to USD 111.8 billion at 3.99%
   • Volume contributes 2.44 points, price about 1.5
5. The Year-by-Year Path: the Parent Is the Sum of Its Children
   • 3.94% in the first three years, 4.01% in the last five
   • The three lines reconcile to USD 111.7 billion in every year, not only on the two dates
   • Non-fuel grades carry 40.1% of parent growth
6. Scenario Analysis
   • Conservative at 2.32%, optimistic at 5.82%
   • What separates them: two governments
7. Regional Outlook
   • North America: largest and slowest, at 2.53%
   • Latin America supplies 28.8% of absolute growth
   • Asia Pacific at 5.11%
8. Segment and Grade Outlook
   • Industrial and chemical at 6.90%
   • Pharmaceutical and personal care at 6.34%
   • Beverage-grade neutral spirit at 4.11%
9. Demand Drivers
   • Brazil’s blend ladder and the shrinking cane crop
   • India’s first mandate complete, the second in drafting
   • China’s stalled programme
   • Europe’s gasoline pool and the 2035 retreat
10. Feedstock and Co-Product Economics
    • The corn crush and the co-product credit
    • Sugar against ethanol at the Brazilian margin
11. Trade and the Export Premium
    • 8.4 billion litres and USD 4.7 billion
    • The US-UK quota and the British plant closures
12. Competitive Landscape
    • A hundred firms and no share
    • The four competitive layers
    • Cost position as the only durable moat
    • What consolidation actually looks like
13. Value Chain
    • The crush: corn in, ethanol, distillers grains and corn oil out
    • The cane mill: sugar against ethanol at the margin
    • The merchant, the blender and the obligation
    • Where the value accrues
14. Investment Activity
    • Capex has gone to co-products, not to litres
    • Carbon capture: permitted, not built
    • The 45Z window and its end date
    • Second generation: real, and small
15. Technology Innovation
    • Yield is the technology
    • Cellulosic co-production finally earns money
    • Corn ethanol to jet
    • What is not working
16. Market Risk
    • Six risks, one dominant
    • Policy duration, feedstock basis and technology substitution
    • Quantifying the downside
17. Regulatory Landscape
    • The mandate map
    • The United States: RFS, 45Z and the 2029 cliff
    • Brazil, India, Europe and China
    • Trade rules and the certification layer
18. Strategic Outlook
    • Own the export channel or own nothing
    • The value is in the non-fuel grades
    • Aviation fuel as optionality, not as a plan
19. Methodology and Data Sources

Competitive Landscape

The American ethanol industry is the most structurally unconcentrated large commodity business in the energy complex, and that single fact explains almost everything else about how it competes. The Federal Trade Commission’s 2025 report on ethanol market concentration records approximately 100 firms producing or marketing ethanol in the United States, with the largest holding 17% of domestic capacity and total capacity including plant under construction at 18.541 billion gallons a year. A market in which no participant holds a fifth of capacity has no price leadership. It has a cost curve and a weather forecast.

Four layers that do not compete on the same axis

The merchant dry-mill producer. Single-site or small-portfolio, typically 100 to 500 million gallons, dry mill, corn in and ethanol, distillers grains and corn oil out. The asset is a spread between two commodity prices. This is the numerically dominant form: 191 plants with 18.477 billion gallons of capacity at 1 January 2025, of which 42 sit in Iowa at 5,039 million gallons and another 24 in Nebraska at 2,404 million. Fixed costs are sunk, so the plant either runs or it does not. Implied utilisation of about 89% in 2025 is healthy by any process-industry standard, and it still leaves roughly a tenth of the fleet acting as marginal supply.

The integrated agricultural processor. Wet milling rather than dry, with ethanol sitting inside a portfolio of starches, sweeteners, proteins and feed. Ethanol becomes a by-product of a larger grind, which means the operator can absorb a negative ethanol margin for longer than a merchant can — and it means the ethanol capacity is asymmetric, because it turns down before the rest of the complex does.

The integrated energy major. Refining-based ownership, where ethanol capacity exists to satisfy a blend obligation and to capture value on product the refiner already sells. The logic is vertical rather than industrial: owning the molecule removes a purchased input from the refinery’s cost stack. These are the natural buyers at the bottom of a cycle, because the asset’s value to them does not rest on the ethanol margin alone.

The sugarcane mill. Structurally different again, because the same tonne of cane makes sugar or ethanol and the mill chooses at the margin. Raízen produced 2.7 million cubic metres of ethanol and 4.8 million tonnes of sugar in its 2025-26 crop year against R$11.3 billion of adjusted EBITDA, and its reported constraint was not the market but the field: wildfires across the Centre-South during 2024-25, below-average off-season rainfall, above-average rain at the start of 2025-26 and frost, all compressing the crush. That is a business whose output is set by agronomy and then allocated by price.

Beneath these four sit two more forms that matter for volume but not for margin: the Indian distiller operating under an administered procurement price, where the return is set by government and the risk is that the government changes its mind; and the licensed Chinese producer, whose capacity is effectively frozen by an investment negative list.

Cost position is the only durable moat

There is no brand, no switching cost and no contract that protects a higher-cost producer. The competitive variable is conversion efficiency, and it moves in increments small enough to look trivial and large enough to decide who survives. The representative Iowa plant converted 2.75 gallons of ethanol per bushel in 2007-2011 and 2.95 by 2023-2025 — a gain of roughly 7% in fifteen years, achieved mainly through enzyme chemistry and corn-kernel-fibre co-production rather than through new equipment. Add 15.25 pounds of distillers grains and 0.95 pounds of corn oil per bushel and co-products contribute something like a third of plant revenue by value, which is what allows the plant to keep running with corn near USD 5.00 a bushel.

Two plants with identical technology but a different corn basis can be fifteen cents a gallon apart on cost. That is why capacity migrated to the corn belt rather than to the customer, and it is why the industry’s profit series looks the way it does: an average of USD 0.21 a gallon in 2025 against a long-run average of USD 0.13, with a range from minus USD 0.05 to plus USD 0.56 inside a single calendar year.

What consolidation actually looks like

The United Kingdom is the cleanest recent example, and it is not a consolidation story at all. The US-UK Prosperity Deal created a tariff-free quota from 30 June 2025 for 1.4 billion litres of American ethanol, prorated to 913 million litres in that year, removing a 19% ad valorem tariff. Within months the two plants that produced virtually all British ethanol had closed. A protected domestic industry of two assets was not outcompeted plant by plant; it was removed when the protection was. Any producer whose case rests on trade defence rather than on cost should read that sequence twice.

Inside the United States the axis has moved from gallons to grams. Because the clean fuel production credit is calculated from a fuel’s emissions rate, a plant’s carbon intensity — its process energy, its dryer configuration, its access to a sequestration pipeline — now sets the price it receives rather than only the cost it bears. Summit Carbon Solutions’ Midwest Carbon Express holds an Iowa permit issued in August 2024 for 688 miles of main line, with eighteen lateral dockets covering roughly 340 miles to connect facilities including POET and Valero plants. Three laterals were withdrawn in June 2026, eight counties came out of the main route, and hearings on two surviving dockets are scheduled for February and March 2027. The network is permitted. It is not built.

That gives the competitive future in one line: the fleet is dividing into plants with a low-carbon pathway and plants without one, and the second group is worth less than its replacement cost. No amount of operating excellence compensates for a carbon intensity the credit will not pay for.

Value Chain

The ethanol value chain is a feedstock conversion business with a credit-trading overlay, and the two halves obey different logic. The physical half is a crush: buy a carbohydrate, ferment it, sell an alcohol and three co-products. The financial half is an obligation: someone must blend a percentage, and the right to satisfy that obligation carries a price. Confusing the two is the most common analytical error in this industry, because the crush decides whether a plant survives and the obligation decides whether it earns anything.

The crush: corn in, ethanol, distillers grains and corn oil out

Corn is roughly 80% of the cost of American ethanol, which makes the upstream relationship unusually simple and unusually brutal. The 2025 model is well documented: 2.95 gallons of ethanol, 15.25 pounds of distillers grains and 0.95 pounds of corn oil from a bushel, with 82.5 cubic feet of natural gas consumed in the process and non-corn variable costs of USD 0.17 to 0.22 a gallon. The plant is therefore short corn and long three products, and its margin is the residual between them.

Two of those products are strategically more interesting than the ethanol. Distillers grains substitute directly for corn and soybean meal in animal feed, so their value rises when feed grain is expensive — a natural partial hedge. Corn oil is the feedstock for biodiesel and renewable diesel and has become the most volatile line in the plant’s revenue: it started 2025 at USD 0.44 a pound, peaked at USD 0.56 in August and fell back to USD 0.52. A plant that installed oil extraction a decade ago is a materially different business from one that did not.

The physical constraint is the harvest. Corn prices ranged from below USD 4.00 to nearly USD 5.00 a bushel in 2025, and the plant’s margin trough and peak followed that range rather than any product-market event. This is the structural reason ethanol margins look cyclical: they are a spread between two harvests with a chemical plant in the middle.

The cane mill: sugar against ethanol at the margin

The Brazilian chain inverts the logic. Cane yields both sugar and ethanol and the mill allocates between them, so ethanol supply responds to the sugar price rather than only to the ethanol price. In the 2025/26 Centre-South harvest the mix settled at 50.38% sugar and 49.62% ethanol, a shift toward sugar from the prior cycle’s 48.06 and 51.94. Cane processed fell to 611.15 million tonnes from 621.93 million, sugar held essentially flat at 40.43 million tonnes against 40.18 million, and ethanol fell 3.56% to 33.72 billion litres.

Within that, the composition moved hard because the mandate moved. Hydrous ethanol — the E100 product sold directly at the pump to flex-fuel cars — fell 7.82% to 20.83 billion litres, while anhydrous, the grade blended into gasoline, rose 4.22% to 12.89 billion. When the government lifted the blend from 27.5% to 30% on 1 August 2025, the mill did not produce more ethanol. It produced a different grade of it.

The most important structural change in this link is corn ethanol inside Brazil. Corn ethanol reached 9.19 billion litres in 2025/26, up 12.26%, and now accounts for 27.28% of all Centre-South ethanol. Brazil has imported the American feedstock model into a country with a cane model, which gives it a second, uncorrelated source of supply in the off-season between cane harvests. That is why the country could add a billion litres of mandate-driven demand in a year when its cane crop shrank.

The merchant, the blender and the obligation

Between the producer and the pump sits the obligation, and its economics are now in transition. Under the Renewable Fuel Standard, the blender demonstrates compliance with credits, and in 2025 the D3 cellulosic credit averaged USD 2.34 a gallon — a large number against an ethanol price of USD 1.60. It is also why corn-kernel-fibre co-production spread quickly: generating 0.10 gallons of cellulosic ethanol per bushel earned the plant roughly USD 0.23 a gallon of additional revenue against about USD 0.20 a gallon of enzyme cost, and it was worth USD 0.06 of the plant’s USD 0.21 net margin.

The clean fuel production credit changes who holds that value. Because it is a production credit measured from a fuel’s emissions rate rather than a blender’s credit, the incentive moves upstream to the party that can actually lower the carbon intensity — which means the producer, and which means the technology that determines intensity becomes the object of investment.

The export channel is the other half of the merchant’s economics, and the numbers are stark. American producers earned USD 4.7 billion on 8.4 billion litres of exports in 2025, or USD 0.56 a litre, against a domestic plant gate of about USD 0.42. Even after the cost of moving product to a port, that differential exceeds the industry’s entire average net margin. Export access is not a nice-to-have for a merchant producer. It is the only place in the chain where price is set by something other than a corn farmer.

Where the value accrues

Four parties take value out of this chain before the producer does. The feedstock owner takes it through the basis. The technology licensor takes it through enzymes, yeast and process licences — a small percentage of revenue that is nearly all margin, and the one position in the chain with genuine pricing power. The logistics provider takes it through the cost of moving a hazardous, water-absorbing, low-value-per-litre liquid, which is why ethanol moves by rail and barge and almost never by truck beyond a short radius. And the obligation holder takes it through the credit.

What is left lands with the producer, and it splits by grade. A litre of specification-grade solvent, or of USP ethanol for pharmaceutical use, sells for roughly twice a litre of fuel and is governed by qualification, audit and supply agreements rather than by a mandate. That is where a producer converts a commodity position into a business, and it is why the non-fuel grades — 12.9% of volume and 24.5% of value — carry a disproportionate share of the industry’s durable earnings.

Investment Activity

Almost no money in this industry is being spent on making more ethanol. That is the single most important fact about its capital cycle, and it is why a forecast of 2.14% fuel volume growth is not pessimism but arithmetic. Capacity in the United States stood at 18.477 billion gallons at the start of 2025 and 18.541 billion including plant under construction — a difference of 64 million gallons across an entire industry. New greenfield capacity is not being financed anywhere in the American market.

Capex has gone to co-products and to carbon intensity, not to litres

What the industry does build falls into three categories. The first is co-product extraction: corn oil presses, distillers-grain drying and feed fractionation, all of which raise revenue per bushel without adding a gallon of ethanol. The second is energy efficiency in the plant itself — process integration, heat recovery, dryer upgrades. The third, and by far the largest in ambition, is carbon capture.

The last of these is the interesting one, because it is the only capital programme in the sector large enough to change a plant’s economics outright. Summit Carbon Solutions’ Midwest Carbon Express was permitted by the Iowa Utilities Commission in August 2024: 688 miles of main line, with separate applications for eighteen trunk and lateral pipelines covering roughly 340 miles to connect POET and Valero facilities into the network. Since then the project has moved sideways rather than forward. Eight counties were removed from the proposed route in a June 2026 amendment, three lateral dockets were withdrawn the same month, and hearings on the two surviving laterals are set for February and March 2027.

Read that sequence carefully, because it is the whole investment story in miniature. A permitted pipeline has not been built, and the plants whose carbon intensity depends on it cannot count on it. Capital allocated to carbon capture is currently allocated to legal process, land easements and engineering studies, not to steel.

The credit window, and its end date

The clean fuel production credit is what makes that capital programme rational, and it is also what makes the industry’s 2029 nervous. Section 45Z provides an income tax credit for clean transportation fuel produced domestically after 31 December 2024 and sold by 31 December 2029, at an emissions rate not greater than 50 kilograms of carbon dioxide equivalent per million British thermal units. The emissions rate is determined by the credit’s own model rather than by a producer’s own accounting, and the qualifying pathways published for calendar 2026 confine ethanol to US corn starch, US sorghum grain, US corn stover and — for the alcohol-to-jet pathway only — Brazilian sugarcane.

Three consequences follow. First, the credit is not permanent, so any project justified by it has to pay back inside a five-year window; that favours retrofits over new plants. Second, the emissions threshold is a technical bar, not a commercial one, and it is calculated on a model a producer does not control. Third, and most consequentially, a Brazilian cane producer can only reach the credit by converting its ethanol into jet fuel. The policy does not merely reward lower carbon intensity; it specifically routes one class of producer into one product.

Second generation: real, and smaller than the narrative

Raízen’s second-generation programme is the largest operating cellulosic ethanol investment in the world, and it is worth stating its size precisely. The company produced 120.2 million litres of E2G ethanol in its 2025-26 crop year, up 104% year on year, from projects at the Vale do Rosário and Gasa mills, while stabilising and ramping Univalem, Barra and Bonfim. Second-generation output converts bagasse and straw — residues the mill already owns — into additional ethanol without additional cane.

That 120.2 million litres is 0.36% of Brazil’s 33.72 billion litres. A doubling of a technology that has been commercially pursued for fifteen years still leaves it as a rounding error in the supply base. The investment case for cellulosic ethanol is therefore not a volume case. It is a carbon-intensity case: the residue-derived litre carries a better emissions score than the starch-derived one, and under a credit calculated from emissions, a better score is the product.

Where the capacity is actually being built

Two places, and neither is the United States.

India is building distillation capacity, not just demand. Capacity reached 19.9 billion litres a year by February 2026, against a blending programme that has moved from 1.53% to 20% in roughly a decade and against projected ESY 2025-26 demand above 12 billion litres. Indian capacity is being financed by an administered price that is set above the American and Brazilian markets, which makes it the only large market where a producer is effectively guaranteed a margin. The corollary is that the margin is a policy decision and can be revised.

Brazil is building corn ethanol, not cane ethanol. Corn ethanol reached 9.19 billion litres in Centre-South Brazil in 2025/26, up 12.26%, and now accounts for 27.28% of the region’s output. Corn ethanol plants run in the cane off-season, use a feedstock Brazil produces in surplus, and are cheaper and faster to build than a greenfield mill. In a year when the cane crop fell to 611.15 million tonnes, corn was the growth feedstock.

What a buyer is paying for

Because no one is building greenfield plants, the transaction market is the investment market, and the valuation axis has changed. Historically an ethanol plant was worth a multiple of its gallons of capacity, benchmarked against replacement cost. Under a production credit calculated from emissions, that is no longer the right question. The right question is whether the plant has a credible path to a low carbon intensity score — which in practice means access to sequestration, either by pipeline or by a geological formation close enough to pipe to, or a process configuration that gets there on its own.

That is why the sector’s capital is being spent on equipment it never used to buy, and why the plants most likely to change hands are the ones furthest from a sequestration route. The investment case in this market has stopped being about ethanol and started being about the by-products of making it — the carbon, the oil, the feed, and the aviation fuel.

Technology & Innovation

The technology that matters most in this industry is not new. It is the number of gallons a bushel yields, and it has moved from 2.75 to 2.95 over roughly fifteen years. Everything else — enzymes, yeast, process control, artificial intelligence applied to a fermenter — is in service of that number, because a 1% yield improvement on 62.5 billion litres of American output is worth more than any single plant ever built.

Yield is the technology, and the gains come from chemistry

The mechanism of that improvement is worth understanding, because it explains where the next increment will come from. The ethanol yield per bushel rose from 2.75 gallons in 2007-2011 to 2.80 in 2012-2014, 2.85 in 2015-2018, 2.90 in 2019-2022 and 2.95 in 2023-2025. Almost none of that came from new tanks or bigger fermenters. It came from better enzymes, more heat-tolerant yeast, tighter process control and — in the last two steps — from a change in what the plant counts as feedstock.

Adding the fibre of the corn kernel to the fermentable sugar stream is the most consequential change in American ethanol technology since the industry scaled. Corn-kernel-fibre technology uses specialised enzymes on the fibrous fraction of the kernel inside the same fermentation tank used for starch, so it requires no additional physical equipment. It is the rare process improvement that is available to an existing plant at near-zero capital cost.

Cellulosic co-production finally earns money

What made it economic was not the ethanol. It was the credit. Generating 0.10 gallons of cellulosic ethanol per bushel at a time when the cellulosic credit averaged USD 2.34 a gallon in 2025 produced roughly USD 0.234 a gallon of additional revenue across the plant’s total output, against an estimated USD 0.20 a gallon of added enzyme and marketing cost. That is a net contribution of about six cents a gallon — and against a 2025 industry net margin of USD 0.21 a gallon, it was more than a quarter of the profit.

The strategic reading is uncomfortable for anyone who wants technology to be the story. Cellulosic ethanol became viable not because the conversion economics improved but because a policy instrument paid for the carbon attribute separately from the molecule. The chemistry was ready; the revenue was not. That is a general lesson about this sector, and it will apply again to aviation fuel.

Corn ethanol to jet: the first commercial plant

On 13 November 2025, LanzaJet’s Freedom Pines Fuels facility in Soperton, Georgia produced jet fuel from ethanol at commercial scale and to ASTM specification — the first plant in the world to do so. The configuration is a chain rather than a single step: ethanol is dehydrated to ethylene using Technip Energies’ Hummingbird technology, then oligomerised into jet-range molecules using a process developed with the US Department of Energy and LanzaTech, then fractionated. Capacity is designed at roughly 10 million gallons a year. The company lists deployments under development across the United States, Australia, Japan, India, the United Kingdom, Colombia, the European Union, the Middle East and Kazakhstan.

Two things make this more than a press release, and one makes it less. More: alcohol-to-jet is the only pathway that converts the world’s existing ethanol surplus into a product with aviation’s willingness to pay, and it does so without competing for the waste oils that the hydrotreated-ester pathway needs and cannot get enough of. More: the qualifying pathway table for the clean fuel production credit lists Brazilian sugarcane as an eligible feedstock only for the alcohol-to-jet pathway, which means the policy is deliberately pointing cane ethanol at aviation rather than at blending.

Less: 10 million gallons is 38 million litres. Against 122.9 billion litres of world fuel ethanol it is three hundredths of one percent. The technology has crossed from demonstration to commercial. The industry has not crossed with it yet.

Second generation, at its true scale

Raízen’s 120.2 million litres of E2G output in 2025-26 — double the prior year — is the largest operating cellulosic ethanol programme anywhere, and it is 0.36% of Brazil’s ethanol. The technology works, uses bagasse and straw the mill already owns, and improves the carbon score of the litre it produces. It has not changed the supply curve and will not within this forecast window.

What is not working

Three things deserve to be named, because each has consumed real capital.

Carbon capture at scale in ethanol. The capture technology is proven; the transport and storage network is not. The Iowa main line was permitted in August 2024, three lateral dockets were withdrawn in June 2026, and the surviving hearings are in 2027. Until carbon dioxide can physically move from a plant to a formation, capture equipment at the plant is a stranded asset with a tax credit attached.

Chinese cellulosic ethanol. Two decades of investment have produced no commercial-scale cellulosic capacity, no sustainability criteria, and no lifecycle carbon standard. The country’s fuel ethanol consumption is forecast to fall 11% in 2025 to 4.3 billion litres, and its investment negative list now makes fuel ethanol projects subject to provincial verification. There is no technology race in China to lose.

The electrification of the light-duty fleet. This is not an ethanol technology failure, but it is the substitution that the technology roadmap has no answer for. China’s fuel ethanol consumption is falling because its gasoline pool is shrinking, and Europe’s gasoline demand is growing only because drivers are replacing diesel cars with petrol hybrids. Every ethanol technology described above competes for a molecule whose main market is contracting in the fastest-adopting region.

The honest conclusion is that this sector’s technology frontier has moved from making ethanol to making ethanol worth more — a better carbon score, a specification grade, a jet molecule. The next eight years of innovation here will be about the attribute, not the volume.

Market Risk

Six risks matter in this market. One of them dominates, and it is not the one the industry talks about. The overall risk level here is moderate — demand is legally underwritten and the product has no substitute in its own application — but two of the underlying axes are high: geopolitical exposure, because a third of world volume crosses a border or is governed by a trade instrument, and substitution risk, because the fuel’s largest use case is the thing electrification is aimed at. The distinction between a moderate aggregate and high components is the useful framing, since the components are what move the price.

One: policy duration, and the 2029 date

Every litre of fuel ethanol in this report is sold because a government requires it, subsidises it, or protects it. The mandates differ in design but share one property: none of them is permanent, and several have an explicit end.

The clearest is the American production credit. Under section 45Z as amended, the credit applies to clean transportation fuel produced domestically after 31 December 2024 and sold by 31 December 2029. That is not a phase-down; it is a cliff. A plant’s carbon-intensity investment has to pay back inside a five-year window, which is why the industry’s capital programme is a retrofit programme. The guidance published for calendar 2026 fixes emissions rates through a government model and restricts ethanol to four named pathway-and-feedstock pairs, one of which — Brazilian sugarcane — is permitted only where the ethanol becomes aviation fuel.

The European exposure is the mirror image: a supportive instrument that is not yet law. The Commission’s December 2025 proposal would cut the 2035 fleet target from a 100% to a 90% carbon reduction, with the residual compensable through sustainable renewable fuel credits, capped at 10% of the 2021 fleet-wide target. If adopted, that is the most valuable single instrument this market has been offered in a decade. If it fails in Council or Parliament, the post-2030 European demand line reverts to a declining trend. It is a proposal, and it is being treated here as one.

India’s next step is similarly unlegislated. E20 was delivered in 2025, five years early, and a roadmap beyond 2026 is in formulation. Between now and its publication, the second-largest growth market in this forecast has no stated target.

Two: feedstock and basis risk

Corn is about 80% of the cost of American ethanol. In 2025 the corn price ranged from below USD 4.00 to nearly USD 5.00 a bushel, and plant net margins ranged from minus USD 0.05 to plus USD 0.56 a gallon inside the same twelve months. That is not a margin; it is a distribution. A producer cannot hedge its way out of a 60-cent intra-year range on a product priced at USD 1.60, and the average of USD 0.21 conceals as much as it reveals.

The Brazilian feedstocks carry agronomic risk on top of price risk. Centre-South crush fell to 611.15 million tonnes in 2025/26 from 621.93 million, with the decline attributed to wildfires during the prior crop year affecting regrowth, below-average off-season rainfall, above-average rain at the start of the season and frost. Total recoverable sugar per tonne fell to 137.79 kilograms from 141.09. A mandate can be legislated in a week; a cane field takes three years to respond.

Three: substitution, which is already happening

This is the risk the industry underweights because it is not a biofuel policy question at all. Chinese fuel ethanol consumption is forecast at 4.3 billion litres in 2025, down 11%, and the stated cause is declining gasoline consumption driven by electric-vehicle adoption — not a policy reversal, not a price effect. Europe’s gasoline pool is growing for a structurally fragile reason: drivers replacing diesel cars with petrol hybrids. Every one of those hybrids is a future battery-electric buyer, and each purchase transfers ethanol demand into a powertrain that does not use it.

The December 2025 European retreat on the 2035 target buys the combustion engine time. It does not buy it indefinite life, and it does not change the direction of the Chinese market, which is where the decline is actually occurring.

Four: trade policy

A third of American output’s value depends on access that is negotiated annually. 2025 was a good year — a record 8.4 billion litres and USD 4.7 billion, with the US-UK Prosperity Deal creating a tariff-free quota from 30 June 2025 for 1.4 billion litres, prorated to 913 million that year, and removing a 19% ad valorem tariff. The same instrument closed the two plants that produced virtually all British ethanol within months. Access giveth and taketh away, and it does both from the same document.

The corollary is that the export premium — USD 0.56 a litre FAS against USD 0.42 at the Iowa plant gate — is a political rent as much as an economic one. Any forecast that assumes it persists without assuming the trade relationships that create it is assuming the outcome of a negotiation.

Five: execution risk on carbon intensity

The industry’s value now partly depends on a pipeline network that has been permitted since August 2024 and is not built. Three lateral dockets were withdrawn in June 2026, eight counties were removed from the main route, and the surviving hearings run into 2027. A plant that has installed capture equipment and has no route to a formation holds an asset worth less than its book value, and the credit it was relying on is time-limited to 2029 at the other end.

Six: concentration

Three countries produce 86.4% of the world’s fuel ethanol. A failure in any one of them is a global supply event, and two of the three are agricultural economies whose output is decided by weather. That concentration is also the reason this market has never had a sustained price spike: three independent harvests are harder to fail simultaneously than one.

What the downside is worth

The conservative scenario in this report — no further Indian or Brazilian mandate step, flat American output, continued Chinese decline, industrial grades growing below trend — produces USD 98.2 billion in 2033 against the baseline’s USD 111.8 billion. The gap is USD 13.6 billion, and the optimistic case adds a further USD 30.4 billion above the baseline. In other words, the difference between a market that disappoints and a market that surprises is worth more than twice the downside, which is normally a reason to be long. Here it is a reason to be specific: the optimistic case requires an Indian E30 and a Brazilian E35, and neither has been announced.

Regulatory Landscape

This is the most regulated commodity in the energy complex after tobacco, and the regulation is not a constraint on the market — it is the market. Remove the mandates and roughly three-quarters of the volume in this report has no reason to exist at its current price. What follows is the regime that actually applies, jurisdiction by jurisdiction, and the dates that matter.

The mandate map

The United States: a static mandate and a moving credit

The Renewable Fuel Standard has not delivered a blend step of consequence in years, and it is no longer where the policy action sits. That has moved to the tax code, and the tax code is now more consequential to a plant’s profitability than the obligation it supplies.

Section 45Z provides an income tax credit for clean transportation fuel produced domestically after 31 December 2024 and sold by 31 December 2029. Qualification requires an emissions rate no greater than 50 kilograms of carbon dioxide equivalent per million British thermal units, determined using the credit’s own model rather than the taxpayer’s. The emissions-rate table published for calendar 2026 restricts ethanol to US corn starch, US sorghum grain, US corn stover, and Brazilian sugarcane for the alcohol-to-jet pathway only. The guidance takes effect on and after 8 September 2026, and it follows amendments made in July 2025.

Three features of that design are underappreciated. The credit is measured per unit of fuel and scaled by emissions performance, so two plants making identical ethanol can receive materially different prices. The eligible feedstock list is prescriptive, which makes agronomy and geography part of the tax base. And the sale-date limit of 31 December 2029 creates a five-year investment horizon for equipment with a twenty-year life — which is why the observed response has been retrofits and co-product extraction rather than new capacity.

Brazil: a blend ladder and a tax change

Brazil raised the anhydrous blend in gasoline from 27.5% to 30% with effect from 1 August 2025, and operates a separate E100 market for flex-fuel vehicles — the only large market in the world where a consumer can choose the biofuel or the fossil fuel at the pump on price. The mechanism is elegant and it is also a trap for forecasters: when the mandate rises, mills do not necessarily produce more ethanol, they produce a different grade. In 2025/26 anhydrous output rose 4.22% to 12.89 billion litres while hydrous fell 7.82% to 20.83 billion.

Tax treatment changed in the same year in a way that affects the netback. From 1 May 2025 the federal social contribution on fuel ethanol moved to a single incidence of BRL 192.20 per cubic metre, from BRL 130.90 previously, under the complementary law implementing the 2023 constitutional reform. That is a material per-litre cost on a product selling for roughly BRL 2.70 to 2.90 a litre, and it lands on the producer’s realised price rather than on the consumer’s pump price.

India: administered prices, and a mandate that arrived early

India delivered E20 in 2025, five years ahead of the 2030 target set in the roadmap for 2020-2025, with public-sector oil marketing companies reaching a 19.24% average blend across ESY 2024-25 on 10.4 billion litres of supply. The National Policy on Biofuels, adopted in 2018 and amended in 2022, sets the framework, and an administered price mechanism sets the return.

What comes next is already legislated elsewhere in the same policy: 5% compressed biogas blending in city gas distribution by 2028-29, and sustainable aviation fuel blending in aviation turbine fuel at 1% in 2027, 2% in 2028 and 5% by 2030, initially on international flights. An ethanol roadmap beyond 2026 is in formulation and is the single most valuable undecided policy in this market.

Europe: a cap, a sub-target and a retreat

The Renewable Energy Directive, as revised in 2023, sets an overall renewable target of at least 42.5% by 2030 and a 14% transport target, and it caps conventional crop-based biofuels at 7% of final energy in transport while limiting each member state’s share of crop-based fuel to one percentage point above its 2020 level. Advanced biofuels carry a binding sub-target of 3.5% by 2030. The architecture matters because it means ethanol is capped exactly where it is cheapest, and the growth is pushed toward waste-derived and residue-derived material.

National decisions have already moved demand in both directions. Bioethanol consumption rose 4.5% to 7.21 billion litres in 2025 on gasoline-pool growth and E10 expansion in France, Germany and Poland, while Sweden reduced its blending mandate — the single largest national reduction in the bloc.

Then came December 2025. The Commission proposed replacing the 2035 100% carbon-reduction target for new cars and vans with a 90% target, lowering the 2030 van target from 50% to 40%, and permitting the residual emissions to be compensated through low-carbon steel credits or sustainable renewable fuel credits, capped at 10% of the 2021 fleet-wide target. For the first time, a European instrument contemplates renewable liquid fuel as a compliance route for new vehicles after 2035. It requires approval by the member states and the Parliament, and it is not yet law.

China: the mandate that was announced and abandoned

China announced a nationwide E10 mandate in 2017 for implementation by 2020. It was not implemented. Fifteen provinces maintain E10 mandates and none has been added since 2019; the national average blend rate is 2.1%; and policy language moved from support for grain-based ethanol in 2017 to strict control of capacity expansion from 2020. There are no sustainability criteria, no lifecycle emissions standard, and biofuels are excluded from the emissions trading system. A negative list issued in April 2025 subjects all fuel ethanol investment to provincial verification.

The octane gap that an ethanol mandate would fill is already filled by something else: China’s methyl tertiary butyl ether capacity is estimated at 25 billion litres with an average blend rate of 8%. Ethanol is not competing for that position. It has lost it.

Trade rules and the certification layer

Two regulatory layers sit above the mandates and are becoming more important than them.

The first is trade. The US-UK Prosperity Deal removed a 19% ad valorem tariff on a 1.4 billion litre quota from 30 June 2025 and, within months, the two plants that produced virtually all British ethanol had closed. That is what a tariff decision is worth in this industry: not a margin adjustment, an industry relocation.

The second is carbon certification. A litre is no longer simply a litre; it carries a score, and the score is set by a model, and the model is set by a regulator. The 45Z emissions-rate table, the EU sustainability criteria and the aviation fuel certification regimes all do the same thing — they convert a chemical commodity into a graded one, where the grade is determined in a rulebook. The producers who benefit from the next eight years are the ones who read those rulebooks early enough to build against them.