Market prices an AI-loser; this DCF prices a muddle-through with regression β · +61.9% margin of safety
Market sits below MC p5 floorIpsos trades at €36.70 against a base-case intrinsic of €96.23/sh after a 10% governance haircut. Even the Monte Carlo p5 of €66 sits well above today's tape — the market is pricing the AI-loser branch in full, not the probable muddle-through.
What it sells, where it sells
Operating segments
Ipsos is a single-segment business by IFRS — the operating debate is really between the small, fast-growing Ipsos.Digital/GMS growth driver (the platform-shift hedge) and the 95% legacy survey base whose pricing power AI is actively repricing.
Country mix (revenue-weighted CRP input)
US + EU-4 (FR/UK/DE/IT) = ~59% of revenue — a US recession or sustained EUR strength against USD (Q1 2026 already took 5.4 pts off reported revenue) is the macro path that hurts most.
Background — five things to know before the case
Context the share price doesn't carry on its face. Skim this once and the bull / bear bullets below stop reading as inside baseball.
- Founder is a buyer, not a seller. Didier Truchot's DT & Partners vehicle bought aggressively at €50-€65 through 2023-2024 — well above today's €36.70. Truchot stepped down as Chair Feb 28 2026 on health grounds (cancer); named Chairman Emeritus, not entrenched.
- The 2022-2025 plan missed. Oddo BHF flagged the credibility problem when the new Horizons plan launched Jan 2026 targeting 3-4% organic 2026-28 and 5%+ by 2028. Q1 2026 organic landed at -1.4%, exactly the wrong directional start.
- Russia deconsolidated Jan 1 2026. Russia was a high-margin contributor — its removal is an explicit drag on 2026 margin reset and is already baked into the FY2025 base.
- M&A is the operating model — 100+ deals. 15 acquisitions since 2023, €179M deployed in FY2025 (BVA Family + infas, largest since 2018). Goodwill ~€1.4B sits on €2.5B revenue; ROIC has held at 12-14% across the pace.
- Net dilution is negative. Share count shrunk 44.25M (Dec 2022) → 43.20M (Dec 2025). New €100M buyback (~6.7% of cap) launched March 2 2026 for cancellation, on top of recurring dilution-offset buybacks.
Where we diverge from the sector
Each row asks: if we'd used the sector median for this one assumption instead of our override, how much would the share price change? Negative = our override is more conservative (worth less); positive = more aggressive (worth more). Skim down the rightmost column to see which bets are doing the work.
The story & the five claims
The 10-year story this DCF is built on, plus the five anchor claims that translate the story into model inputs.
The 10-year story
Ipsos is a muddle-through story. Year-1 revenue grows ~2.5%, the midpoint of management's reaffirmed FY2026 guide trimmed for the Q1 -1.4% miss. Years 2-3 ramp to ~3.5% as BVA Family and infas annualize and the Horizons spend starts to compound; Years 4-7 taper to 3% as easy comparables fade and AI compresses pricing on the low end of survey work; Years 8-10 converge to the 2.0% terminal, modestly below the EUR risk-free. Operating margin grinds from 12.3% (FY2025) to 13.0% by Year 7 — half the Horizons uplift, capturing the Ipsos.Digital / GMS mix shift and continued gross-margin expansion (64.2% in 2020 → 68.7% in 2025) without betting on the full 13.5-14% management target. Reinvestment runs at 1.05 sales-to-capital, dominated by tuck-in M&A goodwill rather than capex — the model Ipsos has actually executed for 100+ deals. Terminal ROC pinned at 12%, a ~3pp spread above the new ~4.4% WACC.
Two debates worth pressure-testing
- β re-anchored to regression (0.59 vs industry 1.00). Sell-side uses bottom-up sector β; we trust the 5Y regression that two sources agree on. Lifts intrinsic ~10/sh vs the industry-β run.
- Margin lift only halfway to Horizons. Sell-side bulls model the full 13.5-14% management target; we land at 13.0% to discount the 2022-2025 plan miss and AI pricing pressure.
- Sales-to-capital 1.05, not industry 2.71. The M&A-goodwill operating model is the right baseline — applying the industry default would massively understate the capital base and overstate FCFF.
- Governance haircut 10%, above the 5% standard. French double-voting-rights gives DT & Partners 20% voting on 11.5% economic; not extractive, but minorities cannot block. Modest premium over the public-company baseline is warranted.
- AI displaces 30-50% of survey revenue by 2032. Synthetic data + PersonaBots become "good enough" for concept tests and tracking; organic turns -1 to -3% from 2027; margin compresses to 8-10% on volume deleverage.
- Plan miss repeats. Same management that missed 2022-2025 misses 2026-28; Q1 2026 -1.4% is the first datapoint of a multi-year underperformance, not a Horizons-ramp artefact.
- EUR strength persists. ~33% of revenue is Americas (mostly USD); sustained EUR appreciation continues to take 4-6 pts off reported revenue annually.
- M&A pace breaks ROIC. One bad large deal among the next 15 tuck-ins impairs goodwill (~€1.4B stock) and breaks the 12-14% ROIC band that supports the terminal spread above WACC.
- CEO transition stumbles. Stoclet is internal but Poitou is new in the Global CEO seat — Horizons execution risk above the baseline plan-slippage history.
Risks to thesis (tail, not bear case)
10-year forecast
Revenue €2.60B → €3.33B over 10y (~2.75% CAGR); operating margin grinds from 12.3% to 13.0% by Y7.
Monte Carlo distribution
With β re-anchored to the 0.59 5Y regression, the entire MC distribution sits above market: even the p5 floor at €66/sh is ~81% above today's €36.70. The DCF says either the regression β is wrong, or the market is.
1000 iterations randomising the central uncertainties. P(intrinsic < market €36.70) = 0.0%.
Mean €87.52 ± €13.14.
Cost of capital build
| Risk-free rate | 2.40% |
| Mature-market ERP | 4.23% |
| Levered β | 0.59 |
| Weighted CRP | 0.90% |
| Cost of equity | 5.43% |
| Pre-tax cost of debt (synth Aaa/AAA) | 2.09% |
| D / V | ~15% |
| WACC | 4.43% |
Full year-by-year DCF
| Year | Revenue | Op mgn | EBIT | EBIT(1−t) | Reinvest | FCFF | PV |
|---|---|---|---|---|---|---|---|
| 1 | €2.60B | 12.35% | €321M | €238M | €72M | €165M | €158M |
| 2 | €2.68B | 12.46% | €334M | €247M | €74M | €173M | €158M |
| 3 | €2.76B | 12.56% | €347M | €256M | €77M | €180M | €158M |
| 4 | €2.84B | 12.67% | €360M | €266M | €79M | €188M | €158M |
| 5 | €2.93B | 12.78% | €374M | €277M | €81M | €196M | €158M |
| 6 | €3.01B | 12.89% | €389M | €288M | €84M | €204M | €158M |
| 7 | €3.11B | 13.00% | €404M | €300M | €86M | €213M | €158M |
| 8 | €3.19B | 13.00% | €414M | €308M | €79M | €229M | €161M |
| 9 | €3.26B | 13.00% | €424M | €316M | €71M | €245M | €162M |
| 10 | €3.33B | 13.00% | €433M | €322M | €62M | €260M | €161M |
Methodology & flags
Damodaran FCFF DCF, 10y explicit + perpetuity. R&D capitalised;
IFRS-16 leases in book debt. Synth credit Aaa/AAA. CRP from
revenue-weighted country mix × Damodaran 2026 CRPs. Levered β
0.59 (5Y regression Yahoo/SimplyWallSt, overrides industry
prior 0.87→1.00). Monte Carlo: 1000 iterations. Engine v1.0.0 ·
result:
/Users/valentin/Documents/notes_jiliac_labs/Finance/Damodaran/valuations/ips/output/2026-05-25-result.json
- R&D cap: OFF · Lease cap: OFF · Failure: OFF · ESO: OFF
- Governance haircut: 10% (€106.92 > €96.23)
- Sensitivity tornado: not run