Cocoa is falling, and Mondelez should be the winner. Whether it keeps the winnings depends less on the cocoa market than on whether the world's chocolate makers keep the peace.
At $58.19, its close on October 2, Mondelez is already priced for a full recovery. Put the business back at its 2024 margin, grow it 4% a year for a decade, and the shares offer about 8.5% a year. The strongest bull case we could build with discipline offers 9.1% to 9.4%. Only one version clears 10%, and it quietly assumes that every dollar Mondelez reinvests will earn 43% forever. Meanwhile the company says it is spending its cocoa relief on its brands, prices in Europe are already falling, and Lindt has just told investors it will use easing cocoa to reprice for volume. Two numbers decide this, and neither appears in any filing: how much of the relief survives the promotion calendar, and what the new emerging-market volume earns.
The bull case fits on an index card, which is part of its charm. Cocoa spiked, squeezed every chocolate maker's margin, and is now coming down. Mondelez, owner of Cadbury, Milka, Toblerone and Oreo, should get its margin back, while its brands and its reach in emerging markets keep the top line growing. The repair pays for the patience, and the growth pays for everything after.
Its latest advocate is Ben Shuleva, Fidelity's consumer-sector leader, in a Barron's interview with Evie Liu published October 1. Mondelez is among the top holdings of the staples fund he co-manages, and his case is the index card: chocolate and cookies are indulgences where brands matter and private labels struggle, the emerging-market business grows faster than America does, and cocoa, which squeezed margins on the way up, is on the way down.
It is a good hypothesis because it can be wrong in specific, checkable ways. This note goes looking for them.
Shuleva summed up the whole industry in a line: "Staples companies buy commodities and sell brands to capture the spread in between." Adrian Slywotzky and David Morrison catalogued the ways a business can capture a spread like that in The Profit Zone (1997). Several of their profit models describe a piece of Mondelez. One describes the whole of it.
Brand profit. The same chocolate earns more in a purple wrapper. Slywotzky's favorite demonstration was a single California plant that built one car under two nameplates; the Toyota badge fetched about $300 more than the GM one (The Art of Profitability, 2002, p. 132). Cadbury is that badge, and so are Oreo and Milka.
A model that fits is only half the question. The other half is whether the company is winning at it, and for a brand the test is old and unforgiving: can it raise its price and keep its volume?
Here Δp / p is the change in price and Δq / q the change in volume and mix. Mondelez's second quarter, region by region:
| Region | Price | Volume and mix | The brand test |
|---|---|---|---|
| North America | +2.2 | +1.2 | Winning |
| AMEA | +1.9 | +5.2 | Winning, though operating income fell 0.4% |
| Latin America | +7.9 | +0.5 | Winning, flattered by about 1.5 points of trade inventory |
| Europe | −1.4 | −2.1 | Losing: price cut, volume still down |
| Mondelez | +1.5 | +0.7 | Winning, narrowly |
Percentage points of organic revenue growth, second quarter of 2026, from the company's release.
The brand is winning nearly everywhere except the one region where the chocolate truce is decided.
The profit pattern the company has been living through comes from Slywotzky's sequel, Profit Patterns (1999, pp. 107 to 109): the value chain squeeze, in which suppliers on one side and customers on the other both grow stronger and take more of the value, leaving the step in between unable to cover its cost. For two years the step in between was the chocolate maker, caught between the cocoa market and Europe's grocers. The squeeze is now releasing on the supplier side. This note is really about one question: who collects the release, Mondelez, the retailer, or the shopper at the shelf?
So far, nobody: the release has not reached the income statement. Higher raw material costs were still among the reasons the company gave for lower operating income in the second quarter, because a large maker buys its cocoa well ahead. The scoring starts when it arrives, and the test is the same one in reverse: whoever's share of the chain's profit rises as cocoa falls is collecting it.
Underneath the pattern sits one of the oldest models in game theory, the repeated game. Chocolate makers do not agree on prices, and they don't need to, so long as each expects the others to answer a price cut with one of their own. The truce holds while
Here G is what one maker gains, once, by cutting price first while cocoa falls; Vtruce is each year's profit if everyone holds price; Vwar is each year's profit after a price war; and δ is how heavily each maker weighs next year against this one.
Falling cocoa raises G: there is suddenly room to cut price and still earn more per bar than last year. Weak volume raises it again, because the shelf space a price cut wins is worth more when shoppers are scarce. And on September 29, Lindt, the category's premium leader, said it expects 2027 volume growth from "increased brand investments, innovations, and an adjusted pricing strategy, supported by cost savings and easing cocoa prices." In the language of the model, someone has just mentioned G out loud.
The second piece of math decides what Mondelez's growth is worth. A business earning NOPAT and growing at g is worth
where RONIC is the return on the new capital that growth consumes and r is the return the owner requires. Growth adds value only when RONIC is above r; at RONIC equal to r, growth is worth exactly nothing. For Mondelez the new capital is outlets, production lines, acquisitions, and the advertising that the accounts treat as an expense.
We rebuilt the ten-year owner-cash model behind this note line by line and turned it around, in the manner of expectations investing (Mauboussin, 2006). Instead of asking what the shares are worth, we ask what $58.19 offers a year on each set of assumptions. Vista does not publish what a share is worth. What a price assumes is the more honest question anyway.
| The case | What it assumes | What $58.19 offers a year |
|---|---|---|
| Downside | 1.5% growth; margins stuck at 13% | 5.0% |
| Base, with growth earning 20% on new capital | The base case below, with its terminal value built from an explicit 20% return on new capital | 8.3% |
| Base | 4% growth for a decade; margins back to 16.2% by Year 4 and held there forever | 8.5% |
| Base, working capital tied to growth | Two treatments of supplier finance | 8.7% to 8.9% |
| The bull case at full strength | 5% growth, 17% margins, 20% return on new capital, about $375 million of interest | 9.1% to 9.4% |
| Upside, disciplined | 5% growth, 18% margins, 20% return on new capital | 9.9% |
| Upside as first built | The same, with a terminal value implying 43% on new capital forever | 10.3% |
Annual return at which each case's owner cash equals the October 2 close. Year-end cash timing; the bull case range spans year-end to mid-year timing. Inputs and method are at the end of this note.
Four things stand out.
The base case is already generous. It returns margins to 16.2% by Year 4 and keeps them there forever. 16.2% was 2024, the better of the two pre-spike years we checked: 2023 was 15.9%, 2025 fell to 13.2%, and the second quarter of 2026 printed 13.1%. It also grows revenue 4% a year for a decade, inside the company's long-term target of 3% to 5% but above the 2.2% it managed in the latest quarter.
To offer 10%, the base needs a margin nobody has seen. With everything else unchanged, margins would have to average 20.05% from Year 2 to Year 10, nearly four points above any full year we checked.
The terminal value is doing quiet heavy lifting. Nearly 55% of the base case lives beyond Year 10, and the shortcut that produces it implies that new capital earns about 27%. Pin that return at 20% and the base offers 8.3%; at 12%, 7.9%. The upside case is the only one above 10%, and its terminal value implies 43%.
Year 1 asks for a fifth more cash than the company is promising. The base case's first twelve months produce about 21% more owner cash than the company's own 2026 free cash flow guidance of about $3 billion implies, once stock compensation, lease payments and an acquisition allowance are taken out.
This reminds me of Warren Buffett's test of a business, given to the Financial Crisis Inquiry Commission in 2010: "if you have to have a prayer session before raising the price by 10 percent, then you've got a terrible business." This autumn's chocolate question is the mirror image: who can keep a price when the cost underneath it falls. Judging by Europe's second quarter, the prayer meetings have started.
In the second quarter, adjusted gross profit rose 3.0% at constant currency and adjusted operating income fell 6.1%. The gap is the reinvestment. The company names the reasons for the decline itself: higher raw material costs, higher other selling, general and administrative expenses, and higher advertising and consumer promotion costs. Its chief executive describes "reinvesting behind our brands to enable sustained performance for years to come." That may prove wise. It is not, yet, cocoa savings kept.
The regions tell four different stories:
One more line belongs in any model. Mondelez runs on negative working capital, but part of it is financed: $2.9 billion of supplier finance sits inside its payables and $585 million of receivables have been sold, and both balances were lower in June than in December. That is a stress to disclose, not a gift to capitalize.
The literature is not kind to permanent margins. Fama and French (2000) found that profitability reverts toward the average at roughly 38% a year, and Chan, Karceski and Lakonishok (2003) found sustained high growth rare and impossible to pick in advance. Companies that grow their investment or their assets fastest have tended to earn less afterward (Titman, Wei and Xie, 2004; Cooper, Gulen and Schill, 2008), which is a reason to pin the return on new capital rather than let a formula assume it. And the restructuring and system costs that companies leave out of adjusted earnings tend to come back (Doyle, Lundholm and Soliman, 2003), which is why the model keeps charging for them.
This reminds me of Proverbs 27:23-24 (BSB): "Be sure to know the state of your flocks, and pay close attention to your herds; for riches are not forever, nor does a crown endure to every generation." A 16.2% margin is a crown. The base case wears it for eternity.
Here is the best case we could build without cheating. Margins climb to 17.0% by Year 5, eighty basis points above 2024. Revenue grows 5% a year from Year 2, the top of the company's long-term range. Interest runs at about $375 million, the company's 2026 planning figure. Working capital grows with sales, without assuming the supplier-finance programs grow too. And new capital earns 20% in perpetuity, which is generous but not magical. On that case, $58.19 offers 9.1% to 9.4% a year, depending only on when in each year the cash is assumed to arrive.
That is not a poor return from a company that sells biscuits and chocolate on every continent. Claus and Thomas (2001) estimated that the equity premium implied by prices and analysts' forecasts is closer to three percent than to the historical average. An investor who agrees can defend a 9% hurdle on a steady consumer staple, and at that hurdle the bull case works. An investor who requires 10% will find that no disciplined case we built gets there. Which hurdle is right is the reader's decision, not ours.
Likely, we can get some of it from primary research, which is what Vista is all about. The filings will tell us, a quarter at a time, whether the truce held. People who ran these businesses through earlier cocoa cycles can tell us now what usually happens, and how fast.
We would askWhen cocoa fell after a spike, how much of the gross margin recovery did the business keep after six quarters of promotion, price-pack resets and retailer funding?
The answer that would change the view"Less than half."
We would askWhen volume grows 5% and profit doesn't, is the gap front-loaded distribution that pays back within three years, or the permanent cost of competing with local brands?
The answer that would change the view"The new volume settles at a structurally lower margin."
We would askAs a matter of past practice, how far forward do the large makers buy, and how long before a fall in futures reaches the cost of a delivered bar?
The answer that would change the view"Most of 2027's cost is already locked near the peak."
We would askWhen a supplier's costs fall, do retailers take it as price, as promotional funding, or as their own margin, and how quickly? Are this year's share gains bought with promotion?
The answer that would change the view"The retailer takes it, within two quarters."
Every conversation would draw on past, general experience only: no current employees of Mondelez or of its customers and suppliers, nothing confidential, and every Advisor screened for conflicts before a word is said. For a client engagement, Vista would combine this note with those interviews: two or three Advisors, structured conversations, and one written brief that says where they agree, where they split, and what would change the answer.
We will score this note in February 2027, after Mondelez reports its fourth quarter, and publish the result on the scorecard whether it flatters us or not. The tests are fixed today, so they cannot drift:
Four or five passes, and the relief is being kept: the base case's margin path starts to look earnable. Two or fewer, and the truce is not holding. Three is too early to say. We will also report what the price then offers on each case, and read Lindt's January sales report for what "an adjusted pricing strategy" turned out to mean.
Willy Wonka, in the 1971 film, described the investor's position better than any strategist: "The suspense is terrible. I hope it'll last." It will last at least two more quarters. Until then, the price already assumes the happy ending.
Owner cash each year = (revenue × adjusted EBIT margin − net interest) × (1 − tax rate) + depreciation and amortization − capital spending − working-capital investment − after-tax ERP and restructuring cash − finance-lease principal − an acquisition allowance + equity distributions − minority distributions − legacy pension cash. Year 1 runs from October 3, 2026 to October 2, 2027. Year-10 cash grows into perpetuity at the terminal rate, or, where stated, the terminal value is built from an explicit return on new capital: NOPAT × (1 + g) less the investment that growth at that return requires. Shares are fixed at 1.290 billion; no buybacks are assumed. The return each case offers is the discount rate at which its owner cash, per share, equals the $58.19 close. Every figure in this note was recomputed in code from these inputs.
| Input | Downside | Base | Upside | Bull case |
|---|---|---|---|---|
| Year-1 revenue ($bn) | 39.5 | 40.5 | 41.0 | 40.5 |
| Annual growth after Year 1 | 1.5% | 4% | 5% | 5% |
| Adjusted EBIT margin, Years 1 to 4 | 12.5, 13, 13, 13% | 14.5, 15.5, 16, 16.2% | 15.5, 17, 18, 18% | 14.5, 15.5, 16.2, 16.7% |
| Margin from Year 5 | 13% | 16.2% | 18% | 17% |
| Tax rate | 27% | 25% | 24% | 25% |
| Net interest ($bn) | 0.60 | 0.45 | 0.40 | 0.375 |
| Capex, D&A (% of revenue) | 4.0, 3.4 | 3.8, 3.4 | 3.7, 3.4 | 3.8, 3.4 |
| Working capital | 0.7% of sales | 0.6% of sales | 0.5% of sales | 6.4% of new sales |
| ERP and restructuring cash ($bn) | 0.35 falling to 0.15 | 0.30 falling to 0.10 | 0.25 falling to 0.075 | 0.30 falling to 0.10 |
| Lease principal, acquisitions ($bn) | 0.17, 0.20 | 0.16, 0.20 | 0.16, 0.20 | 0.16, 0.20 |
| Terminal growth | 2% | 3% | 3.5% | 3% |
Every case also adds $0.05 billion a year of distributions from equity investments and deducts $0.016 billion of minority distributions and $0.05 billion of legacy pension cash. The working-capital variants of the base case apply either the June ratio of trade working capital to sales (negative, with the financing programs in place) or the same ratio with the programs added back (6.4%), to each year's new sales.
This note is research, not investment advice. It states what a market price assumes under labeled assumptions; it is not a recommendation to buy, sell or hold any security, and the decision belongs to the reader. As of October 3, 2026, Russ Rosenzweig, Vista's founder, owns no Mondelez shares. The note began as a research question Russ Rosenzweig put to the Vista Research Desk, which challenged the model over three rounds; the model and every figure were then rebuilt and checked against the company's filings.