CHAPTERS Group: The Loss Is Accounting, the Price Is the Problem
HGB accounting forces a book loss while cash grows 64% a year. Still: watchlist, not a core position. At 20x EBITDA there is no margin of safety.
The CHAPTERS Group annual report for 2025 contains two numbers that appear to contradict each other: a consolidated net loss of €29.7m and free cash flow of €16.8m [1]. A company that loses money and makes money at the same time. Anyone who reads only the first number sees an overpriced loss-maker: earnings per share of −€0.69, a negative P/E, EV/sales above 12. Anyone who reads only the second sees a compounder in its build-out phase. I have worked through both, and the answer does not lie in the middle.
CHAPTERS is a serial acquirer modelled on Constellation Software, with one difference that changes everything: the group reports under HGB, German GAAP. German accounting rules require goodwill to be amortised on a scheduled basis, and it is precisely this amortisation that produces the loss. Cash is untouched by it. So the question is not whether CHAPTERS is profitable. It is. The question is whether the market has long since paid for that.
The core business earns around 19% on its operating capital, free cash flow has grown 64% a year since 2021, and the reported loss is an HGB artefact. On the purchase price actually paid, however, CHAPTERS earns only around 4%, less than its cost of capital; interest consumes 62% of operating cash flow, and close to a third of the cash flow belongs to minorities. At €31.35 the share price is pricing in a finished compounder that has yet to become one. Watchlist, not a core position.
This article is part of a six-part series on vertical-market-software compounders. The other parts: Constellation Software, Lumine Group, Topicus, Asseco Poland and the VMS comparison. The topic dossier Tokenisation provides the framework for the group’s AI risk.
What CHAPTERS is
CHAPTERS buys small and mid-sized companies with mission-critical software: administrative solutions for public authorities, industry software for the German Mittelstand, financial services for international students. At the end of 2025 that meant 60 operating companies with more than 1,300 employees, pro-forma total output of €195m and adjusted operating EBITDA of €49.1m [2]. The logic behind it is Constellation’s: fragmented niches, high switching costs, long customer relationships, and a capital allocation system that collects the many small cash flows and channels them into the next acquisition.
The difference from the model lies in age. Constellation has three decades of history, CHAPTERS in its current form five years. That is not a flaw, but it does mean that the decisive question for any serial acquirer, whether the capital paid earns its cost, is still open here.
The foundation: a loss that costs no cash
At an IFRS reporter like Constellation, true earnings power disappears behind the amortisation of acquired customer and technology assets; goodwill itself stays untouched. At CHAPTERS it is the other way round. HGB requires straight-line amortisation of goodwill, here over ten years. In 2025 that meant €35.5m of goodwill and €8.3m of intangibles from purchase price allocation, €43.9m in total, running through the income statement and pushing EBIT down to −€22.9m [1].
I consider the usual split “including and excluding goodwill” too narrow. The capitalised brand and software assets are economically part of the purchase price paid as well. I therefore calculate including and excluding all acquisition intangibles. That produces the bridge on which the entire analysis rests:
| Item (2025) | €m |
|---|---|
| Reported EBIT | −22.85 |
| + Goodwill amortisation (required under HGB) | +35.54 |
| + Amortisation of PPA intangibles | +8.33 |
| = Leonard EBIT (excl. acq. intangibles) | +21.02 |
The cash flow statement confirms this. In 2025 a consolidated net loss of €29.7m stood against operating cash flow of €18.5m and free cash flow of €16.8m [1]. Depreciation and amortisation exceed the loss by one and a half times. And cash is not growing in leaps but in a straight line:
ROIC the Leonard way: two truths
From the Leonard EBIT, at a 30% tax rate, comes NOPAT of €14.7m. I checked the tax rate against the actual cash tax rate; in 2025 it was around 31% [1]. I set this NOPAT against invested capital in both readings:
| Reading | NOPAT | Invested capital | ROIC | Spread to 7.9% WACC |
|---|---|---|---|---|
| Incl. acq. intangibles (purchase price paid) | 14.7 | 563.4 | 2.6% | −5.3pp |
| Excl. acq. intangibles (core business) | 14.7 | 113.2 | 13.0% | +5.1pp |
The year 2025 carried three one-off burdens: three restructuring cases in the Public segment with a combined −€10.1m of EBITDA, the costs of the FinTech merger and the placement of the bond [1]. Strip them out and ROIC excluding intangibles rises to around 19%, and including them to around 4%. Both figures are estimates based on the adjustments disclosed in the annual report; the order of magnitude is robust, the decimal place is not.
“The core business is highly capital-efficient. On the purchase price paid, however, CHAPTERS does not yet earn its cost of capital. That gap separates an early-stage serial acquirer from the mature Constellation.”
Where the value creation comes from and what it costs
Value creation is the product of spread, reinvestment rate and growth. The core of CHAPTERS ties up hardly any capital: €1.7m of organic investment with high cash conversion [1]. Growth comes almost entirely from acquisitions; in 2025, €131.8m flowed into acquisitions, 7.8x free cash flow [1]. This is financed externally, filtered through an internal hurdle: shareholder loans to the operating units cost 10%, and an acquisition has to earn considerably more to justify them.
But the cash flow statement also shows what this flywheel costs. Interest payments rose from €0.4m in 2021 to €11.5m in 2025, a 28-fold increase [1]. They now absorb 62% of operating cash flow. The causes are the €72m bond at 7% and the acquisition loans; net debt stands at 3.87x EBITDA [3]. As long as ROIC on the purchase price sits below the cost of capital, every debt-financed acquisition widens this gap rather than closing it.
A second deduction follows: a substantial part of the cash flow does not belong to CHAPTERS shareholders. Minorities hold €89.7m of €313.6m in consolidated equity, or 29% [1]. The largest EBITDA block, the FinTech segment with around 39% of adjusted EBITDA, is only 61.7% owned by CHAPTERS. Of the reported €16.8m of free cash flow, roughly €12m is left for the shareholder on a back-of-the-envelope calculation.
The optionality: AI on proprietary data
Since October 2025, CTO Tobias Pook has been driving a portfolio-wide AI initiative: AI Hub, maturity framework, the Momentum initiative [4]. The logic convinces me: the operating companies sit on decades of structured customer data that an AI start-up cannot replicate overnight. The first recurring revenues are already emerging, at Icomedias in the police sector, at HUP in publishing [4]. Management does not expect the broad effect on the portfolio before 2027, however. That is an option, not a base-case expectation, and I therefore leave it out of the valuation core.
Valuation: two methods, one result
I value today’s owner cash flow with an FCFE model: levered free cash flow after interest, discounted directly at the cost of equity, without deducting net debt a second time. Three scenarios:
Bear Case
€7.68
- FCF₀ €11.0m, growth 8%, terminal 1.5%, Ke 9.5%
- Margin normalisation disappoints
- −76% vs. share price
Base Case
€14.39
- FCF₀ €12.5m, growth 14%, terminal 2.5%, Ke 8.55%
- DCF anchor value
- −54% vs. share price
Bull Case
€24.56
- FCF₀ €14.0m, growth 20%, terminal 3.0%, Ke 8.0%
- M&A machine scales
- −22% vs. share price
No scenario justifies the price of €31.35. Even the bull case sits 22% below it. Now, I know that an FCFE model by construction underestimates a serial acquirer in its build-out phase, because it does not capitalise the acquisition machine. Hence the cross-check via an EBITDA multiple, calculated on the adjusted operating EBITDA of the holdings of €49.1m, less net debt and a minority share of around 27%:
| EV/EBITDA | Equity CHG (€m) | per share | vs. share price |
|---|---|---|---|
| 14x | 399 | €16.73 | −47% |
| 16x | 471 | €19.74 | −37% |
| 18x | 542 | €22.75 | −27% |
| 20x | 614 | €25.75 | −18% |
The market is currently paying around 19.9x [3]. Both methods say the same thing: the price already contains a successful compounder that has grown into its growth rates and margins. For the growing-in itself, the buyer gets nothing extra.
Bear versus bull
- Core business with around 19% ROIC excluding intangibles: a genuine, capital-efficient engine
- FCF growth of 64% p.a. since 2021; the book loss is an HGB artefact
- Disciplined allocation via the 10% hurdle, no dividend, full reinvestment
- Guidance for organic EBITDA growth in 2026 raised at the AGM from 14–17% to more than 22% [2]
- AI optionality from 2027 on a data base that cannot be copied
- ROIC on the purchase price below WACC: every acquisition at today’s prices destroys value on paper
- Interest service absorbs 62% of operating cash flow
- Around 29% of equity and an even larger share of FinTech cash flow belong to minorities
- VSOP dilutes (strike price €22.27, capped at share prices above €46) [1]
- 46% of adjusted operating EBITDA 2025 consisted of adjustments [2]; the metric is half definition
- 20x EBITDA with no margin of safety
What has come in since June
The AGM on 13 July 2026 delivered two things and failed to deliver three [2]. Delivered: guidance for organic growth of adjusted operating EBITDA in 2026 was raised from 14–17% to more than 22%, and the company named for the first time a target range for adjusted earnings per share in 2027, €0.80 to €1.10, after −€0.05 in 2025. The presentation itself also discloses that €22.7m of the €49.1m of adjusted operating EBITDA in 2025 were adjustments, 46%, after 14% in 2023; for 2026 the aim is normalisation to below 15%.
Not delivered were answers to the three questions I had formulated for the AGM: on what metrics do the largest operating companies license their software, how high is the share of recurring revenue, and will the AI modules be priced separately by usage or bundled into existing maintenance fees? As long as that stays open, the pricing resilience that a 20x multiple presupposes remains unproven.
The scenarios are deliberately unweighted, because none of them reaches the share price; a weighted fair value would only add false precision. Four further points rest on my own assumptions: the 30% tax rate (checked against the 31% cash tax rate, but not identical); the classification of the €10.1m of one-off burdens as entirely non-recurring; the 27% minority share of equity value in the multiple calculation, derived from the 29% share of equity; and the 20x comparison multiple, which is anchored to the Constellation group rather than to German small caps. Anyone who sets one of these differently shifts the fair value, but not the sign of the gap to the share price.
Conclusion
CHAPTERS is a structural Constellation clone with a genuine, capital-efficient core. Two things separate it from its model. First, HGB hides the profitability: the loss is accounting, cash is growing 64% a year. Second, the capital paid does not yet earn its cost, and the cash flow that reaches the shareholder is, after interest and minorities, considerably thinner than the group figure suggests.
Anyone buying today is betting that margins normalise in 2026 and 2027 and that the company grows into its interest burden. That is a legitimate bet. It is just not a bet on a mispricing, because both valuation methods show a negative margin of safety at the price of €31.35.
Rating: Hold, category Watchlist. A small, observing tranche is defensible. A significant position requires either a share price heading towards €20 to €22, which would create a real buffer, or robust evidence that ROIC on the purchase price exceeds the cost of capital. Monitoring: the 2026 adjustment ratio (target below 15%), organic EBITDA growth against the raised guidance of more than 22%, and whether the next AGM answers the three open questions.
Update log
- 05.09.2026 — Revision: AGM materials of 13.07.2026 incorporated (guidance upgrade, 2027 EPS target, adjustment ratio), list of sources added, rating changed from “Watchlist” to the schema rating “Hold” with category “Watchlist”, language revised. Valuation model and fair values unchanged; price basis still 05.06.2026.
- 05.06.2026 — First published, based on the 2025 annual report.
Sources
- CHAPTERS Group AG, Annual Report 2025 (HGB consolidated financial statements, audited by BDO, unqualified opinion), published May 2026. Consolidated income statement, cash flow statement, notes on goodwill amortisation, minorities, VSOP. → https://www.chaptersgroup.de/wp-content/uploads/2026/05/CHAPTERS-Group-AG-Geschaftsbericht-2025.pdf
- CHAPTERS Group AG, Presentation for the 2026 Annual General Meeting, Hamburg, 13.07.2026. Pro-forma 2025 key figures, reconciliation of adjusted EBITDA, guidance upgrade, 2027e EPS target, balance sheet as at 31.12.2025. → https://www.chaptersgroup.com/wp-content/uploads/2026/07/20260713_CHAPTERS_HV.pdf
- TIKR Terminal, market data for CHAPTERS Group AG (XETRA: CHG), as at 05.06.2026: price €31.35, EV/EBITDA, net debt/EBITDA.
- CHAPTERS Group AG, “AI Interview” with CTO Tobias Pook, May 2026. → https://www.chaptersgroup.de/wp-content/uploads/2026/05/Chapters_AI-Interview_Pook_05-25.pdf