Instructor outcome
This episode is a bounded success: Microsoft completed the acquisition, continued to report LinkedIn as a distinct business, said LinkedIn had more than 700 million registered members in FY2020, and disclosed USD 19.817 billion of LinkedIn revenue in FY2026. The latest filing supplies a matched FY2024–FY2026 revenue series whose deterministic CAGR is 10.019100 percent. [judgment.linkedin.outcome.bounded-success; claim.linkedin.outcome.close-and-purchase-accounting; claim.linkedin.outcome.registered-member-scale; claim.linkedin.outcome.latest-matched-revenue-scale; table.linkedin.outcome.matched-latest-revenue]
The perimeter matters. This is success for transaction completion, disclosed business continuity and revenue scale—not proof of integration quality, attractive acquisition IRR, security alpha, causal synergy, target-price accuracy or an authorized exposure decision. The selected record lacks a complete standalone post-close profit and cash-flow series, attributable integration economics, timed acquisition cash flows, a counterfactual, point-in-time capitalization and security prices. [judgment.linkedin.outcome.return-valuation-abstention; claim.linkedin.outcome.return-and-valuation-unknown; table.linkedin.outcome.unresolved-return-inputs]
The frozen public-equity decision
Part A's decision time is June 14, 2016 at 14:30:00 UTC, with a knowledge cutoff of June 13, 2016 at 23:59:59 UTC. The learner is a public-equity underwriting committee assessing Microsoft after the signed announcement, not Microsoft's acquisition board. The packet therefore could not redesign or cancel the signed deal on the issuer's behalf. [claim.linkedin.cutoff.signed-terms; evidence.linkedin.cutoff.per-share; evidence.linkedin.cutoff.closing-conditions]
At the cutoff, LinkedIn had substantial reported revenue, registered-member and product scale, but it also reported an operating loss, material stock compensation and incomplete cash-economics evidence. [claim.linkedin.cutoff.revenue-and-member-scale; claim.linkedin.cutoff.product-diversification; claim.linkedin.cutoff.profit-and-cash-boundary; evidence.linkedin.cutoff.income; evidence.linkedin.cutoff.stock-comp]
The announcement said LinkedIn would retain its brand, culture and independence and that Jeff Weiner would remain CEO. That was a company plan, not evidence that the operating model had been executed. [claim.linkedin.cutoff.announced-operating-model; evidence.linkedin.cutoff.deal-terms]
A contemporaneous independent report characterized the bid as poor capital stewardship and argued that LinkedIn's revenue growth masked losses. Warren preserves those as third-party opinions and does not adopt their proprietary valuation. [claim.linkedin.cutoff.independent-price-discipline-warning; evidence.independent.cutoff.capital-stewardship-critique; evidence.independent.cutoff.growth-loss-critique]
Because capitalization, valuation, mandate, current exposure and acquisition cash-flow inputs were incomplete, Part A put Microsoft on a diligence watchlist, authorized no trade, authorized no initiation or increase, and referred every existing-exposure maintain, reduce or exit decision for authorized human re-underwriting. [judgment.linkedin.cutoff.defer-reunderwrite; judgment.linkedin.cutoff.return-abstention]
The current successor freeze is bound to SHA-256 bcad9cd8f23fa50e4f11c01519a2d40a22a248b72fbdd921e995785dda8ff1be and chains to the prior frozen digest. The successor change binds the deterministic model's inputs to fact IDs and records re-execution; it does not change the cutoff facts or conclusions. Retrospective reconstruction and the successor freeze enforce the cutoff and Part A/Part B phase separation, but they do not prove cognitive blinding. Later outcome evidence does not revise that knowledge state.
What happened after the freeze
Regulatory commitments: valid time and knowledge time remain separate
The European Commission decision is dated December 6, 2016 and says final commitments were sufficient to remove its serious doubts. The commitments described user ability to disable LinkedIn features in major Office products, OEM freedom not to install LinkedIn Windows applications or tiles, and a five-year duration. [claim.linkedin.outcome.eu-commitments; evidence.ec.outcome.commitments-sufficient; evidence.ec.outcome.user-disable-commitment; evidence.ec.outcome.oem-choice-commitment; evidence.ec.outcome.five-year-duration]
The selected government PDF does not provide a verified earlier page-publication timestamp. The corpus therefore preserves December 6, 2016 as the decision's economic/occurred date but conservatively sets public availability to retrieval on August 16, 2026. It cannot strengthen the June 2016 packet or be represented as contemporaneously known there. Exact evidence comes from a separately hashed pdftotext -layout derivative bound to the immutable PDF. source · ec.outcome.m8124-decision-pdf source · ec.outcome.m8124-decision-text
The decision documents remedy design, not later compliance, competitive welfare or the remedies' causal effect on LinkedIn. Those remain unknown. [claim.linkedin.outcome.regulatory-compliance-unknown]
The acquisition closed and purchase accounting was concentrated in goodwill and intangibles
Microsoft reported that it completed the acquisition on December 8, 2016 for a USD 27.009 billion GAAP purchase price and that the purchase price consisted primarily of approximately USD 26.9 billion of cash. Both differ in perimeter from the signing announcement's USD 26.2 billion transaction value inclusive of LinkedIn net cash. None supersedes another: announced transaction value inclusive of LinkedIn net cash, rounded cash consideration and purchase-accounting price remain separate sourced facts. [fact.linkedin.cutoff.announced-value; fact.microsoft.outcome.cash-paid; fact.microsoft.outcome.purchase-price]
The disclosed allocation included:
| Reported allocation | USD millions |
|---|---|
| Approximate cash paid | 26,900 |
| Acquired cash | 1,328 |
| Acquired short-term investments | 2,110 |
| Acquired intangible assets | 7,887 |
| Goodwill | 16,803 |
| Assumed short-term debt, positive liability magnitude | 1,323 |
Structured lineage: table.linkedin.outcome.purchase-accounting; [claim.linkedin.outcome.close-and-purchase-accounting; evidence.microsoft.outcome.close-and-ppa].
The deterministic outcome script calculates goodwill at 62.212596 percent, acquired intangibles at 29.201377 percent, and the two together at 91.413973 percent of purchase price. [model-run.linkedin.outcome-descriptive-checks] These are descriptive purchase-accounting shares, not impairment forecasts or realized returns.
The first reported post-close period was partial and amortization-heavy
For the period from the December 8 close through June 30, 2017, Microsoft reported USD 2.268 billion of LinkedIn revenue and a USD 948 million operating-loss magnitude. It said USD 866 million of the loss was acquired-intangible amortization. [claim.linkedin.outcome.early-results-boundary; evidence.microsoft.outcome.partial-results; evidence.microsoft.outcome.amortization-boundary; table.linkedin.outcome.early-partial-results]
The deterministic script computes amortization at 91.350211 percent of the reported loss magnitude and USD 82 million of loss magnitude after subtracting only that disclosed amortization. The USD 82 million figure is not normalized operating profit, cash flow or a full accounting adjustment; it simply shows why the reported partial-year loss cannot be read without its noncash purchase-accounting content. [model-run.linkedin.outcome-descriptive-checks]
Disclosed member and revenue scale expanded
Microsoft's FY2020 filing said LinkedIn had more than 700 million members. The structured fact records 700 million as a lower bound for registered members. It is not active use, retention or monetizable activity, and the cutoff filing had warned that member metrics were internally tracked and not independently verified. [claim.linkedin.outcome.registered-member-scale; evidence.microsoft.outcome.members-2020; evidence.linkedin.cutoff.measurement-risk]
The same FY2020 filing reported LinkedIn revenue of USD 5.259 billion in FY2018, USD 6.754 billion in FY2019 and USD 8.077 billion in FY2020. [claim.linkedin.outcome.revenue-scale-2020; evidence.microsoft.outcome.revenue-2020-presentation; table.linkedin.outcome.revenue-2018-2020]
The FY2026 filing provides the preferred matched series:
| Fiscal year | LinkedIn revenue, USD millions | Deterministic year-over-year growth |
|---|---|---|
| 2024 | 16,372 | Not computed from this three-year table |
| 2025 | 17,812 | 8.795505% |
| 2026 | 19,817 | 11.256456% |
The deterministic FY2024–FY2026 CAGR is 10.019100 percent, and the absolute increase is USD 3.445 billion. All three revenue values come from the same FY2026 filing presentation. [claim.linkedin.outcome.latest-matched-revenue-scale; evidence.microsoft.outcome.revenue-2026-matched; table.linkedin.outcome.matched-latest-revenue; table.linkedin.outcome.deterministic-scale-checks]
Original and recast revenue facts are both preserved
Microsoft's FY2023 filing originally presented LinkedIn revenue of USD 13.816 billion in FY2022 and USD 15.145 billion in FY2023. The FY2024 filing later presented recast values of USD 13.631 billion and USD 14.989 billion and said prior periods had been recast to conform to how Microsoft internally managed and monitored the business. [claim.linkedin.outcome.original-2022-2023-presentation; claim.linkedin.outcome.recast-2022-2023-presentation; evidence.microsoft.outcome.revenue-2023-original; evidence.microsoft.outcome.revenue-2024-recast]
| Fiscal year | Original presentation, USDm | Recast presentation, USDm | Recast minus original, USDm |
|---|---|---|---|
| 2022 | 13,816 | 13,631 | -185 |
| 2023 | 15,145 | 14,989 | -156 |
Structured lineage: table.linkedin.outcome.presentation-versioning; [claim.linkedin.outcome.presentation-differences; judgment.linkedin.outcome.presentation-versioning].
The original facts remain in the ledger. Each later recast fact has a supersedes_fact_id, so the corpus can reproduce what was reported at each knowledge date. The USD 185 million and USD 156 million differences are presentation changes—not evidence of period-over-period operating deterioration.
The long-horizon comparison is descriptive, not the primary growth measure
Comparing FY2020 revenue from the FY2020 filing with FY2026 revenue from the FY2026 filing gives a USD 11.740 billion increase, a 2.453510-times multiple and a 16.135405 percent six-year CAGR. [claim.linkedin.outcome.cross-presentation-scale; model-run.linkedin.outcome-descriptive-checks]
Those endpoints come from different presentation vintages, and intervening product-revenue figures were recast. The result is therefore labeled cross-presentation descriptive scale. It is useful for broad magnitude only. The primary matched growth measure remains the 10.019100 percent FY2024–FY2026 CAGR from one FY2026 filing. [assumption.linkedin.outcome.cross-presentation-comparability; table.linkedin.outcome.deterministic-scale-checks]
Why scale does not establish acquisition return
The selected evidence still lacks:
- a complete standalone post-close LinkedIn operating-profit series;
- a complete standalone cash-flow series;
- audited realized synergies, dis-synergies and integration cash costs;
- counterfactual LinkedIn or Microsoft cash flows without the transaction;
- timed incremental acquisition cash flows and terminal value;
- point-in-time diluted capitalization, net debt, security prices and forecasts; and
- an approved mandate, valuation methodology and risk budget.
Those gaps are explicit cells in table.linkedin.outcome.unresolved-return-inputs. A revenue-to-purchase-price ratio is not IRR, and Microsoft-wide shareholder returns cannot be attributed to LinkedIn from this record. [claim.linkedin.outcome.return-and-valuation-unknown; judgment.linkedin.outcome.return-valuation-abstention]
The outcome classification, evidence confidence, business-scale assessment and price attractiveness therefore remain separate. Later scale does not retroactively authorize Part A's missing trade decision.
Causality and the nearest rival explanation
The primary hypothesis is deliberately qualified: retained brand and operating continuity, combined with selective Microsoft product and distribution options, may have supported LinkedIn's continuity and revenue growth. The announced autonomy model, pre-close product breadth and later disclosed scale are consistent with that path. [hypothesis.linkedin.bounded-autonomy-and-product-optionality]
A credible rival is that LinkedIn's pre-existing trajectory, secular digitization, labor-market cycles and advertising demand explain a material but unquantified share of the later scale. [hypothesis.linkedin.secular-growth-and-preexisting-trajectory]
The selected public record is observational. It cannot allocate causal shares among autonomy, Microsoft distribution, LinkedIn execution, acquisitions, macroeconomic conditions or regulation. [assumption.linkedin.outcome.causal-attribution-unavailable]
Feasible counterfactual for the public-equity learner
Because the transaction was already signed, the learner could not turn the assignment into pre-signing board advice. The feasible counterfactual was to keep Microsoft on a research watchlist, defer every exposure action, and require authorized human re-underwriting only after separate closing, funding, autonomy, talent, customer, product, accounting, privacy, cash-economics and valuation gates cleared. [counterfactual.linkedin.watchlist-with-gated-reunderwriting; assumption.linkedin.outcome.gated-integration-counterfactual]
That counterfactual authorizes no hold: current exposure, mandate, valuation and risk budget remained missing. The record also cannot estimate the return from waiting.
What a financial-analysis agent should learn
Two candidate rules emerge:
rule.acquisition.separate-business-scale-from-acquisition-return— treat closing, continuity, scale, economics, return, business quality and price attractiveness as separate conclusions.rule.acquisition.gate-integration-under-announced-autonomy— treat an autonomy promise as a testable plan and require non-substitutable integration and underwriting gates before any human capital decision.
Microsoft–LinkedIn supports the rules as a bounded scale success; HP–Autonomy is the paired failure/counterexample case. That pairing does not prove either issuer used or omitted the rules, and two cases do not make either rule corpus-validated.
Remaining abstentions and hindsight limits
- FY2020 supplies the selected registered-member lower bound; the report does not project that member fact through FY2026.
- Revenue through FY2026 does not establish standalone profit, cash flow or integration quality.
- The selected Commission decision does not establish later compliance or welfare effects and was not verified as publicly available before its 2026 retrieval.
- Original and recast product-revenue presentations remain versioned; neither is silently erased.
- No acquisition IRR, intrinsic value, target price, position size, maintain/reduce/exit decision or shareholder-return attribution is produced.
- Human publication approval is absent, so this case is not approved for external release.
Source map
- Microsoft FY2017 Form 10-K — close, purchase accounting and partial-period results (
src.microsoft.outcome.2017-10k). - Microsoft FY2020 Form 10-K — FY2018–FY2020 revenue and the more-than-700-million member statement (
src.microsoft.outcome.2020-10k). - Microsoft FY2023 Form 10-K — original FY2022 and FY2023 product-revenue presentation (
src.microsoft.outcome.2023-10k). - Microsoft FY2024 Form 10-K — recast FY2022/FY2023 and FY2024 product-revenue presentation (
src.microsoft.outcome.2024-10k). - Microsoft FY2026 Form 10-K — matched FY2024–FY2026 product-revenue presentation (
src.microsoft.outcome.2026-10k). - European Commission Case M.8124 decision — immutable decision PDF and remedy text (
src.ec.outcome.m8124-decision-pdf; deterministic exact text insrc.ec.outcome.m8124-decision-text).
Every material reported number resolves through structured facts to hash-bound evidence. All derived arithmetic runs in deterministic Python with isolated Decimal context. The run is marked non-authoritative because no frozen model-registry execution receipt exists, and it deliberately computes no acquisition return or valuation.