ASK YOUR STAKEHOLDER   ·   WARREN WISE REPORT 2026

PRESS RELEASE

An incorruptible reader enters the capital market — and hardly any German blue chip is ready

For the first time, an AI has analysed 40 DAX and MDAX corporations — from A for Adidas to Z for Zalando — the way a long-term value investor would: without self-interest, rule-based, solely on the basis of audited annual reports. The verdict: plenty of substance, little top class. A grade of 3.1 out of 5, with the top tier “Premium” left empty. And very few companies appear ready for what is about to become the standard.

• No “Premium”: 0 of 40 reach the top, average 3.1 of 5 — a watchlist universe.
• Substance yes, renewal and stewardship thin: 34 of 40 strong on substance, but only 11 with demonstrated innovative strength (e.g. SAP, Infineon, Merck, Siemens Healthineers, Siemens), only 5 also cleanly steered (Airbus, Heidelberg Materials, SAP, Scout24, Deutsche Telekom).
• Three recurring patterns: asymmetric adjustments (e.g. Zalando, Continental), distributions above free cash flow (e.g. BASF), goodwill above equity (e.g. Hochtief, 215%).
• Transparency paradox: of all companies, it is the embattled ones — Bayer, E.ON, BASF, Lufthansa, RWE — that report most openly.
• Agentic Finance: AI reads annual reports en masse — hardly any corporation appears prepared for it.
• The full “Warren Wise Report 2026” is available at warren-wise.com.

Wiesbaden, 30 June 2026 — A new reader is entering the capital market: one that analyses every annual report, never tires and knows no self-interest. The multi-agent system Warren Wise has analysed 40 heavyweights from the DAX and MDAX the way a long-term value investor would — and paints a sober picture: not premium, but a watchlist universe full of caveats and curated metrics.

A single large watchlist
The average grade is 3.1 on a scale from 1 (knockout zone) to 5 (Premium): Warren Wise rates 14 corporations as strong, 20 sit at watchlist level, 6 below. “Premium” remains unoccupied; SAP comes closest.

With each of the three assessment levels, the picture of the top corporations weakens. Substance — business model and earnings power — is high almost everywhere (34 of 40 strong). On future-fitness the field splits: 30 corporations are financially robust, yet only 11 show demonstrated innovative strength. And on stewardship — capital discipline, governance, risk reporting — only five come close to real strength: the same five that also lead on substance.

The pattern is the same across the whole sample: the strength of these corporations lies in what they have built over the years — not in what they make of it today. From a value perspective, a well-known name and a strong balance sheet do not yet amount to a convincing equity story.

“We started with the value investor because he answers the question almost everyone has about a company — the supervisory board, employees, suppliers, customers, each in their own currency: will it remain sound and adaptable enough to endure? The value investor turns that into method: figures against the narrative, strategy against reality.”
— Ariane Hofstetter, Managing Director, CPT

Accountability — or a curated self-portrait?
An annual report fulfils a legal duty — accountability to the owners. But what is reported is prescribed; how it is framed, weighted and emphasised is up to management. In that latitude, a flattering, carefully curated self-portrait takes shape. It catches no one’s eye — it becomes visible only through patient, rule-bound analysis: relating figures the report discloses in separate places, sometimes buried deep in footnotes or the appendix, and comparing 40 reports with one another. Importantly: all companies operate within the bounds of permissible, audited accounting. Warren Wise makes no allegation of any rule-breach but assesses the effect from a value perspective.

Adjusted asymmetrically. 35 of the 40 corporations place a self-defined “adjusted” figure alongside the reported result. The range is enormous: the median is 12.4%, yet eleven corporations adjust beyond the 25% reference mark. What matters to Warren Wise is not the magnitude alone but the direction — when every adjustment runs in the company’s favour (add-backs only), earnings appear more favourable than they are. At Zalando, for instance, the adjustment ratio jumped within a single year from 30.4% to 52.6% — a gap of €203.5m between reported (€387.2m) and adjusted EBIT (€590.7m). The extreme cases are Continental at 648% (spin-off year) and Bayer at 574% (litigation) — both for an identifiable reason. It can be done differently: six corporations adjust conservatively, and 5 corporations — among them Adidas and Nordex — forgo adjusted metrics entirely.

Distributions above free cash flow. At seven corporations, the 2025 distribution to shareholders (dividend and share buyback) exceeded free cash flow. Most markedly at BASF: dividend and buyback summed to €2,363m, while free cash flow came to €1,342m — the difference drawn from balance-sheet reserves, for the second year running. In the study’s image: here the seed corn is being consumed instead of securing the next harvest.

Hidden impairment risk. At four corporations the goodwill carried on the balance sheet exceeds total equity — most pronounced at Hochtief at 215% with an equity ratio of just 5.7%, followed by Hensoldt (112%), Nemetschek (110%) and Bayer (108%). Goodwill is not amortised on a schedule but tested annually for impairment (IAS 36); on the numbers, a single larger impairment cycle could here largely consume the equity base.

The transparency paradox. The most surprising twist: it is, of all companies, the recently embattled ones that disclose their risks most openly — Bayer, E.ON, BASF, Lufthansa and RWE.

The bigger picture: Agentic Finance
Exactly this way of reading is now becoming the standard. Warren Wise is no isolated case but part of a development known by the term Agentic Finance: AI systems evaluate disclosures against fixed criteria, condense them into signals and pass these on to trading agents. Such readers do not negotiate, do not soften out of politeness — and do not warn in advance.

“This study is, to the best of our knowledge, the first public demonstration of how such a system reads annual reports at scale, independently recalculates the figures and assesses the equity stories.”
— Michael Diegelmann, Managing Director, CPT

How Warren Wise judges
Warren Wise is not a clever one-off prompt but a self-contained multi-agent system with clear responsibilities — so every judgement remains reproducible, comparable and traceable to a concrete passage in the report.

Four sector types were left out for now, because their balance-sheet logic would distort a comparison on the same scale: banks, insurers, real estate and captive-finance automotive groups. The study was conducted between 1 May and 8 June 2026; the assessment framework was validated with independent value investors. Every finding is produced with AI support and then checked by a qualified person against the disclosed passages of the reports (human in the loop).

About Ask Your Stakeholder / Warren Wise
Warren Wise is a rule-based multi-agent system that analyses equity stories from the perspective of a long-term quality and value investor. “Ask Warren” makes the framework available as an interactive Q&A application and helps companies prepare for Agentic Finance. The study is published by CPT GmbH. The full study is available at warren-wise.com.

Press contact
Ariane Hofstetter · hofstetter@askyourstakeholder.com
Michael Diegelmann · diegelmann@askyourstakeholder.com
Web: warren-wise.com

Disclaimer: This publication serves general information purposes only and does not constitute investment advice, research in the regulatory sense, or any recommendation to buy, hold or sell financial instruments. The scores, categories and classifications it contains are model-based value judgements as at a given date, following the disclosed methodology, and are not statements of fact about the companies concerned. The assessments are produced with AI support and may contain errors. “Warren Wise” is a fictional, synthetic figure; echoes of the names Benjamin Graham, Warren Buffett or Charlie Munger describe solely the underlying investment philosophies and establish no connection. DAX® and MDAX® are registered trademarks; the study has no connection to ISS STOXX or Deutsche Börse AG. Business or personal relationships (including with the cometis group) may exist that could be relevant as a potential conflict of interest.