What is a buyer paying for when a hundred-year-old coffee brand changes hands at less than one times revenue? In November 2026, Kraft Heinz, the US packaged-food group, agreed to sell its North American coffee business — Maxwell House, Gevalia, Tassimo and four other brands — to the private equity firm One Rock Capital Partners at an enterprise value (EV, the price of the whole business including debt) of USD 650 million. Read as a brand deal, the number is an obituary. Read as a Maxwell House carve-out — the separation of a division from its corporate parent into a standalone company — built on a plant and a calendar, it is something else.
The desk’s shorthand is a bakery. A century-old shop has lost its younger customers to cafés and supermarket own-labels, and last year’s record flour price is still sitting in this year’s books. The owner lists the shop on those books. The buyer has not come for the bread; he has come for the only large oven within a hundred kilometres, and for a harvest that suggests flour will be cheaper by the time he gets the keys.
Only five facts about this transaction are public. Everything beyond them in this note is the desk’s reconstruction of what a disciplined buyer would have to build on top, and it is labelled as such. One Rock’s actual financing and contract terms are not disclosed.
The Deal: A Maxwell House Carve-Out at 0.78x Revenue
The target is the North American coffee division of Kraft Heinz; the buyer is One Rock Capital Partners. The announcement discloses five things and nothing else.
- An enterprise value of USD 650 million.
- A 30% non-controlling stake retained by Kraft Heinz.
- After-tax net cash proceeds of approximately USD 500 million, to be used for debt reduction.
- Closing expected in mid-2027.
- The transfer of roughly 250 functional employees across supply chain, sales, marketing, finance, HR and R&D, together with two plants: Jacksonville, Florida and Vancouver, Canada.
The size of the acquisition financing, the contract terms and the unit’s current EBITDA (earnings before interest, tax, depreciation and amortisation) are not disclosed. The division’s fiscal 2024 revenue was reported at roughly USD 835 million, which puts the enterprise value at 0.78 times sales. Listed packaged-food companies trade at two to three times revenue, and Keurig Dr Pepper’s acquisition of JDE Peet’s, completed in April 2026, was reportedly struck at around 14 times EBITDA. Set beside those, 0.78 times is not a discount; it is a price attached to a different kind of asset.
The perimeter is North America only. Maxwell House in Korea is made by Dongsuh Foods, a joint venture founded in 1968 with General Foods in which Mondelez still holds 50%; Kraft Heinz and Mondelez descend from the same General Foods lineage and parted in 2012. Rights to the name outside North America sit with JDE Peet’s, now owned by Keurig Dr Pepper. One name, three owners — a detail that returns later as both an exit route and a fuse.
The seller’s position explains the timing. Kraft Heinz announced in 2025 that it would split in two; chief executive Cahillane, who arrived in January 2026, halted the split and redirected the company toward a USD 700 million investment plan in marketing, R&D and people. Within that plan, coffee and processed meats were treated not as growth platforms but as businesses to be managed for cash or sold.
In February 2019, CNBC, citing sources, reported that the coffee business generated roughly USD 400 million of EBITDA and could fetch at least USD 3 billion; a sale was explored and shelved. Seven years later the asset goes for about a fifth of that figure. The desk reads the bargaining leverage as the buyer’s from the outset.
Structure Breakdown
The structure’s centre of gravity lies in four places, and the brand is not one of them.
A price equal to one trademark
Against the 2019 reported EBITDA, USD 650 million is 1.6 times. Current EBITDA is undisclosed; the desk’s estimate is roughly USD 100 million, which would mean earning power has fallen to a quarter in seven years. At USD 150 million the entry multiple would still be only about four times. Either way the conclusion is the same: this is not a price on the brand’s future but on what is left once the brand has melted.
The trademark ledger says what is left. Kraft Heinz impaired the Maxwell House trademark by USD 213 million in 2019 and then disclosed a carrying value of USD 823 million; further impairments followed in 2021 and 2023, and another USD 100 million in the second quarter of 2025. The desk estimates the current carrying value at roughly USD 550–650 million. On that estimate, the enterprise value is approximately the book value of a single trademark — and the two plants, the six other brands and the working capital come, in effect, for nothing.
A ledger heavier than the business
The trailing twelve months on which the price rests carry 2025’s beans. Arabica futures traded above USD 4 per pound in 2025, a record, and roasters buy and hedge six to twelve months ahead, so the peak cost arrives in the 2026 income statement with a lag. Kraft Heinz’s second-quarter 2025 filing recorded coffee price increases taken against input costs and falling coffee volumes in the same quarter.
Closing is set for mid-2027. As of October 2026, arabica trades around USD 3 per pound, about 29% below the peak of a year earlier, and Brazil’s 2026/27 crop is forecast at a record of more than 70 million bags. A drought or a re-imposed tariff could reverse that picture. As it stands, the buyer pays on a ledger weighted with expensive beans and takes the keys under a forecast of cheaper ones.
A second effect follows the first. As green coffee bought at high prices is consumed in a lower-price environment, the money tied up in inventory shrinks and working capital is released as cash. It never appears in the income statement, but it repays debt. The alpha the architect is after is not a brand revival; it is the lag between the price and the cash flow.
The 30%: buffer and bond
The press release presents the retained stake as conviction in the brands’ long-term potential. The arithmetic permits another reading. Seventy percent of the enterprise value is USD 455 million, yet Kraft Heinz says it will receive about USD 500 million in after-tax net cash. Either the new company borrows and passes cash up to the seller, or repeated impairments have left a tax basis high enough that little tax is due; the public record does not settle which.
In both cases the 30% works as a buffer between the cash the seller wants and the price the buyer can pay. It also reads as collateral. Roughly 250 functional staff is unlikely to be the whole head-office function of a business with more than USD 800 million in revenue; plant workers transfer separately, but much of logistics, procurement, IT and large-retailer sales probably remains inside Kraft Heinz’s shared services.
The day the transition services agreement (TSA, under which the seller keeps running back-office functions for the carved-out unit after closing) expires is the day the true standalone cost appears. Until then, the most realistic lever for keeping the seller’s service quality intact is the value of its own 30%. It is a stake that can be written as a restraint rather than a vote of confidence.
A buyer with a script
In 2021 One Rock, with Metropoulos, carved Nestlé’s North American water business out for USD 4.3 billion, stood it up as BlueTriton, and later reverse-merged it with the listed Primo Water to form Primo Brands, with revenue of about USD 6.5 billion. The script is not brand revival. It is the binding of plant and cash flow that a conglomerate has discarded into a platform, and the mounting of that platform onto something larger.
The desk considers the coffee portfolio likely to be the first chapter of the same script: fill the two plants with retailer private-label and contract-manufacturing volume, then add other orphaned coffee brands to the same plants and sales force so that fixed costs are shared. How loved the brand is does not enter this structure as a core variable. The desk’s investment-committee conclusion follows from that: as a brand turnaround, rejected; as a plant-and-cash-flow platform, conditional interest.
The Governing Logic
An architect builds structure in order not to lose, and only then to earn. What follows is therefore not a description of what One Rock did — that is unknown — but of what the seat requires. The emphasis falls on the downside hedge rather than the exit narrative because the three variables that move this asset’s earnings — the green coffee price, the pace of category decline and the standalone cost base — all sit outside the sponsor’s control.
By the desk’s estimate, a move of USD 0.50 per pound in green coffee shifts annual operating profit by USD 20–30 million, about a quarter of estimated EBITDA. Standalone costs revealed after the TSA could, in the worst case, erase 40% of EBITDA; that too is an estimate. Mainstream ground-coffee volume shrinks by about 3% a year, and the only brake is a price cut that reduces revenue. The sponsor controls three things only: the price paid, the amount of debt, and the wording of the contract.
Downside Protection: Five Layers
The plant, bought at a price that assumes the brand does nothing. This is the one layer already public. The desk estimates the replacement cost of the plants at USD 200–300 million and the royalty value of licensing the brands to another roaster at USD 150–250 million, for a liquidation floor of roughly USD 400–500 million — a 25–35% discount to enterprise value.
Running retailer private label and contract volume through the plants is defence, not growth. Even if brand volume falls 6% a year in the downside case, the fixed costs stay absorbed and the covenants (the financial tests lenders write into loan agreements) hold. The tipping point is not a sales breakout but the moment private-label volume passes 20–25% of output and begins to offset the brand’s decline; until then EBITDA is on a falling path.
Debt held beneath the liquidation floor. The floor protects equity only if the debt sits below it. On estimated EBITDA of USD 105 million, borrowing capped at three times — about USD 300 million — means that even in the worst case plant and licence value repay the lenders and leave USD 100–200 million for equity. Mezzanine (subordinated debt layered between senior loans and equity) has no place here: stack two layers of fixed interest on cash flows that move with the bean, and the liquidation value is absorbed entirely by creditors. If the actual financing proves to be above four times, the margin of safety existed only on paper.
Settlement in bags, not dollars. Fix the working-capital peg — the agreed level of inventory and receivables delivered at closing — in bags of green coffee rather than in money. However much high-cost inventory remains at closing, the price difference is settled with the seller and the buyer takes quantity only, which nails the cost lag to the buyer’s side of the table. Index-linked green-coffee clauses in retailer supply contracts then perform the pass-through that the brand no longer can.
The 30% as a performance bond. The TSA should run 36 months at a fixed price with an extension option, and a service-level failure should be punished not with damages but with dilution of the seller’s residual stake or a discount to a call-option strike. A drag-along clause (the right to compel a minority holder to sell alongside the majority) keeps the seller from blocking an exit. A non-controlling holder can still neglect the business; the design makes that neglect return as an impairment of its own holding.
Tail risks go back the same way. Environmental liability at the Jacksonville plant, which has operated since 1924, carve-out tax exposure and pre-closing recalls are matters for special indemnities and an escrow of 5–8% of enterprise value held for two to three years, not for warranty insurance. Insurers do not underwrite those tails; the seller should hold them.
A reading of the rights map. The North American trademarks transfer in full, but Gevalia and Tassimo are divided along the same line as Maxwell House, Tassimo brewers are made by Bosch, and global leadership of that system now sits with the dominant player in North American single-serve. The new company’s machine roadmap depends on a competitor’s decisions, and the exact contractual form of that dependence is a diligence item. A brand held by three owners cannot be grown globally and sold; that exit was closed from the start.
Read in reverse, the same map is an opportunity. Keurig Dr Pepper, holding the rights outside North America, is the most natural strategic buyer to reunify the brands. It has an incentive of equal size to squeeze rather than buy, and the architect has to hold both faces in view.
The value-creation bridge summarises the design. In the desk’s base case roughly a third of the return comes from debt repayment, nearly half from green-coffee normalisation and plant utilisation, and only about a fifth from multiple expansion. Exit is an option — a reverse merger into a listed consolidator, or reunification under Keurig Dr Pepper — and the structure must still return capital through debt paydown and plant value if neither opens. It is a design for not losing, with a single upside bet placed on top: the lag in the cost cycle.
What Must Be True
Three propositions carry the read. Each comes with the observation that would break it.
1. The weight on the ledger is cyclical, not structural. The trailing earnings must be depressed by 2025’s beans rather than by a permanent loss of earning power. The evidence against is already on the table: reported EBITDA of about USD 400 million in 2019 against an estimated USD 100 million now, a pace that would halve it again inside a five-year hold, and a seller that cut its long-term cash-flow forecasts for the trademark four times, in 2019, 2021, 2023 and 2025.
Futures gains from the 2025 spike may also sit inside the earnings base, and high-cost inventory may depress margins in the first quarters after closing. Falsifier: margins that fail to widen once that inventory has cleared, despite cheaper green coffee — or arabica back near USD 4 before closing.
2. The standalone cost fits inside the cushion. The EBITDA on view stands on the parent’s allocated costs. The desk estimates the gap between allocated and standalone cost at USD 15–40 million a year, which removes 15–40% of estimated EBITDA on its own. Falsifier: a standalone cost base at or above the top of that range when the TSA ends, or a decline in the seller’s service that carries no consequence for its stake.
3. The margin of safety exists in numbers, not on paper. Debt must sit under the liquidation floor, and the plants must run whether or not the brand sells. Falsifier: acquisition financing reported above four times estimated EBITDA, or plant utilisation that falls in step with brand volume because private-label and contract work does not arrive.
The stress case is a single year. Suppose that in 2028 the TSA expires and standalone costs surface in full, the collective agreement at Jacksonville — where a March 2024 labour-board vote kept union representation by 107 of 131 ballots — comes up for renewal, a La Niña drought sends green coffee back to USD 4 a pound, and Walmart, which accounts for about 21% of Kraft Heinz’s company-wide sales, strengthens its own-label coffee. EBITDA falls from an estimated USD 105 million to around USD 50 million, USD 300 million of debt becomes six times leverage, and the covenants are breached. The desk puts the probability at 10–15%; if it happens, the equity is on a path to total loss.
The sequence matters more than the trigger. Shelf space shrinks, volume falls, utilisation drops, unit fixed cost rises, price goes up, and the shelf shrinks again; once inside that loop, EBITDA halves in two to three years. A covenant breach leads to a forced sale at plant value, which impairs the seller’s 30%, and a seller disclosing a failed carve-out ensures that the next buyer appears only at a lower price.
Valuation does not collapse at once: the utilisation floor gives way first, and the multiple follows. Whether the five layers are in the actual contract turns on three diligence items — standalone cost, the contractual structure of the brand rights, and concentration among the top retailers.
The Architect’s Takeaway
If you were building your own deal room, the thing to carry out of this transaction is not USD 650 million but the order in which the number was read. That order can be copied at any size of capital.
Separate the sign from the plant. Every declining business has a sign people look at and a plant that makes the money. A newsletter losing subscribers has a name for a sign and a mailing list and sending infrastructure for a plant; a neighbourhood tutoring school has its principal’s name, and beneath it a lease, an instructor network and parent data. The first question is always what remains when the sign comes off, and whether it can run under someone else’s sign.
Price the lag. A seller always arrives with the heaviest ledger. Before negotiating, take apart why the last twelve months were bad: separating structural decline from a temporary cost shock is the whole of diligence. If the shock is temporary, enter at the price it produced and push the effective date past the point where it unwinds — this deal was signed in November 2026 and closes in mid-2027, and the desk does not treat those eight months as an accident. The lag is a bet, not a guarantee, so fix the settlement of inventory and receivables in quantities: subscriber counts, stock units, contract numbers.
Turn the counterparty’s residual stake into collateral. When a seller offers to keep 30%, the novice reads conviction and relaxes. Write it instead so that a poor handover dilutes the stake or lowers a call-option strike; a buyer without capital has then purchased the seller’s diligence with a clause. When you are the one keeping a stake, check that it carries a co-sale right, or your 30% is paper that can never be sold.
Keep fixed promises beneath liquidation value. In a personal deal room, debt is not only bank debt: fixed salaries, prepaid subscriptions, long leases and minimum volumes promised to contractors all count. Work out what the list, the infrastructure and the contracts would fetch if handed to someone else, and keep total fixed commitments below that floor. If revenue hangs on an algorithm or a platform policy you do not control, one turn of fixed-cost leverage is already too much.
Read the partition line. Your subscribers are yours but the recommendation algorithm is the platform’s; your brand is yours but the payment rails belong to the processor. Ask where your rights end, who holds the far side, and whether that party is a buyer or a squeezer. By the desk’s reading, a line of this kind is why Dongsuh Foods has not exported Maxim in more than half a century: once drawn in a contract, it lasts.
Platform the orphans. Brands that a large company has labelled non-core tend to come out in sequence, and whoever owns the first plant is the likeliest buyer of the second brand. Abandoned community servers, the user base of a game with no sequel, the asset library of a closed studio and a vertical newsletter dropped by a large publisher are all orphans. Design the first acquisition so that the second can be mounted on the same plant; one plant under many signs is built by design rather than capital, and that is the definition of a micro-GP.
One principle binds the six. Build not to lose, and place a single upside bet on top. Place two, and it is no longer design; it is hope.
Syndicate’s Note
This is not a coffee deal. It is a design that converts the time of a decline into cash flow: five public facts, five layers of defence, one bet. Brand revival appears nowhere among its premises, and a reader who needs it to appear has misread the price.
The desk does not find this mechanism among the registered patterns. It is the mirror of Proof-Deferred Pricing: there a seller prices a peak and the buyer hands the proof risk back through contingent consideration, while here the price is struck on a trough ledger and the buyer keeps the cycle risk as its one bet. The retained 30% borrows the hostage logic of Mutual Hostage Architecture, but the hostage is equity rather than captive revenue, and it is one layer of five rather than the governing logic.
The working name is The Heavy Ledger: a carve-out priced on the seller’s heaviest books and closed after the weight has lifted. One disclosed case resting on a reconstructed term sheet does not make a pattern, so it enters the declassification queue of the Structure Pattern Library rather than the index.
House Verdict: TRACK. Under the Verdict Protocol, the financing disclosure, the closing in mid-2027 and the first standalone results will each test this read. Falsifier: if the acquisition financing is reported above roughly four times estimated EBITDA, the liquidation floor belongs to the lenders and the claim that this structure was built not to lose is wrong. The verdict classifies structural learning value; it is not an investment judgment.
FAQ
What is a carve-out in private equity?
A carve-out is the sale of a division or business unit out of a larger corporate parent, usually to a financial sponsor, which then stands it up as an independent company. The reported earnings of such a unit rest on costs allocated by the parent, so the true standalone cost base only appears once transition services end. That gap, rather than the headline multiple, is where carve-out underwriting is tested.
Who is buying Maxwell House, and what does the deal include?
One Rock Capital Partners is acquiring Kraft Heinz’s North American coffee business at an enterprise value of USD 650 million, with closing expected in mid-2027. The perimeter includes Maxwell House, Gevalia and Tassimo in North America, plants in Jacksonville, Florida and Vancouver, Canada, and about 250 functional employees. Kraft Heinz keeps a 30% non-controlling stake.
Why did Kraft Heinz keep a 30% stake in its coffee business?
Kraft Heinz describes the stake as confidence in the brands’ long-term potential. The desk’s structural reading is that it also bridges the gap between the cash the seller wanted and the price the buyer could pay, and that it can serve as collateral for the seller’s performance under transition services. The actual shareholder terms are not disclosed.
Does the sale affect Maxwell House or Maxim in Korea?
No. The transaction covers North America only. Maxwell House and Maxim in Korea are produced by Dongsuh Foods, in which Mondelez holds 50%, and rights elsewhere sit with JDE Peet’s, now part of Keurig Dr Pepper.
This article is for informational purposes only and does not constitute investment advice, an offer, or a solicitation. Figures are based on public reports and may be estimates.
For more structural insights and deep-dive video breakdowns, visit Structure Syndicate on YouTube.