Question: AI mega-IPO issuance-surge base rate — does an IPO wave at all-time highs predict a market top Verdict: issuance-surge-is-a-yellow-flag-not-a-sell-trigger; robust-baker-wurgler-signal-needs-an-aggregate-wave-not-three-deals; lone-mega-ipo-broad-top-hit-rate-~1-in-7; spacex-glencore-analog-top-tick-for-ai-infra-complex-not-broad-sp; ritter-ps-over-40-rule-fires-on-spcx-itself-86pct-underperform

What we're asking

Does the "blockbuster IPO at the top = bubble burst" intuition hold? Calibrates the lane's Regime-Top / Issuance-Surge hypothesis (added 2026-06-02). Three sub-questions: (1) is equity-issuance volume a reliable forward-return predictor? (2) can a single mega-IPO top-tick a market, or does it take a wave? (3) which template does the 2026 giga-IPO stack (SpaceX + Anthropic + OpenAI, ~$200B raised / ~$4T value, into ATHs with QQQ RSI ~78) actually match?

What we found

1. The robust signal is the WAVE, not the deal (Baker-Wurgler)

Baker & Wurgler, "The Equity Share in New Issues and Aggregate Stock Returns" (J. Finance 2000): the share of equity in total new issuance (S = equity ÷ (equity+debt)) predicts one-year-ahead market returns; top-quartile S → ≈ −8% next-year equal-weighted real return, stable across 1928–1997, survives controls; the small-sample/"pseudo-market-timing" critique explains only ~1% of the power. Mechanism: managers collectively time equity issuance into peak sentiment, and the supply surge reverses equity scarcity. By construction this is an AGGREGATE signal — a single deal can't move the ratio. Equity share + IPO volume + IPO first-day pops are 3 of the 6 inputs to the Baker-Wurgler sentiment index.

Caveat (disconfirming, for balance): a contrarian camp (Statsedge) argues the crash-prediction story is overfit to 1999–2000 and 2021; IPO-heavy years since 1995 actually averaged higher forward returns, and IPO droughts (2002/2009/2020) are the cleaner bullish signal. Reconciles with Baker-Wurgler being a low-R², 1–3yr predictor: high issuance lowers expected returns and raises tail risk but is not a dated timing edge — momentum can run for many months first.

2. Lone mega-IPO → broad-market top: ~1 in 7 (case study)

Clean broad-market top-ticks: 1/7 (~14%); ~2–3/7 if you count sector tops + credit Aramco. A lone mega-IPO is a coin-flip-or-worse standalone top signal. The discriminator: cyclical-peak deals top-tick (BX/credit, Glencore/commodities — the IPO is the late-cycle monetization event); secular- growth platforms don't (GOOG/BABA/FB kept compounding). Crisis-window deals (Visa) can even be anti-signals.

3. The IPO-itself underperformance base rate is the clean one (Ritter)

Ritter (U. Florida): IPOs with >$100M sales AND price/sales >40 at offer → 12 of 14 (≈86%) underperformed the market over the next 3 years (bought at first-day close), 1980–2025. SpaceX at ~$1.75T / >90× revenue sits squarely in that bucket. This is a high-confidence statement about $SPCX itself (child-case), not the broad tape. NB: the "−20pp over 3y / −58pp for >40× rev" magnitudes the Economist/Fool cited were not verifiable in sourcing — use the 86% frequency, flag the magnitudes as unconfirmed.

Verdict + reasoning

Issuance surge is a yellow flag / risk-budget input, NOT a sell trigger — and for this lane its strongest, most defensible form is a top-tick for the AI-infra/high-beta complex, not the broad S&P.

• Broad-market regime-top: weak. Three concentrated deals ≠ the 1999/2021- style aggregate flood that powers the robust Baker-Wurgler prediction; lone- mega-IPO broad-top hit rate is ~1/7. Don't trade the broad index off this.
• AI-complex top-tick: credible (Glencore analog). SpaceX's business is secular, but the deal's pricing/bundling are cyclical-peak in character (>90× revenue; bundling xAI imports the most-crowded theme at its valuation peak). That makes it most like Glencore — a dominant franchise sold at its sector's valuation top — so the top-tick, if any, fires for the AI-infra / high-multiple-growth complex, with a multi-week-to-multi-month lead (BX led by ~4 months / +4%), not the whole market.
• $SPCX itself: likely underperforms (Ritter P/S>40 → 86%), high confidence — routes to the SpaceX child case, not parent-owned.

Escalation tell (the thing that would upgrade regime-top from yellow to red): the issuance surge broadens — a rising count of speculative IPOs, rising day-1 pops, and a rising aggregate equity share (the Baker-Wurgler wave). Three giga-deals are the leading edge test; watch whether a flood follows.

Follow-through

• Folded into the lane as the Regime-Top / Issuance-Surge hypothesis (thread Hypotheses D3 + classifier tag confirms-regime-top).
• New standing watch filed: 2026 IPO-issuance-broadening monitor (count of speculative listings + day-1 pops + equity share) as the regime-top escalation gate.
• $SPCX Ritter-underperformance base rate noted for the SpaceX child case (not parent-owned).

Sources

Price, RSI and trend figures read from the desk's pre-computed scan summaries (summaries). No number in this note was computed in prose.