Editorial Article
Feed Is 62.5 Per Cent of the Bill. That Is Why the Mill Cannot Raise Its Price
In July 2026, a broiler producer in Paraná spent R$4.77 to put a kilogram of live weight on a bird. 62.51% of that was feed. One-day chicks were 19.83%. Everything else — labour, energy, housing, health, finance, the lot — shared what was left. In Santa Catarina the same month, a pig producer’s cost was R$6.29 a kilogram and feed was 72.52% of it.
Those two numbers come from Embrapa’s CIAS cost indices, and they are the most useful figures published about the Latin American feed industry. Not because they size the market — they don’t — but because they describe the negotiating position of every feed mill in the region. When your product is two-thirds to three-quarters of your customer’s cost, you are selling into the least price-elastic demand in agriculture and you still cannot earn a margin on it. That contradiction is the market.
Cost passes through. Margin does not.
A feed mill can recover a rise in the price of corn, because its customer has no substitute for energy in a ration. What it cannot recover is anything above that: a technical service, a formulation advantage, a brand, a relationship. The customer will pay the cost of the ration and will not pay a premium on top of it, because the customer’s alternative to paying more is to reformulate — and reformulating is cheaper than paying.
This is why feed businesses are structurally large, thin and cash-hungry, and why the industry’s consolidation has happened at the ingredient layer rather than at the mill. A mill that earns a cost pass-through and nothing else has to earn its return on volume, logistics and working capital. Every commercial decision in the region follows from that.
There is a second-order effect that is easy to miss. Because feed is 62–73% of the customer’s cost, the customer’s own margin is volatile in a way that makes long contracts unattractive to both sides. A three-year feed supply agreement signed in a low-grain year is a bet on corn, not a service contract. So the region’s feed trade is transacted in short cycles, with quarterly contracting and index-linked pricing — which is another way of saying that the mill has transferred the risk it cannot hedge onto the contract structure.
The other 37.5% is where the industry’s profit went
If a mill cannot earn above cost on the ration, the money migrates to whoever reduces the ratio. That is the entire commercial logic of the additive and premix layer, and it explains why those businesses grow faster than the tonnage they sit inside.
A nutritionist who takes one percentage point of soybean meal out of a broiler ration is worth more to a Brazilian integrator than a percentage point of market share, because the saving is booked monthly against a cost line that is 62.5% of the total. Phytase, synthetic amino acids, alternative proteins and gut-health products are all sold against that arithmetic rather than against the price of a tonne. This is the mechanism developed in Lysine Below Cost — molecules that grow faster than the market they sit in, while the molecules themselves trade at or below the cost of production.
It is also the wedge embedded in this market’s forecast. Latin American feed volume grows 2.40% a year to 2033, from 204.4 to 247.2 million tonnes. Value grows 4.00%, from USD 91.7 billion to USD 125.5 billion. The 1.6-point gap is not inflation: it is additive intensity per tonne rising, aquaculture and pet feed taking share from commodity broiler rations, and realised revenue per tonne moving from USD 448.45 to USD 507.61 — a 1.56% CAGR. If that mix shift does not happen, and revenue per tonne stays flat for eight years, 2033 value is USD 110.9 billion, not USD 125.5 billion. One assumption, a 12% swing.
The demand subsidy that expires
2025’s volumes were bought with cheap grain. Brazil’s 2025/26 harvest reached a record 361.7 million tonnes (+2.6%), including 180.4 million tonnes of soybeans (+5.2%), and the region’s mills priced their rations against it. That is a demand subsidy, and like all subsidies it has an expiry.
When corn and soybean meal revert toward their means, the feed cost line rises before the protein price does — because the producer’s selling price is set by a market where supply responds with a lag, while the mill’s input cost resets every quarter. The producer’s response to a squeezed margin is not to buy cheaper feed. It is to cut stocking density, extend the finishing period, or skip a cycle. That means the volume effect arrives one to two quarters after the cost effect, and it is negative in both directions at once: fewer tonnes sold and a customer base under stress.
This lag is the region’s main cyclical risk, and no forecast that extrapolates 2025 contains it. It is also asymmetric: grain falling is good for volumes within a quarter, grain rising is bad for volumes with a two-quarter delay, so a full cycle leaves the industry with less volume than the average of the two halves.
Why the industry is being built from both ends at once
The ratio explains the shape of the capital being spent in the region, and the two halves of that spending are mirror images of each other.
- Integrated protein companies build captive mills to control the 62.5%. BRF’s merger into Marfrig — cleared by CADE on 5 September 2025 and renamed MBRF Global Foods — put poultry, swine and beef feed buying inside one balance sheet. JBS and Minerva run their own feed and supplement capacity for the same reason. When feed is two-thirds to three-quarters of your cost, owning the mill is a margin strategy, not a supply strategy.
- Additive and premix companies build local plants to sell into the 37.5%. dsm-firmenich opened a 100,000 tonne-a-year beef and dairy supplement plant at Sete Lagoas in October 2024, on 12,000 square metres of built area; ADM lifted production at its Paraná nutrient facility by 40%. Neither investment is a bet on tonnage. Both are bets that additive intensity per tonne rises faster than the feed pile itself.
An integrator with its own mill can buy corn at harvest, store it, and price its own feed below the merchant market for as long as its balance sheet allows. That is the competitive weapon in this industry, and it has nothing to do with nutrition.
What would change the number
Three things are worth watching, because each would falsify a piece of the argument above.
- Feed’s share of producer cost falling below 55%. That would require a step change in feed conversion or a grain collapse, and it is the one condition under which mills could begin earning a margin above cost pass-through. Nothing in the current data suggests it.
- Additive intensity stalling. If premix and supplement volumes stop outgrowing the feed pile, the value CAGR converges on the volume CAGR and the market’s 2033 value lands nearer USD 110 billion than USD 125 billion.
- A quarter in which feed cost falls and mill margins do not expand. That would show the ratio is not the binding constraint, and that something else — logistics, working capital, customer concentration — is setting the price.
The monthly Embrapa CIAS index is the only frequent read on this ratio published anywhere in the region. For a market that is otherwise measured once a year, mill by mill, it is worth more attention than it gets.
The full cost structure, the country and species splits, and the forecast behind these numbers are set out in Latin America Animal Feed Market Size, Share & Forecast 2026–2033, under the Animal Nutrition & Feed segment. The input side of the same cost line is sized in the Global Feed Phosphates report.
Companion analysis. The geography behind this cost base is in Brazil Is 44 Per Cent of the Region and Mexico Buys Its Corn in Dollars, and the corporate consolidation is in The Meat Merger That Was Really a Feed Order. The full sizing and the valuation method are in the press release, Latin America’s Feed Market to Reach USD 125.5 Billion by 2033.
Frequently Asked Questions
What share of a Brazilian producer's cost is feed?
Embrapa's CIAS cost indices put feed at 62.51% of a Paraná broiler producer's total cost in July 2026, when the cost of production was R$4.77 per kilogram of live weight and one-day chicks were a further 19.83%. For a Santa Catarina pig producer the same month, feed was 72.52% of a R$6.29 per kilogram cost. No other single line item in either operation exceeds 20%.
Why can't a feed mill earn a margin when feed is most of its customer's cost?
Because the customer can reformulate more cheaply than it can pay. When feed is two-thirds to three-quarters of a producer's cost, its first response to a price rise is to change the ration, not to accept the increase. A mill therefore recovers the cost of grain and nothing above it, which caps what it can earn from formulation and pushes it toward services, additives and financing instead.
Why does the value of the Latin American feed market grow faster than its tonnage?
The forecast separates them deliberately. Volume grows 2.40% a year from 204.4 million tonnes in 2025 to 247.2 million tonnes in 2033, while value grows 4.00% from USD 91.7 billion to USD 125.5 billion. The 1.6-point gap is additive intensity per tonne, aquaculture and pet feed taking share from commodity broiler rations, and realised revenue per tonne rising from USD 448.45 to USD 507.61. If revenue per tonne stayed flat, 2033 value would be USD 110.9 billion instead — a 12% difference.
What is the biggest cyclical risk to Latin American feed demand?
Record grain. Brazil's 2025/26 crop reached 361.7 million tonnes, up 2.6%, including 180.4 million tonnes of soybeans. Those volumes were bought with an unusually low feed cost base. When corn and soybean meal revert toward their means, the mill's cost line rises before the producer's selling price does, and the producer's response is to cut stocking density or skip a cycle rather than buy cheaper feed — a volume effect that arrives one to two quarters after the cost effect.