Sugar Mill Unit Economics in São Paulo: The Margin Stack From Cane to FOB

Greenfield sugar mills in São Paulo can be profitable under the right conditions, but margins are highly sensitive to sugar prices, cane costs, and scale. This report analyzes the margin stack from sugarcane purchase to Free-on-Board (FOB) sugar export, using realistic 2024–2026 data. We find that at mid-cycle prices (raw sugar ≈15–16¢/lb and ATR ≈R$1.25/kg), a 3 million ton/year crush mill in São Paulo generates an EBITDA on the order of R$150–R$200 per ton of sugar (≈US$25–35/t) – an ~8% EBITDA margin{1} {2}. Smaller mills (~1.2 Mt) barely break even at these prices, while mega-mills (~5 Mt) achieve comfortable profits due to economies of scale. A sustained raw sugar price above ~15¢/lb or an ATR below ~R$1.30/kg is needed for most greenfield mills to break even, assuming typical cost structures.

In summary, scale and efficiency are decisive: larger mills can tolerate higher cane prices and lower sugar prices than small mills. Co-generation of electricity from bagasse significantly bolsters margins (often adding ~R$150–R$200 per ton sugar). Benchmarking against industry leaders like São Martinho and Raízen shows that top-quartile Brazilian operations produce raw sugar at a cost around US$330–$350 per ton FOB {3} {1}– the lowest globally. New São Paulo mills can approach these benchmarks with high recoverable sugar (ATR ~140 kg/ton cane) and disciplined cost control. The attached financial model allows investors to test scenarios for sugar price, ATR, exchange rates, logistics, and other inputs, providing insight into EBITDA/ton FOB under various cases. We present below a breakdown of the margin stack from cane to FOB, scenario analysis for 1.2M vs 3M vs 5M ton/year mills, sensitivity charts for break-even, and an investment outlook.

Definitions & Context

ATR (Total Recoverable Sugar) – A measure of sucrose content in cane used in Brazil’s cane payment system. It is expressed in kg of TRS (total recoverable sugar) per ton of cane. In São Paulo, Consecana-SP publishes a monthly ATR price in BRL per kg TRS, which determines cane payments {2}. For example, an ATR price of R$1.20/kg implies roughly R$160 per ton of cane if cane quality is ~135 kg ATR/ton. ATR is analogous to TRS; this report uses the terms interchangeably.

Consecana-SP – The council that sets the cane payment formula in São Paulo. It links cane prices to sugar and ethanol revenues. The Consecana ATR price averaged about R$1.17–1.20/kg in the 2024/25 harvest {2}, equating to ~R$158–168 per ton of cane (at ~135 kg ATR/ton). We use these benchmarks for cane cost assumptions.

VHP Raw Sugar – “Very High Polarization” raw sugar (~ICUMSA 600–1200) is the standard raw sugar for export, priced in USD (ICE #11 world market). We assume FOB Santos pricing based on ICE #11 futures plus local basis. Recent prices ranged from ≈15¢/lb (US$330/t) in late 2025 {4} to >20¢/lb (US$450/t) during 2024 spikes {5}.

White Sugar Premium – São Paulo mills can refine a portion of production to white sugar (ICUMSA <45), which commands a premium. In 2025, Brazil’s FOB refined sugar traded ~$120–140/ton above raw sugar on the world market {4}, though netback premium to mills was lower (~$50–$80/t) after accounting for refining costs. We assume a US$60/t premium for refined sugar in our model, in line with market conditions {6} {7}.

FOB Santos – Free on Board export price at the Port of Santos. This is the price a mill realizes for export sugar, after inland freight and port fees. We assume inland logistics + port costs ~R$120 per ton sugar (trucking/railing to port, elevation, storage, etc.). For context, trucking sugar ~300 km to Santos costs on the order of R$80–90/t, and port handling around R$30–40/t, totaling ~R$120/t as used.

EBITDA – Earnings before interest, tax, depreciation, and amortization. In this analysis we focus on EBITDA per ton of sugar (FOB) as the key unit economics metric. Positive EBITDA/ton indicates the operation covers cash costs (cane, processing, overhead) and contributes to fixed cost recovery and profit. All figures are real 2025 BRL unless stated, with USD conversions at the prevailing BCB PTAX exchange (approximately R$5.2–5.8 per US$ in 2024–25 {8} {9}.

Seasonality & Working Capital – The Center-South harvest runs April to November. Mills crush cane and produce sugar mostly in these months, but sales revenue may be distributed year-round. This creates a working capital need to finance inventory and operations during and shortly after the harvest. We will discuss the impact of carrying sugar inventories (which can average 1–2 months of production) and receivables on cash flow.

Assumptions Summary

The financial model (downloadable Excel) is built for a greenfield mill in São Paulo starting in the 2025/26 harvest. Key assumptions are:

• Scale Scenarios: 1.2 million, 3 million, and 5 million tons cane per year (t/year). All scenarios assume 210 crushing days/year (typical seasonal operation). This implies roughly 5,700 t/day capacity for the 1.2M scenario, 14,300 t/day for 3M, and 23,800 t/day for 5M.
• Cane Quality: ATR (TRS) content of 135 kg per ton of cane (close to the 2023/24 CS Brazil average ~138–141 kg {10} {11}. Industrial recovery of ATR to sugar is assumed at 92%, meaning about 124 kg of sugar per ton of cane is produced. This yield (12.4%) translates to ~8.0 tons of cane per ton of sugar. It is achievable with modern equipment; UNICA reported CS 2025/26 ATR around 142 kg/t (a good year){12}.
• Product Mix: We assume a mix of 60% raw sugar and 40% white sugar in the sugar production portfolio (0% ethanol, since we focus on a fully sugar-oriented mill). This reflects many new mills maximizing sugar production when economics favor sugar over ethanol (as was the case with a ~52% sugar mix in CS Brazil 2025 {13} {14}. The white sugar portion incurs additional refining cost but earns a premium as noted.
• Pricing: Base case uses NY#11 raw sugar ≈15.9¢/lb (US$350/ton) and a white sugar premium of US$60/t, consistent with late-2024/early-2025 market levels {4} {7}. The USD/BRL is assumed at 5.58 (average early 2025) [8]. Domestic crystal sugar in São Paulo was around R$130–145 per 50kg in late 2024 [15] [16], which aligns with these FOB export parity prices. We will also examine downside (14¢) and upside (20¢) sugar scenarios.
• Cane Cost: Based on Consecana-SP ATR price ~R$1.25/kg, roughly the early 2025 level [17] [2]. At 135 kg ATR/ton, this is ~R$169 per ton of cane paid to suppliers (or the internal opportunity cost for owned cane). The model treats cane cost as variable, directly tied to ATR and thus sugar/ethanol prices. (Notably, if sugar prices rise, ATR price will rise as mills pay growers more; this linkage is considered in scenario discussion.)
• Processing Costs: Non-cane processing opex of R$55 per ton cane (includes labor, maintenance, chemicals, etc. for the factory). Fixed overhead (management, administration) is assumed at R$40 million/year for a 1.2M t mill, scaled up modestly for larger mills. This equates to ~R$33/t cane at 1.2M, improving to ~R$20/t at 3M and ~R$16/t at 5M (economies of scale in overhead). These figures are in line with industry cost breakdowns and reflect that larger operations spread fixed costs over more tonnage.
• Co-generation: The mill exports surplus bagasse electricity. A modern medium-sized plant might export ~80 kWh per ton of cane crushed. We assume this for the 3M case; smaller mills might export a bit less (some may only cover internal needs), while a state-of-the-art large mill could reach 100+ kWh/t cane by using high-pressure boilers and cane trash. The power sales price is assumed at R$280/MWh (≈R$0.28/kWh) – representative of Brazil spot prices in recent years. Thus, ~80 kWh/t yields ~R$22/t cane in energy revenue. This credit is important to the margin stack.
• Logistics and Taxes: Inland freight + FOB handling cost is R$120/t sugar (as noted). We assume no export tax (Brazil does not tax sugar exports; PIS/Cofins are zero-rated for exports). We also ignore ICMS since sugar for export is exempt; for domestic sales, ICMS/PIS would effectively reduce the net price realized[18]. Thus, the export margin stack is analyzed pre-tax. Corporate income tax and financing costs are considered in the investment conclusion but are outside the unit cash cost scope.

These assumptions represent a mid-cycle baseline. The attached model allows changing all key inputs (sugar price, ATR, FX, yields, costs, capacity, capex, etc.) to stress test the economics. Next, we break down the margin stack under base conditions.

Margin Stack Breakdown: Cane to FOB

Under our mid-cycle base case (NY#11 ≈15.9¢, ATR R$1.25, USD/BRL 5.58), a 3M t/year São Paulo mill producing 60% raw and 40% white sugar would realize approximately R$1,970 per ton of sugar net FOB (after freight to port). From this, all costs are paid, leaving an EBITDA margin of roughly R$162 per ton sugar, or about US$29/t (≈8% of FOB value). Below is the approximate per-ton margin stack:

• Gross FOB Price: ~R$2,090 per ton sugar. This corresponds to a blend of raw and white export prices. For example, raw sugar at 15.9¢/lb = ~$350/t, white sugar ≈$410/t (with premium)[6]. Weighted 60/40 and converted at 5.58 BRL/USD, we get ~R$2,090/t. (For reference, domestic São Paulo crystal prices were in this range – ~R$2.6/kg or R$130/50kg in early 2025[15], equivalent to ~R$2,600/t, slightly higher due to local market factors and taxes).
• Less: Inland Logistics: ~R$120/t. This includes cane-to-port transport and port fees, bringing the net mill gate price to about R$1,970/t FOB.
• Less: Cane Cost: ~R$1,356 per ton sugar. This is the dominant cost, ~69% of net revenue. It comes from paying ~R$168/t cane[2] with ~8.05 tons of cane needed per ton of sugar (given 124 kg sugar/t cane recovery). We derived this using the Consecana ATR price (R$1.2478/kg) for the period[17]. In other words, at 15.9¢ sugar, ~R$1.25 ATR, the cane cost is nearly R$1.36 per kg of sugar produced.
• Less: Processing & Overhead: ~R$604 per ton sugar. This includes all non-cane operating costs. Variable factory costs (~R$55/t cane for labor, fuel, maintenance, chemicals) contribute ~R$443/t sugar, and fixed overhead (~R$20/t cane at 3M scale) adds ~R$161/t sugar. Notably, the overhead portion would be higher (≈R$270/t) for a small 1.2M mill and lower (≈R$130/t) for a 5M operation, illustrating scale efficiency.
• Less: Refining Cost (for white): ~R$28 per ton sugar (blended). White sugar incurs additional refining expenses (power, filtration, etc.), assumed ~R$70/t for the white portion. With 40% of output refined, the average cost is R$28/t across all sugar. If the mill produced only raw sugar, this line would be zero – but then we would also lose the premium in price. In our scenario the premium (~R$330/t in FOB price) far outweighs the refining cost, so refining 40% of output is net-profitable.
• Add: Co-generation Credit: +R$180 per ton sugar. Selling electricity from bagasse contributes additional revenue. At 80 kWh/t cane surplus and R$280/MWh, each ton of cane yields ~R$22.4 in power sales. Per ton of sugar (8.05 tons cane), that’s ~R$180 added revenue. This offsets ~30% of the processing + overhead costs – a significant boost. Mills like São Martinho export on the order of 700 GWh from ~23 Mt crush, about 30 kWh/t; newer mills with bigger boilers can do considerably more.

After these, we arrive at EBITDA ≈ R$162 per ton sugar (i.e. 1,970 – 1,356 – 604 – 28 + 180). That is roughly US$29/t at 5.58 BRL/USD, equating to an EBITDA margin of ~8% on FOB revenue. Cane cost is by far the largest component, around 2/3 of total cost, followed by processing costs. The margin stack is sensitive: a 10% move in sugar price or ATR can swing the EBITDA from positive to negative for a small mill.

It’s worth noting that export-focused mills avoid domestic sales taxes – if this same sugar were sold internally in Brazil, the mill might face ICMS/PIS/COFINS taxes that could exceed R$200/t (though some exemptions exist for sugar). By exporting (FOB), the mill realizes the full market price{15} {16}. This highlights Brazil’s incentive to export sugar and the importance of efficient logistics to get sugar from interior São Paulo to the port at low cost.

For comparison, global cost benchmarks show Brazil’s cost competitiveness: Czarnikow estimates Brazil’s average raw sugar production cost at ~US$335/t FOB in 2024/25 {3} {19} – which matches our modeled mid-cycle cost (~$320/t total cost, $350/t price, yielding ~$30/t margin). Brazil sits at the low end globally; India is around $420/t (with subsidies) and the EU $750+t (as refined)[20] [21]. Within Brazil, top-tier operators achieve even lower costs. São Martinho, for instance, reported R$1.17/kg TRS average price for cane in 2024/25[22], and they crushed 23.1 Mt cane to produce ~1.5 Mt sugar and 700 GWh power, with an adjusted EBITDA margin around 45% in that crop-year (including ethanol) – indicating a cash cost well below 15¢/lb. Raízen, the largest player, crushed a massive 30.9 Mt in just Q1 2024/25 with ATR ~124 kg/t (early in season), and leverages scale to minimize unit costs. These benchmarks underscore that efficient São Paulo mills can thrive at mid-cycle prices, whereas smaller or less efficient operations might struggle without higher sugar prices or excellent agro productivity.

Scenario Analysis: 1.2M vs 3M vs 5M Tonne Mills

To illustrate the impact of scale, we compare unit economics for three greenfield mill sizes:

• Small Mill – 1.2M t/year: A single-train mill on the smaller end for Brazil. Perhaps an independent mill or new co-op venture. We assume overhead R$40M (as above) and slightly less efficient co-gen (~50 kWh/t cane exportable). This mill would produce roughly 150,000 t of sugar per year.
• Mid-size Mill – 3M t/year: A more typical large mill or expansion project (producing ~375,000 t sugar/year). Overhead is higher in absolute terms (assume ~R$60M) but lower per tonne. Co-gen ~80 kWh/t cane (base assumption).
• Mega Mill – 5M t/year: A very large scale operation (common for integrated groups like Raízen or Cofco in Brazil, often as a cluster of plants). Produces ~620,000 t sugar/year. We assume overhead ~R$80M and co-gen ~100 kWh/t (leveraging scale and perhaps straw feeding). Many new projects may not start at 5M, but could reach this with expansion; we include it for strategic comparison.

Using the base prices (15.9¢, R$1.25 ATR, 5.58 FX), the EBITDA per ton for each scenario comes out to approximately:

• 1.2M t Mill: ~R$55/t sugar EBITDA (only ~US$10/t). This is essentially break-even, a margin of just ~3%. The small mill barely covers costs at mid-cycle pricing. Any adverse shift (e.g. slightly lower sugar price or higher cane cost) would push it into the red. It relies on either above-average cane yields or price upside to be worthwhile.
• 3M t Mill: ~R$162/t (US$29) EBITDA, as detailed. Margin ~8%. This suggests a viable project – indeed many existing mills operate around this scale and remained profitable in 2024/25 when sugar averaged ~17¢/lb and ATR ~R$1.17 {2} {5}. The mid-size scenario represents a typical efficient mill in São Paulo.
• 5M t Mill: ~R$194/t (US$34) EBITDA, or ~9–10% margin. The large mill’s lower unit costs (especially overhead per ton and better energy credit) give it roughly 20% higher EBITDA/ton than the 3M case under identical prices. This cushion means a big mill could withstand lower sugar prices better. Many newer projects aim for scale (sometimes by integrating nearby smaller mills) to capture these economies.

In practice, not only overhead but also conversion efficiencies improve with scale: bigger mills often invest in more advanced technology (higher pressure boilers, better recovery rates, etc.). We did not assume a higher sugar recovery for the 5M (keeping 92%), but a top-tier large mill could perhaps achieve 93–94% recovery or utilize cane residuals more fully, adding further to margin.

Refinery Uplift: All scenarios assumed 40% refined sugar production. A greenfield project can choose to include a refinery or produce only VHP raw. Including a refinery adds capital cost and operating cost, but as long as the white premium stays above the refining cost, it yields a net margin benefit. In our case, premium $60 minus refining ~$12 (70 BRL) gives +$48/t on 40% of production, boosting total EBITDA by ~$19/t. If the white premium collapsed or refiners overseas had an advantage, the mill could revert to raws. The flexibility to switch product mix is valuable. São Paulo refineries (producing Icumsa 45) often enjoy premiums in markets like Africa or the Middle East, as seen in 2024’s export patterns (e.g. Algeria and UAE taking refined sugar) {4} {23}.

Ethanol Option: We have focused on sugar, but Brazilian mills are dual-capable. In a down-cycle for sugar (or up-cycle for ethanol), a mill might produce more ethanol to preserve margins. However, our “greenfield sugar mill” premise assumes investors are targeting sugar output (perhaps counting on export markets). Still, it’s important to note that ethanol provides a floor via the sugar-ethanol parity. In 2025, despite sugar-ethanol price convergence, mills stuck with sugar to maximize revenue {13} {24}. A new mill would likely do the same unless ethanol’s economics clearly dominate. For now, we assume sugar-max configuration (over 50% mix), which in recent years has been the profit-maximizing choice.

Break-Even Sensitivities

What sugar price or ATR level makes a new mill profitable? We stress-tested the scenarios for break-even points:

Sugar Price Sensitivity: Holding ATR at ~R$1.25 and other costs constant, the 1.2M t mill requires roughly 15.5¢/lb raw sugar to hit EBITDA breakeven. The 3M and 5M mills break even at lower prices, around 14.6¢/lb and 14.3¢/lb respectively. In other words, a large operation could survive in a 14¢ world, while a small mill would be underwater below ~15¢. This is illustrated in the chart below.

needs ~15.5¢/lb to break even, whereas a large mill (red) can break even closer to 14.3¢. Note: In reality, low sugar prices would also reduce ATR (cane cost), partially mitigating the impact – here we isolate the price effect {2} {1} {3}.

As the chart shows, EBITDA per ton increases linearly with sugar price. At 18–20¢, all scales see healthy margins (e.g. at 18¢, ~R$200/t for small mill, R$300+ for large). Conversely, at 13–14¢, only the biggest mills remain slightly profitable. This sensitivity aligns with industry cost curves – Brazil’s lowest-cost producers can tolerate prices in the 12–14¢ range (with some co-gen and ethanol credits), but higher-cost and smaller mills might shut production if world prices approach those lows for long {3} {1}.

ATR (Cane Cost) Sensitivity: Keeping sugar around mid-cycle (~16¢/lb), we vary the ATR price (which proxies cane cost). Higher ATR means mills pay growers more per ton of cane. A large mill can afford to pay up to ~R$1.43/kg ATR before losing margin at 16¢ sugar, whereas a small mill is breakeven around ~R$1.29/kg. The 3M case is in between, ~R$1.37/kg.

In practice, ATR and sugar prices are correlated by the Consecana formula. If world sugar spikes to 20¢+, ATR will rise (growers capture ~60% of revenue). Conversely, at 14¢ sugar, ATR drops, providing cost relief. Thus, actual breakeven dynamics are a bit buffered: a price crash is partially offset by cheaper cane, and a price boom is shared with farmers. Our sensitivity charts above hold one factor fixed to show the relationship. The key takeaway is that scale extends the breakeven envelope – larger mills have more room to absorb price or cost shocks before dipping into negative EBITDA.

Working Capital & Seasonality: Another critical aspect is the cash flow timing. A sugar mill’s expenses (cane payments, wages, maintenance) are front-loaded in harvest months, while some sales revenue comes later. For instance, a mill may build up 2 months of sugar inventory by end of crush and have typical payment terms of 30–60 days on shipments. Meanwhile, it pays farmers typically within 15 days of cane delivery. This mismatch can create a peak working capital requirement on the order of 2–3 months of COGS. In our 3M example, that equates to roughly R$150–200 million (~US$30–40M) needing financing at peak (just after harvest). Smaller mills would need less absolute WC (maybe ~R$50M), and larger mills more (R$300M+), but as a percentage of revenue it’s similar. High Brazilian interest rates (e.g. ~12% Selic in 2025) mean carrying this inventory isn’t cheap – it can shave off a few percent of margins if not managed.

Mills commonly use strategies to ease this strain: bank credit lines and sugar forward sales. For example, they secure ACC/ACE (advance on export contracts) loans or use crop-backed loans to get cash earlier for their sugar that’s in inventory or transit. Additionally, some will coordinate ethanol sales (which can be sold from tanks later in the year) to generate cash when sugar inventories are high. Our model’s working capital module indicates ~75 net working capital days in a typical cycle (60 days inventory + 30 AR – 15 AP), which matches industry norms. An investor should ensure the project has sufficient liquidity or credit to finance ~R$0.1 to 0.2 per R$1 of annual revenue during the peak (e.g. a 3M mill with ~R$1.5B revenue might need ~R$150–200M). The cost of this financing effectively adds a few reais to the cost per ton of sugar if annualized.

Benchmarking Real-World Operators

To sanity-check our unit economics, we compare to leading São Paulo mill operators:

• São Martinho: One of Brazil’s most efficient groups. In the 2023/24 crop, São Martinho crushed 23.1 million tons of cane, producing ~1.5 Mt of sugar (and ethanol from the remainder). This implies an average sugar mix around 50%, which is consistent with an ATR of ~141 kg/t and sugar yield ~6.5% (the rest to ethanol). They sold ~700 GWh of power (≈30 kWh/t). Their reported cane yield and TRS were above industry average (they often achieve >140 kg ATR). São Martinho’s cost of production is not public in detail, but their financials show healthy margins even at mid-cycle prices – a testament to low unit costs. Using our model, a São Martinho-type operation (large scale, high ATR, high co-gen) would likely see EBITDA/ton well above R$200 at mid prices, which aligns with their ~44% EBITDA margin in the sugar/ethanol segment in 2023 (when sugar averaged ~18¢)[3] [1].
• Raízen: The world’s largest sugarcane processor (a JV of Cosan and Shell). In Q1 of 2024/25 (Apr–Jun 2024), Raízen crushed 30.9 Mt of cane – an astonishing volume in 3 months, albeit across many mills. Their average ATR was 124.2 kg/t in that quarter (early season, lower than peak). Raízen’s scale gives it tremendous logistical and purchasing advantages. They reported an adjusted EBITDA of R$2.313 billion in that quarter (across sugar, ethanol, energy and trading segments). While not directly comparable, if one attributes, say, half of that to sugar, and consider the sugar volume exported, the implied EBITDA per ton is in the same ballpark of a few hundred reais. Raízen’s cost per ton of sugar is likely among the lowest – possibly approaching R$1,100–1,200/t ($200–220) cash cost, given their integrated model and ethanol hedging, which would be ~13–14¢/lb. This aligns with our large-mill scenario and the Czarnikow estimate of ~15¢ for Brazil’s cost floor {25} {26}.
• International Comparison: Brazilian mills’ EBITDA/ton can swing widely by year. In boom years (high sugar price, good yields) it’s not uncommon to see >US$100/t profit. In bust years (low price or poor harvest), even large mills can approach breakeven. For context, in 2020 when sugar prices were low (~10–12¢), many mills barely broke even or incurred losses, highlighting the importance of being at the low end of the cost curve. By contrast, a Thai or Indian mill might have higher costs – needing perhaps 18–20¢ for breakeven without subsidies{27}. European beet factories need even more (some >$0.35/kg for refined sugar){20} {28}. Thus, a well-run São Paulo greenfield can be globally competitive, but only if it hits the benchmarks on ATR, conversion efficiency, and scale.

Investment Outlook and Conclusion

Profitability: Given the above analysis, a greenfield sugar mill in São Paulo can generate an EBITDA margin in the high single digits at today’s prices (mid-2020s). In numeric terms, expect US$25–$40 EBITDA per ton of sugar in an average year for a large operation – which on a ~0.6 million ton sugar output (for a 5M crush) is ~$15–$24 million EBITDA annually. Smaller mills would see substantially less. Over a price cycle, margins could expand to >15% in peak price years, or vanish in trough years. The breakeven NY#11 price for a typical new mill is around 14–16¢/lb (assuming no ethanol pivot). This is below the World Bank’s forecasted average price of ~17¢ for 2025–26{5}, suggesting a modest safety cushion. However, it’s tight enough that efficient operations and risk management are critical.

Key Risks:

• Market Prices: Sugar is volatile. A sustained drop below 14¢ (or sharp BRL appreciation) would squeeze even the best mills. Conversely, upside price scenarios (like 18–20¢) offer a windfall – our model shows EBITDA/ton doubling from mid to high scenario. Investors should consider hedging strategies for sugar (and perhaps FX) to lock in acceptable margins for the initial years.
• Cane Supply & ATR: The model assumes ample cane at 135 ATR. Weather or agronomic issues (drought, frost, poor renewal) can lower yields and ATR, raising unit costs. Notably, 2024 saw lower cane yields ~75 t/ha in CS Brazil (vs 80+ normally){29}, partly offset by slightly higher ATR ~142. A new mill must secure sufficient cane through cultivation or supplier contracts. If cane shortfall occurs, fixed costs get distributed over less sugar, hurting economics.
• Cane Pricing Formula: Consecana is reviewing its formula as of 2025{30}. Any change could alter cane cost dynamics. For example, if ethanol’s weight increases or if new formulas guarantee higher payout to farmers, mill margins could be pressured. So far, Consecana has balanced mill and grower interests, but it’s something to watch.
• Logistics & Export Capacity: São Paulo’s advantage is proximity to Santos, but infrastructure constraints can arise (e.g. 2025 saw port queues 30+ days for sugar at peak export rush[31]). Extra logistics costs (demurrage, storage) can chip away at FOB netbacks. Our assumed R$120/t might rise if fuel prices spike or if port fees increase. A greenfield project may need investment in own logistics (trucks, terminals) to mitigate this.
• Exchange Rate: A weaker BRL generally helps sugar exporters by lowering local costs relative to USD revenue (as seen in 2024 when BRL devalued ~13% and effectively reduced costs in USD[1]). However, FX also affects imported capex and debt servicing. We assumed ~5.5 BRL/USD; if BRL strengthens to, say, 4.5, the same world price yields fewer Reais and pinches margins ~20% (unless cane price also adjusts down via Consecana).

Financing and Returns: The upfront capex for a new 3M t mill (with cogeneration and refinery) can easily be in the few hundred million USD range. Returns on investment will depend on long-term sugar/ethanol prices and carbon credits (if any for surplus power or ethanol). Our analysis shows mid-cycle EBITDA margins <10%, which, after depreciation and financing costs, might translate to low single-digit net margins. This means that equity IRR for a greenfield project will likely rely on periods of above-average prices or operational outperformance. That said, strategic investors (e.g. integrated commodity players or energy-offtakers) often target synergies and longer-term positioning. Brazil’s efficient producers like São Martinho trade at healthy valuations, reflecting expectations of sustained cash generation. A new project, with the benefit of modern design, could achieve similar or better unit costs – but ramp-up and attaining promised efficiencies is the execution challenge.

Conclusion: Building a greenfield sugar mill in São Paulo is a viable endeavor if planned with realistic economics. The margin stack from cane to FOB reveals slim but positive margins in normal conditions, heavily contingent on cane cost management and achieving scale. Key success factors include: (1) securing high-yielding cane supply (both quantity and quality ATR), (2) leveraging co-products like electricity (which provided ~R$180/t sugar credit in our base case), {3} optimizing logistics to Santos, and {4} maintaining flexibility between raw vs white sugar (and ethanol) to capture the best market value. Large, efficient mills can operate profitably at prices that would put smaller mills out of business – hence, consolidation and expansion have been prominent in Brazil (e.g., Raízen expanding crush volume in search of cost per ton reduction).

Investors and development finance institutions (DFIs) looking at such projects should stress test their models with conservative scenarios (e.g. sugar at 13¢, BRL stronger at 5.0, or interest rates high during inventory carry). The attached financial model (link below) provides a tool to do this. Our scenario analysis indicates the project NPV is most sensitive to the long-term sugar price and exchange rate assumptions, followed by cane availability (capacity utilization). On the upside, if global sugar remains in deficit (as in 2024–25)[5] or if the mill can monetize more value (e.g. ethanol during lows, carbon credits for bioenergy), the returns improve significantly.

Overall, São Paulo’s sugar sector remains a cornerstone of global sugar supply – a greenfield mill, when executed and operated well, can slot into the lower end of the global cost curve. The margin from cane to FOB may not be large on a per-ton basis at mid prices, but over millions of tons it underpins a solid business that can weather the volatility of commodity cycles better than most competitors worldwide.

Appendix: Sources and Methodology

Sources (2024–2026)

All figures in this report are grounded in recent public data and are cited inline. Primary sources include:

• Consecana-SP: ATR pricing benchmarks and cane payment framework
• CEPEA/ESALQ: São Paulo sugar indicators and export parity references
• UNICA: Center-South production metrics, ATR ranges, and sugar mix context
• USDA FAS (GAIN): exchange rate context, production forecasts, and sector benchmarks
• Czarnikow (CZAPP): global cost benchmarks and comparative cost curves

Methodology

The margin stack is computed per ton of sugar at FOB by converting each cost or credit into the same unit:

• Cane cost per ton of sugar = (ATR price in R$/kg × ATR kg per ton cane) × (tons of cane per ton of sugar)
• Energy credit per ton of sugar = (exportable kWh per ton cane × R$/kWh) × (tons of cane per ton of sugar)

All scenarios assume full utilization at the stated crush capacity. Fixed overhead is allocated fully across output. In practice, ramp-up and commissioning can reduce utilization in year 1 and compress early-year returns.

Working capital and seasonality

Working capital is estimated using inventory, receivables, and payables day assumptions applied to annual COGS. Because production is concentrated in the April to November harvest, cash needs can peak during and shortly after the crush when inventories and receivables are elevated.

Currency and FX

The model uses the PTAX exchange rate as the baseline reference and treats FX as a key sensitivity. In general, a weaker BRL increases BRL-denominated revenue relative to BRL-denominated costs, improving EBITDA per ton, subject to Consecana-linked cane pricing and domestic inflation dynamics.

Validation checks

Outputs were cross-checked against widely cited industry benchmarks, including Brazil’s mid-cycle cost positioning on the global curve, observed ATR price ranges, and typical industrial yields. Where possible, results were sanity-checked against the performance envelope of leading Center-South operators.