Tools chosen: kalshi_direct claude_news gdelt_news kalshi_related polymarket_related wikipedia

Sub-questions (Fermi decomposition)

Planner reasoning

This is a long-horizon Kalshi market on whether Anthropic confirms an IPO before OpenAI (by 2040). Key drivers are each company's corporate structure (OpenAI's restructuring to a PBC, Anthropic's PBC status), recent funding rounds and valuations, executive statements about IPO plans, and investor liquidity pressures. The Kalshi price is the primary anchor, cross-checked against Polymarket and recent news on IPO signals from both companies.

# Current state Both companies have confidentially filed draft S-1s with the SEC (OpenAI ~May 22, 2026; Anthropic ~June 1, 2026), but neither has priced or completed an IPO. Anthropic is reportedly targeting fall 2026 (Sept–Oct) while OpenAI is reportedly weighing a delay to 2027 amid market volatility concerns. Kalshi prices Anthropic to IPO first at 93%. # Timeline of key events - 2025: OpenAI completes restructuring into OpenAI Group PBC, 26% owned by nonprofit OpenAI Foundation (confirmed, Wikipedia). - 2026-04: OpenAI closes funding round at $852B post-money valuation (confirmed, Wikipedia). - 2026-05-22: OpenAI files confidential draft S-1 with SEC (reported, smartasset/claude_news). - 2026-05: Anthropic closes Series H at $965B valuation (confirmed, Wikipedia/claude_news). - 2026-06-01: Anthropic files confidential draft S-1, ~1 week after OpenAI (reported, forgeglobal/benzinga). - 2026-06-03: Goldman Sachs, Morgan Stanley, JPMorgan named lead underwriters for Anthropic IPO (reported, techtimes). - 2026-06-25: Reuters/Forbes report OpenAI considering delaying IPO to 2027 after SpaceX's rocky public debut; original plan was late 2026, possibly Sept (reported, forbes). - 2026-08-19: Reporting details Anthropic's dual-class governance structure (super-voting shares + 3-member Long-Term Benefit Trust) for planned IPO (reported, techtimes). - 2026-10-23 (est.): Market consensus/banking sources point to Anthropic Nasdaq listing as early as this date (rumored, greyjournal/techtimes). - 2026-10-27: OpenAI reaches restructuring deal with Microsoft granting it a 27% PBC stake, resolving an impasse that had hindered OpenAI's IPO plans (reported, tipranks). - Q2 2026: Anthropic revenue ($11.5B) surpasses OpenAI ($6.7B) for first time; Anthropic posts first operating profit (~$559M) (reported, smartasset/greyjournal). # Event Will Anthropic confirm an IPO before OpenAI (i.e., IPO first), resolving by 2040-01-01. # Outcomes to forecast - Yes (Anthropic IPOs first) - No (OpenAI IPOs first, or ties/other resolution) # Kalshi market anchor KXOAIANTH-40-ANTH (Anthropic) currently priced at **93% YES**, up +10pts over 30 days, flat over 7 days. 90-day range 71%–94%. Avg daily volume ~627 contracts — moderate liquidity, market has been trending toward Anthropic with conviction. # Sub-question answers 1. **OpenAI structure/IPO permissibility** — 2025 restructuring converted the for-profit arm into OpenAI Group PBC, 26% (later 27% per Microsoft deal) owned by nonprofit OpenAI Foundation; this structure is generally viewed as IPO-compatible (Wikipedia, tipranks). 2. **Anthropic leadership IPO statements** — No direct Amodei quote found, but Anthropic has hired banks (Goldman, Morgan Stanley, JPMorgan) as underwriters and reportedly is preparing governance (super-voting shares, Long-Term Benefit Trust) for a public listing (techtimes, claude_news). 3. **OpenAI leadership/preparation** — Sam Altman reportedly pushing bankers for a $1T valuation IPO; CFO Sarah Friar cited a 2027 target while some advisers suggest late 2026 is possible; OpenAI filed confidential S-1 May 2026 (forbes, marketscreener). 4. **Funding/valuations/liquidity pressure** — OpenAI: $852B (Apr 2026) valuation, reportedly needs another pre-IPO round at ~$830B; heavy losses (~$14B projected 2026 loss), no profitability until ~2030. Anthropic: $965B (May 2026 Series H), revenue run-rate $47–65B, first profitable quarter in Q2 2026 (claude_news, cmcmarkets, forbes). 5. **Market signals** — Kalshi anchors at 93% for Anthropic; no Polymarket market specifically tracks this question (only unrelated Anthropic model-release market found). No conflicting cross-market signal. 6. **Structural obstacles** — OpenAI: nonprofit/PBC structure required Microsoft revenue-share renegotiation, resolved Oct 27, 2026 with 27% stake deal, removing a key IPO blocker. Anthropic: Long-Term Benefit Trust (shrunk 4→3 members) retains power to elect board majority; dual-class shares preserve founder control — governance complexity but not legally blocking (claude_news, tipranks). # Key facts (high-confidence, factual) 1. [Wikipedia] OpenAI filed for IPO in June 2026; restructured to PBC in 2025 (26–27% nonprofit-owned). 2. [Wikipedia] Anthropic valued at $965B (May 2026); "reportedly plans IPO fall 2026." 3. [claude_news] Both firms confidentially filed draft S-1s in May–June 2026, ~1 week apart. 4. [claude_news] Anthropic overtook OpenAI in quarterly revenue in Q2 2026 and posted first operating profit. 5. [tipranks] OpenAI-Microsoft restructuring deal (27% stake) reached Oct 27, 2026, removing a prior IPO impasse. # Cross-market signals - Kalshi related: "US to take 1% stakes in both OpenAI and Anthropic" at 16% (unrelated but shows active AI-IPO market ecosystem); no direct competing IPO-timing market. - Polymarket: No market directly on OpenAI/Anthropic IPO order found. - Sportsbook implied: N/A. # Analyst opinions and speculation - Advisers reportedly frame OpenAI's choice as "wait until 2027 for $1T valuation vs. list sooner at lower valuation" (Forbes/smartasset) — suggests OpenAI leadership may deliberately delay. - Banking sources speculate Anthropic Nasdaq listing as early as Oct 2026 (rumored, unconfirmed date). - claude_news synthesis: "Anthropic appears to be moving faster toward an actual listing... suggesting Anthropic may currently be favored to IPO first." # Directional lean per outcome - **Yes (Anthropic first):** Supported by faster revenue growth/profitability, earlier reported target date (fall 2026), underwriters already named, Kalshi pricing (93%) trending up. Some risk: exact filing sequence ambiguous (OpenAI filed confidential S-1 first chronologically per Wikipedia). - **No (OpenAI first):** Supported by OpenAI's S-1 filing predates Anthropic's by ~1 week (per one source), Microsoft deal resolution (Oct 2026) removes blocker, some advisers suggest OpenAI could list "late 2026," similar timeframe to Anthropic. But explicit reporting leans toward OpenAI delaying to 2027. # Gaps / unknowns - No confirmed actual IPO date for either company as of latest data (both still pre-listing). - Conflicting reports on filing sequence (OpenAI first per Wikipedia/chronology vs. Anthropic "ahead" per Benzinga framing). - GDELT tool failed (timeouts), limiting independent news cross-check. - No Polymarket equivalent market to validate Kalshi pricing. # Calibration anchors - Kalshi current YES price: 93% (anchor), up from 71% low over 90 days — reflects growing market conviction Anthropic wins the race. - Precedent: In similar "X or Y IPO first" Kalshi markets (e.g., Deel/Rippling 63%, Ramp/Brex 3%), leading candidate by revenue/momentum tends to be heavily favored once banker/underwriter news breaks — consistent with current Anthropic pricing.

1. Both forecasts land on identical 0.92, essentially reproducing the Kalshi price rather than triangulating independently — given the explicitly noted "conflicting reports on filing sequence" and OpenAI's earlier S-1 filing (May 22 vs. June 1), a genuinely independent estimate should show more spread from the anchor, not converge exactly on it. 2. Neither forecast grapples with the extremely long resolution horizon (close 2040-01-01) — the "No" outcome also captures "ties/other resolution" (e.g., one or both companies never IPO, get acquired, or dissolve), yet both rationales treat this purely as a 2026-vs-2027 OpenAI/Anthropic race, ignoring multi-year tail scenarios that could shift probability meaningfully over a 14-year window. 3. The brief flags moderate liquidity (~627 contracts/day) and a rapid 22-point swing in 30 days (71%→93%), which suggests the market itself may be unstable/thin rather than a highly reliable consensus signal — neither forecaster discounts for this volatility or considers reverting toward the 71% low as a plausible re-anchor point. 4. Both cite the calibration anchors "Deel/Rippling 63%" and "Ramp/Brex 3%" as precedent but neither explains why this case resembles the far more extreme 3% case rather than the more moderate 63% case, despite comparably uncertain/conflicting timeline evidence for both companies. 5. Neither forecast weighs OpenAI's earlier confidential S-1 filing (chronologically first) as a meaningful signal that OpenAI's internal IPO machinery could move faster than the "2027 delay" narrative suggests, especially now that the Microsoft blocker was resolved Oct 27, 2026 — this is acknowledged as "residual risk" but not quantified, leaving the 8% "No" allocation potentially too thin given two independent bullish OpenAI signals (filing timing + blocker resolution). 6. Confidence scores differ (0.68 vs. 0.74) yet both produce the exact same probability distribution, indicating the confidence metric isn't actually informing differentiated uncertainty bands (e.g., a wider interval or hedged alternative outcome allocation) despite ostensibly different conviction levels.

90% (71% confidence)

Yes: 90% No: 10%

Both forecasts converge on 0.92, closely tracking the Kalshi 93% anchor, and the core evidence strongly favors Anthropic: named underwriters, IPO-ready governance, first operating profit, and a reported fall 2026 Nasdaq target, while OpenAI's own leadership signals a deliberate delay to 2027 for a $1T valuation. However, the critique correctly identifies underweighted 'No' pathways: OpenAI filed its confidential S-1 first, the Microsoft blocker was resolved in Oct 2026 enabling a possible late-2026 pivot, the Kalshi market is thin and swung 22 points in 30 days, and the 2040 horizon includes tail scenarios (Anthropic delay/withdrawal, ties, non-IPO outcomes) that fall into 'No'. These considerations justify shading slightly below both the market and consensus rather than above, but not dramatically, since no concrete evidence suggests OpenAI will actually list before Anthropic's October window. Final call: 0.90 Yes, 0.10 No.

Total pipeline time: 207.0s