Business
Know the Business — The Magnum Ice Cream Company
Bottom line. A high-quality category (consumable indulgence, 3–4% structural growth, brand-led pricing power, a physical cold-chain moat) wrapped around a freshly-listed, just-levered carve-out whose reported FY2025 P&L is materially understated by separation drag. The right question is not "what is the multiple today" — it is "what does the business earn in 2028 when the €500M productivity programme has fully landed, Transitional Service Agreements with Unilever have ended, and AMEA's 23% adjusted EBITDA margin is a larger share of mix." Consensus appears to underweight the AMEA mix shift and the mechanical FCF snap-back as separation costs roll off, and to overweight how quickly a 16,500-person ex-Unilever subsidiary becomes a clean-execution compounder.
1. How This Business Actually Works
The revenue engine is branded ice cream IP × premium price-per-serving × cold-chain placement. TMICC sells a Magnum stick or Cornetto cone for €1.50–3.00 in markets where the dairy + cocoa + sugar inputs cost a fraction of that, and it controls the freezer cabinet in three million convenience stores, kiosks and grocery aisles where the impulse purchase actually happens. The freezer is the moat: it is paid for, placed and serviced by TMICC, and a new entrant cannot rent that shelf at any price. Four of the world's five largest ice cream brands (Magnum, Ben & Jerry's, Cornetto, Wall's/Heartbrand) sit inside that fleet, alongside premium-pint formats (Talenti, Yasso) and accessible coinage formats (Popsicle, water ices) that drive emerging-market penetration.
Incremental profit is governed by three levers: (a) mix toward premium single-serve and "better-for-you" formats (Magnum Bonbon, Yasso pints, Cornetto Max) which carry 2–3× the gross margin of family tubs; (b) cabinet productivity — a Heartbrand-branded freezer in a Pakistani kiosk has near-zero marginal cost once placed and dictates SKU mix shown to the consumer; (c) AMEA penetration — TMICC earns a 22.9% adjusted EBITDA margin in AMEA on €1.9B of revenue versus 13.1% in Europe & ANZ. The economic ceiling is set by the freezer footprint, which is itself a capex decision. Capex ran 4.5% of revenue in FY2025 — most of it cabinets and supply-chain transformation, not factories — and the FY2025 €1.3B adjusted EBITDA against €357M of capex says the underlying cash machine is already there, even if the reported FCF was crushed to €38M by separation outflows.
The single most important sentence to remember. Reported FY2025 operating profit of €599M and free cash flow of €38M are not the run-rate. Adjusted EBITDA was €1,255M (15.9% margin) and management guides to €0.8–1.0B of FCF from 2028 once Transitional Service Agreements with Unilever expire end-2027. The gap is carve-out plumbing, not lost economics.
2. The Playing Field
There is no listed pure-play comparable. The "main global ice cream competitor" identified by management is Nestlé's 50/50 Froneri JV with PAI Partners — which is private and consolidated only as an associate inside Nestlé. So the peer set is necessarily mixed: the former parent (Unilever), the strategic rival inside a conglomerate (Nestlé), the US premium pint operator (General Mills owns Häagen-Dazs US), the global snacks adjacency (Mondelēz), and the US confectionery indulgence comparator and licensing partner (Hershey).
Read the table this way. On EBITDA margin, TMICC sits at the bottom of the peer set (15.9% on adjusted basis, ~11.8% reported) — that gap is the value-creation prize; management says historical profitability has been "significantly behind the estimated profitability of our main global ice cream competitor." On EV/EBITDA, the company trades cheap (~10x Adj EBITDA) versus Hershey at 21× and Nestlé at 15× — the obvious explanation is that the market is discounting both the carve-out drag and the pure-play execution risk. On EV/Sales (1.4× versus a 2.2–3.5× peer band), the gap looks larger than the EBITDA gap because the carve-out left TMICC with a fresh €3B of bond debt and a depressed FY2025 EBITDA denominator. The cleanest read of what "good" looks like in this industry is Hershey and Unilever: 17–20% EBITDA margins, 10–12% ROIC, and a 12–20× EBITDA multiple. TMICC's stated medium-term algorithm (3–5% OSG + 40–60bps annual margin expansion) is precisely the path from where it sits today to where those peers already are.
3. Is This Business Cyclical?
The category is secular, but the earnings line is moderately cyclical — mainly through commodities, weather and emerging-market FX. Demand is structurally defended ("the lipstick of foods" — CEO; even families under stress make budget room for ice cream), and global category growth has run 3–4% for a decade through recessions, inflation, COVID and lapping. What moves the P&L instead is the input cost cycle (cocoa +100% from 2023 to 2025 cost 380bps of gross margin in FY25), the weather cycle (wet UK summer 2024, severe Philippines weather 2025), and the EM FX cycle — a 4.3% FX translation drag on FY25 reported revenue is mostly Turkish lira and US dollar weakness, with a €31M non-cash hyperinflation loss on top.
Two patterns to internalise. First, the category cycle is benign on volume (1.1 → 1.5 → 2.9 percentage points across FY24 / FY25 / Q1'26) — volume is accelerating into 2026. Second, the earnings cycle is being shaped right now by cocoa: FY25's 380bps of input-cost inflation was a margin event, and management has flagged "a little bit of tailwind" on cocoa, dairy and palm oil into the 2026 second half. If cocoa retreats meaningfully, FY26 adjusted EBITDA margin expands ahead of the 40–60bps guide. If it does not, the productivity programme still delivers, but the reported line stays compressed for another year. Either way, the demand side is not the variable to watch — the input cycle is.
4. The Metrics That Actually Matter
Standard packaged-food metrics (P/E, EBITDA margin) are necessary but not sufficient here. The five numbers below explain value creation and destruction in TMICC better than the income statement does.
The trap to avoid: anchoring on reported FY2025 numbers. Reported operating margin (7.6%) and reported FCF (€38M) are both polluted by one-time separation cash. Adjusted EBITDA (€1,255M) and the per-region margins (13.1% / 14.1% / 22.9%) are the right starting point for any thesis. Track OVG and AMEA margin closely — they prove or disprove the "premiumising-emerging-markets pure-play" thesis. Track the TSA exit progress because that is the mechanical FCF unlock. Everything else is downstream.
5. What Is This Business Worth?
The right primary lens is EV/Adjusted EBITDA at a steady-state margin (not the depressed FY25 reported margin), cross-checked against run-rate free cash flow at the 2028 inflection point management has guided to. A standard P/E is misleading for the next two years because reported EPS is being suppressed by separation cash interest, hyperinflation accounting and TSA markups that all unwind. A pure DCF is honest but largely a function of the three assumptions in the value-driver table below.
A formal sum-of-the-parts is genuinely useful as an analytical lens because the three reporting regions have materially different economics — but it should be treated as a discipline check, not a price target. The cleanest read is in the regional table below.
Where consolidated reporting hides value. AMEA is 24% of revenue but ~34% of adjusted EBITDA, with 10.9% organic sales growth and a margin trajectory that looks like an emerging-market consumer compounder, not a mature staple. Group-level multiples mark this segment to mature-Europe economics. That is the most plausible source of analytical upside — and the single hardest fact to defend if AMEA volume growth slows.
What would make this cheap: cocoa retreats meaningfully into 2026–27, AMEA volume holds 4%+, the €500M productivity programme delivers on schedule, and the market starts crediting 2028 FCF guidance (€0.8–1.0B → ~9% FCF yield on today's market cap). What would make this expensive: the productivity programme is slower than guided, Italy and Brazil turnarounds drag past 2026, AMEA volume disappoints, and the market re-prices to reported (not adjusted) margins for longer than expected.
6. What I'd Tell a Young Analyst
Don't anchor on FY2025 reported numbers. They are clean accounting but a dirty signal — separation cash, TSA markups, hyperinflation and €3B of fresh bond debt all hit at once. The underwriteable number is adjusted EBITDA at €1,255M and the trajectory toward 16.5–18% adjusted EBITDA margins by 2027–28.
Watch AMEA, then cocoa, then the TSA exit calendar — in that order. AMEA's 22.9% margin and 10.9% OSG is the engine of the mix story. Cocoa is the single largest swing factor on reported EBITDA in FY26. TSA exits are the mechanical FCF unlock — management has been specific that they end by year-end 2027, and every quarter of progress is visible in the cash bridge.
The moat is the cabinet, not the brand. The brands are valuable because the cabinets exist; without the freezer placement, Magnum is just another premium chocolate-coated dessert. Cabinet net adds and AFH revenue per cabinet are the structural variables — they get less analyst attention than brand share but matter more for the next decade of revenue ceiling.
The market's main misread is probably the FCF math. A €38M FCF print on a 2.4× levered balance sheet looks alarming. Management's €0.8–1.0B run-rate FCF guide for 2028 is internally consistent with the adjusted EBITDA bridge and the separation cash schedule — but it requires the analyst to trust both the productivity programme and the TSA wind-down. If you believe the company, MICC screens as a high-quality #1-share global staple priced as a stressed carve-out. If you don't, the discount is fair.
Where you'd actually be wrong. Two things genuinely change the thesis. First, share loss returning in any of the three major brand pools (US, UK, China) — TMICC lost share for a decade before this turnaround, and a sustained reversal would say the brand IP is less durable than the bull case requires. Second, the productivity programme stalling at €250M cumulative — that would mean the standalone cost base is structurally heavier than guided, and the margin bridge collapses. Watch both quarterly.