{"paper_id":"2609.24181","title":"Firm Valuation When AI Shapes the Business Model: A Milestone-Based Real-Options Framework for the AI Valuation Uncertainty Problem","authors":["Walter Kurz","Wojtek Stricker","Stefan Marx","Frank Reinhardt","Florian Kollberg"],"paper_date":"2026-09-21","paper_type":"survey-review","paper_url":"https://arxiv.org/pdf/2609.24181v1","page_url":"https://thequant.space/flowcharts/firm-valuation-when-ai-shapes-the-business-model-a-milestone-based-real-options/","scores":{"empirical_rigor":4,"math_complexity":7,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes an industry‑agnostic taxonomy of AI Integrators and Providers with integration‑depth levels, and a milestone‑gated real‑options overlay that decomposes firm value into five components using an AHP‑based Success Readiness Index to derive option‑level probabilities.","quote":"We propose an industry-agnostic taxonomy separating AI Integrators from AI Providers. A milestone-gated real-options overlay decomposes milestone state value into five components, and an Analytic Hierarchy Process-based Success Readiness Index derives per-option probabilities from structured pairwise comparisons for scenario analysis.","summary":"Standard valuation methods, including discounted cash flow, the income approach standard IDW S 1 of the Institute of Public Auditors in Germany, and market multiples, compress milestone probabilities, continuation options, and risk shifts into opaque aggregate parameters; none provides a structured…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Discounted cash flow (DCF) provides the necessary valuation foundation.","type":"model","stated":"explicit","quote":"The standard discounted cash-flow framework values a firm as V0 = T X t=1 CF t (1 + r)t + TV (1 + r)T (1) where CF t denotes expected free cash flow, r the cost of capital, and TV a terminal value capturing perpetual growth.Koller et al., 2020 This architecture works when the cash-flow engine is stable and projectable from historical patterns.","support":"full"},{"assumption":"Each milestone state has an independently estimated probability of success (ps).","type":"model","stated":"explicit","quote":"Each milestone state s carries a conditional value Vs and an independently estimated probability of success ps; the model prices the firm’s AI component as a portfolio of such milestone-gated options.","support":"full"},{"assumption":"Success Readiness Index scores can be calibrated to probabilities using IDL‑specific base rates derived from comparable technology adoption outcomes.","type":"data","stated":"explicit","quote":"The per-option probability ps is then obtained by calibrating the SRI against IDL-specific base rates derived from comparable technology adoption outcomes.","support":"full"},{"assumption":"Expert raters are auditable and their weights are aggregated via geometric mean per Saaty’s protocol.","type":"execution","stated":"explicit","quote":"Multiple experts score each option independently via the same AHP hierarchy; results are aggregated using the geometric mean, preserving the ratio scale per Saaty’s group-aggregation protocol.","support":"full"},{"assumption":"Milestone probabilities update Bayesian‑style as gates are cleared, reducing risk and raising subsequent ps.","type":"model","stated":"explicit","quote":"As milestones clear, ps for the next milestone updates upward (demonstrated capability reduces uncertainty), rs decreases (risk resolved at each gate), and new continuation options may emerge that were not identifiable at the earlier valuation date.","support":"full"},{"assumption":"Option values are additive across milestones, with double‑counting avoided by the framework’s overlap handling.","type":"model","stated":"explicit","quote":"If Option A (AI customer scoring) and Option B (AI pricing) both claim credit for the same revenue uplift, summing their Vs values double-counts the gain.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Case Firm E12 internal data","provider":null,"kind":"other","access":"proprietary","period":null,"universe":"Swiss AI‑native SaaS startup (energy portfolio optimisation)","quote":"o.21901247 | Machine-readable record: https://journal.swissi-ai.institute/doi/f3jtignfyge3 Case Firm E12 is a Swiss AI-native SaaS startup building an automated energy portfolio optimisation platform.1 Its core product, E12-eAI, trades simultaneously across Day-Ahead, Intraday, Futures, power purchase agreements, flexibility markets, and CO2 certificates, Revenues combine platform licences, performance fees tied to","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"none"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Applied to an AI‑native energy SaaS firm, the framework yields a coherent valuation band that traces to identifiable option‑level assumptions.","kind":"theoretical","quote":"Applied to an AI-native energy software-as-a-service firm, the framework yields a coherent valuation band traceable to identifiable option-level assumptions.","support":"full"},{"finding":"Risk concentrates in later‑stage continuation options, matching the structural prediction for AI Providers.","kind":"theoretical","quote":"Risk concentrates in later-stage continuation options, matching the structural prediction for AI Providers.","support":"full"},{"finding":"Shifting a single option’s probability from 0.3 to 0.6 moves the overall AI‑adjusted valuation by 40%, indicating where risk concentrates.","kind":"theoretical","quote":"If shifting one option’s ps from 0.3 to 0.6 moves VAI by 40%, that option is where risk concentrates.","support":"full"},{"finding":"Standard DCF applied to the case firm would produce a single present value whose sensitivity is dominated by terminal growth and discount rate, obscuring AI‑related assumptions.","kind":"theoretical","quote":"A conventional DCF applied to Case Firm E12 would yield a single present value whose sensitivity concentrates in the terminal growth rate and the discount rate, two parameters that absorb everything the analyst believes about AI integration without revealing what those beliefs are.","support":"full"},{"finding":"The decomposition layer (IDL taxonomy, five‑component Vs, AHP‑based SRI) makes embedded assumptions explicit and auditable, allowing analysts to locate disagreements to specific probabilities or cash‑flow components.","kind":"theoretical","quote":"It adds a decomposition layer, the IDL 0–3 taxonomy, the five-component Vs structure, and the AHP-based SRI protocol, that makes the embedded assumptions explicit and individually auditable. Two analysts who disagree on the CHF 100 million pre-money can now locate their disagreement in pM4, in Vc,M4, or in the 30% discount rate.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"not_extracted","supports":[],"contradicts":[],"qualifies":[],"compared_with":0,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":2,"extracted_at":"2026-11-14T19:39:33Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}