Why This Exists

This is the parent market-structure lane for the 2026 AI mega-IPO cycle.

The user compared it to a Strait of Hormuz closure: not because the mechanism is physical, but because it can become a broad attribution lens for market tape. The question is whether compressed index-entry timing, passive-product buying, float/lockup mechanics, and institutional allocation budgets change the price behavior of stocks that look unrelated at first glance.

This is not primarily a SpaceX lane. $SPCX, OpenAI, and Anthropic are child cases.

Source-Truth Correction

"401(k) rule" is the public-facing shorthand. The actual source mechanism is index methodology and passive/index-product plumbing.

• Nasdaq Fast Entry: primary-source real. Fast-entry candidates ranked top 40 and meeting applicable criteria may be added to the Nasdaq-100 after 15 trading days, subject to the 3x float cap where applicable.
• S&P 500: not automatic. S&P DJI's consultation proposed MegaCap exceptions and a six-month IPO-seasoning period, but S&P 500 additions remain Index Committee discretionary.
• S&P TMI / Completion / Dow Jones U.S. TSM: more immediate fast-track watch if the MegaCap proposal is adopted.

The lane should never say "401(k)s automatically buy the IPO." It should say: index methodology may create forced-flow windows that retirement/index products participate in.

Ownership Rules

• This parent lane owns: Fast Entry, index methodology, passive bid, allocation crowding, float/lockup mechanics, broad market oxygen, and tape attribution.
• SpaceX child track owns: $SPCX, xAI/SpaceX economics, supplier/halo basket, SATS proxy, SpaceX-specific lockups and T-dates.
• Anthropic child track owns: Claude demand, hyperscaler circular financing, compute book, AWS/GOOG/MSFT dependency, safety/IP, Anthropic IPO cadence.
• OpenAI child track owns: OpenAI filing/unit economics once an S-1 or equivalent primary disclosure lands.

Cross-link is not ownership. A shared contract, shared IPO window, or shared index-flow mechanism is evidence for both sides, not a merge point.

Hypotheses

Organized by transmission dimension. This is a broad attribution menu, not a ranked forecast — several can be live at once; each desk pass picks which the tape actually supports (per feedback_dont_false_rank_competing_hypotheses). Expanded 2026-06-02 from 5 flat buckets to a 5-dimension set after the user prompted a zoom-out; (•) = original, (★) = added.

Dimension 1 — Flow (mechanical price pressure)

Liquidity Drain (•). Mega-IPO allocations + passive flows pull capital from crowded winners ($NVDA, Mag7, neoclouds $NBIS/$CRWV/$CORZ/$IREN/$APLD, plus leveraged/ETF vehicles on the same cohort). The near-term channel is the book-build "funding pocket" — most-liquid names sold to raise allocation cash (moomoo). Tape: AI-infra/Mag7 weaken while IPO-linked names or liquidity proxies strengthen.

Passive-Flow Bid (•). Forced index buyers arrive after listing → the IPO name / proxies strengthen into entry dates. Tape: abnormal volume/price near FTSE Russell T+5 or Nasdaq T+15.

Index-Arb Front-Run & Fade (★). Arbs buy ahead of the dated, known inclusions and sell into the forced buyers; the passive bid is pre-empted. Tape: ramp/volume BEFORE the entry date, reversal AFTER — opposite timing of a clean passive-flow bid.

Lockup-Supply Overhang (★). The staggered release ladder (Aug/Sep onward) creates persistent supply that caps $SPCX and bleeds the cohort for quarters — distinct from the one-shot drain. Tape: stepwise underperformance on the dated release tranches.

Dealer-Gamma / Vol Event (★). The listing is a volatility-regime shift, not a directional one; options dealer hedging dominates the flow. Tape: single-name + index IV / term-structure spike around T-1/T0, mechanical hedging.

Dimension 2 — Fundamental (demand & disclosure)

Comp Repricing (•). The mega-IPO becomes a new public comp for AI-infra names. Tape: neocloud / AI-power names move with the IPO name after listing.

Capital-Formation Accelerant — bull (★). The ~$200B raised funds compute → the demand side of our bottleneck theses (ai-power-bottleneck, optical-supercycle, memory-supercycle, cpu-shortage-supply-chain) hardens on FUNDAMENTALS. The S-1s are the first audited demand-side disclosure those theses have ever had. Tape: bottleneck baskets re-rate up independent of $SPCX flow; S-1 capex disclosures size the buildout.

Unit-Economics / Burn Reveal — bear (★). The S-1s expose cost-per-token / gross margin / cash burn; if deeply negative, the whole AI complex de-rates on fundamentals. This is the "bust fuse" — and it's the same disclosure that adjudicates the bull Accelerant case. Tape: AI complex sells off on S-1 disclosure days, led by the highest-multiple names.

Circular-Financing Exposure (★). S-1s disclose the hyperscaler↔lab compute web (Anthropic's AWS/GOOG/MSFT book, the $15B SpaceX/xAI contract) → quality of $MSFT/$GOOGL/$AMZN "AI revenue" questioned. Tape: hyperscalers de-rate on disclosure; cross-links ai-counterparty-taxonomy.

Dimension 3 — Behavioral / regime (sentiment & cycle)

Halo Bid (•). The IPO validates the category; space/AI adjacencies run while the rest of the market absorbs it. Tape: $RKLB/$FLY/$ASTS/$LUNR/$RDW/ $PL/$IRDM/$SATS run without broad AI-infra weakness.

Sell-the-News Fade (★). The counter to Halo: runup into the IPO, then sell-the-news on listing. Tape: halo basket peaks ~T-1 and fades post-listing.

Regime-Top / Issuance-Surge Bell (★). The giga-IPO wave at ATHs marks the AI/risk-asset cycle top. Base-rate read (investigation 2026-06-02-ai-mega-ipo-issuance-surge-base-rate): a yellow flag, not a sell trigger — strongest as a top-tick for the AI-infra/high-beta complex (Glencore analog: dominant franchise sold at its sector's valuation peak), weak for the broad S&P (lone-mega-IPO broad-top hit rate ~1/7). Escalation tell: issuance broadens — rising count of speculative IPOs, day-1 pops, aggregate equity share (the robust Baker-Wurgler wave signal). Tape: AI complex rolls over post-listing while the issuance count climbs.

Retail Distribution Top (★). The unusually large retail carve-out = smart-money distributing to retail at the top. Tape: heavy retail allocation + post-IPO underperformance + insiders selling into retail at lockup releases.

Narrative Monopoly / Attention Crowding-Out (★). The giga-IPO monopolizes attention AND flows, starving non-AI themes. Tape: non-AI sectors/themes bleed flows and underperform into the event regardless of their own fundamentals.

Dimension 4 — Reflexive / structural (multi-quarter)

Reflexive Funding Loop (★). High stock → labs raise equity → fund compute → suppliers' revenue up → AI complex up → higher stock (Soros reflexivity); self-reinforcing until it snaps. Tape: tight $SPCX/labs ↔ supplier-basket correlation; intact = melt-up, break = cascade.

Shrinking-Float Reversal / Capital Diet (★). Equity supply expands after years of buyback-driven scarcity → multi-year market-wide multiple compression (Economist / Haghani). Tape: breadth and multiples compress over quarters as issuance ramps and buybacks fall.

Duration Creep (★). Adding >90× revenue mega-weights raises the index's rate-sensitivity → the broad market becomes more correlated to long rates. Tape: rising $SPY/$QQQ ↔ $TLT rate-beta after inclusion.

VC/Private→Public Rotation (★). Direct public access to SpaceX/OpenAI/Anthropic drains the "only-way-to-play-AI" scarcity premium from private/VC and from public AI-proxy names. Tape: scarcity-premium proxies (some neoclouds, miners-as-AI) de-rate as direct access arrives.

Dimension 5 — Counter / null

Nothingburger (•). The market absorbs the event; only direct child-case names move. Tape: $SPY/$QQQ/Mag7/AI-infra/defensives show no repeatable event-window behavior. (Economist's near-term base case: ~0.1% S&P / ~0.5% NDX initial weight on the ~4% float.)

Dispersion not Direction (★). The event raises intra-complex dispersion (relative repricing of compute economics) without moving the index. Tape: rising single-name dispersion / correlation breakdown within the AI complex; pairs diverge.

Dated Gates

Scan Signature

Every desk pass that invokes this lane should pick a primary signature (and any secondaries), grouped by dimension:

• Flow: confirms-liquidity-drain, confirms-passive-flow-bid, confirms-front-run-fade, confirms-lockup-overhang, confirms-vol-event
• Fundamental: confirms-comp-repricing, confirms-capital-accelerant, confirms-burn-reveal, confirms-circular-financing
• Behavioral/regime: confirms-halo, confirms-sell-the-news, confirms-regime-top, confirms-retail-distribution, confirms-narrative-monopoly
• Reflexive/structural: confirms-reflexive-loop, confirms-capital-diet, confirms-duration-creep, confirms-private-public-rotation
• Null/counter: confirms-nothingburger, confirms-dispersion, contradicts-lane, unrelated-or-insufficient

Required basket read:

• AI / Mag7: $NVDA, $MSFT, $GOOGL, $AMZN, $META
• Neocloud / AI-infra: $NBIS, $CRWV, $CORZ, $IREN, $APLD
• Space halo: $RKLB, $FLY, $ASTS, $LUNR, $RDW, $PL, $IRDM, $SATS
• Broad oxygen: $SPY, $QQQ, $IWM, $TLT

Graduation Criteria

Promote this thread to a perspective only if one of these fires:

1. $SPCX explains cross-market behavior beyond the space basket.
2. Nasdaq/S&P methodology produces observable forced-flow effects.
3. OpenAI or Anthropic files close enough to $SPCX that the market treats the IPOs as one allocation regime.
4. Three or more desk passes need this classifier to make sense of tape.

If none fire, keep this as a useful thread and let the child tracks remain separate.

Trading Posture

The standing real-book stance for every name in this lane (SpaceX live; Anthropic/OpenAI pending). Crystallized 2026-06-17 from the SPCX post-IPO tape.

• Good company ≠ good trade. These can be real, large, fast-growing businesses (Starlink; Claude/ChatGPT revenue) and untradeable at the IPO price simultaneously. "Is it a good company" and "is this a good entry" are different questions; here they answer opposite ways. Real fundamentals are the trap's camouflage — they make the bull case feel solid while the quote is unmoored from them.
• The low-float pop is a trap, not an entry. A +20–40% post-IPO run on a ~4% float (SPCX: +41%) is the easiest move to manufacture and the easiest to unwind — a supply-starved overshoot, not a trend. Buying it is textbook buy-pre-confirmation: no base, no flag, no breakout from structure (a 4-session IPO has no uptrend to "dip" into). Worse for price-trigger names — strength pulls lockup supply forward, so chasing the pop accelerates the reversal you'd be buying into. The FOMO is the trap.
• The supported move is to wait — and the wait is the setup. The lockup overhang is the patient buyer's friend: it drags price to where the float is real and a base can form. Then, if it bottoms → bases → breaks out, it's a genuine trend-hold entry on a real company at a sane-er price — months out, post-unload, which for a 6–48mo holder is the only entry that ever mattered.
• Labs = double-patience. SpaceX has a dated supply cliff (the lockup ladder). Anthropic/OpenAI won't until their S-1s — so it's wait-to-be-public, then wait-for-the-post-lockup-reset. Same script, no dates yet.
• Scope: this is the trend-hold real-book posture (boxes/trend-hold). The pop is still a legitimate momentum vehicle for the active-trading personas (claude-momentum/yolo/swing-kid/degen) — their box, their rules, not this stance. Routing a name to "trade the pop" is an active-box decision, never a real-book one.

Desk-Pass Log

• 2026-06-02 — Desk pass #1. Signature: unrelated-or-insufficient on the broad liquidity-drain leg, early confirms-halo on the 30d/3m space horizon (selective, cooling at the 7d edge). Passive-flow / comp-repricing untestable pre-listing. Methodology corrections: S&P decision still pending (06-08 conditional); Nasdaq T+15 confirmed live; CRSP T+5 added as a new earlier gate; SpaceX S-1/A offering blank (press valuations provisional); lockup staggered. No graduation criteria fired — keep as thread + scan lens. Full artifact: 2026-06-02-ai-mega-ipo-lane-desk-pass-methodology-and-pre-event-tape.
• 2026-06-02 — Economist + HN synthesis (same desk pass, user-supplied sources). Economist "Giga-IPOs" (06-01) + HN 48364055 confirm the methodology corrections and adjudicate the Dunn-vs-Friedman tension: near-term float-math (initial weights tiny on ~4% float — S&P ≈ 0.1%, NDX ≈ 0.5% with 3× cap, Russell 1000 ≈ 0.1%), long-term slow "capital diet" as lockups expire and buybacks slow. Gate correction: the T+5 fast-track is FTSE Russell (double-sourced), not CRSP/VTI. Concentration risk: AI ≈ 40% of S&P 500. Classification unchanged; regime framing hardened; no graduation. Captures: 2026-06-01-economist-giga-ipos-can-stockmarket-swallow-anthropic-spacex-openai, 2026-06-02-hn-giga-ipos-index-inclusion-discussion-48364055.
• 2026-06-02 — Hypothesis expansion + base-rate studies + S-1 crosswalk. Hypotheses restructured 5→19 across 5 dimensions; classifier expanded. Two base-rate studies filed: issuance-surge (regime-top = yellow flag, strongest for the AI-infra complex, Glencore analog) and the positive base rate (2026-06-02-large-ipo-winners-positive-base-rate) — winners cluster in the reasonable-multiple / profitable / durable-moat / not-at-peak bucket; no canonical winner priced >20–40× sales at a high and compounded a decade (Snowflake the falsifier). $SPCX at ~94–107× sales / net loss / ATH timing falls in the 86% loser bucket on all four factors (child-case read, not lane-owned). S-1 crosswalk playbook built + validated twice (Cerebras, then xAI 2026-06-02-xai-s1-segment-crosswalk-read): xAI confirms ai-power-bottleneck from the demand side (1 GW, ~320k GPUs, gas turbines), and the "$15B SpaceX/xAI contract" is corrected to SpaceX→Anthropic ($1.25B/mo, ~325k GPUs, primary-sourced).
• 2026-06-03 — Net-flow model (the lane's central tension, quantified + dated). 2026-06-03-spacex-net-flow-model-passive-demand-vs-lockup-supply. Parametric (offering still blank). Early index-inclusion forced bid ≈ $4–30B is an order of magnitude too small to absorb the staggered lockup supply ≈ $150–350B (s=15–50%). Net flips from a modest confirms-passive-flow-bid (Jun–Jul, T+5/T+15) to a large confirms-lockup-overhang (Aug–Oct, at the Q2 then Q3 earnings releases). Dated form of the Economist "near-term bid, medium-term capital diet." Caveats: insiders may hold (Musk 366d, no early release); +10% tranche is reflexive (strength brings supply); rerun when the deal prices.
• 2026-06-03 — Macro-buffer check (governs the broad-market leg). 2026-06-03-ai-mega-ipo-macro-buffer-check-reserves-rrp. Verified moomoo's claim vs Fed H.4.1: right on plumbing, overstated on conclusion. The RRP first-line buffer IS gone (~$2.5T → $1.5B), so drains now hit reserves directly — BUT reserves ($3.07T) are still "merely ample," QT ended 2025-12-01, and the Fed is armed (RMPs/SRF). Verdict: broad-market base case = confirms-nothingburger (a $75B IPO ≈ 2.4% of reserves, mostly a reallocation), with a fatter left tail if the cash sink hits a quarter-end / tax / TGA-rebuild bulge. Composes with the net-flow model: SPCX-specific lockup overhang is high-confidence; broad contagion is a conditional tail, not a base case. Watch SOFR–IORB + SRF around the IPO and the Q2 lockup.
• 2026-06-04 — Desk pass #2 (T-8 pre-position tape + gate resolution + priced deal). 2026-06-04-ai-mega-ipo-lane-desk-pass-2-pre-position-tape. Signature: primary confirms-sell-the-news (space halo +60–116% 3m but −11% to −24% over the last 7d — desk pass #1's "cooling at the 7d edge" has deepened into an active sell-the-rumor unwind into the print); secondaries confirms-nothingburger (broad tape at ATH/overbought — SPY RSI 68, QQQ RSI 77↑; no funding-pocket Mag7 drain — NVDA +1% 7d) and confirms-dispersion (neocloud stays hot — NBIS +157% 3m, CORZ RSI 73↑ — so the cooling is space-specific, basket not one bloc → cuts against graduation #1). Gate resolution: 06-08 S&P decision still PENDING as of 06-04 (comments closed 5/28, no adoption, roadshow ~6/4) — carry the gate; T-7 (~06-05) pre-position reads as a halo unwind, not accumulation (consistent with net-flow: the anticipatory bid was the trade). Priced-deal fold-in: offering no longer blank — $135 / 555,555,555 sh = $75.0B + $11.25B greenshoe, trades ~06-12 (2026-06-03-spacex-s-1a-priced-135-per-share-75b-offering); lands on the net-flow base case (graduates parametric → deal-confirmed; verdict unchanged). No graduation criterion fired — keep as thread + scan lens.
• 2026-06-17 — Desk pass #3 (first POST-listing tape read; the drain-vs-halo tell). 2026-06-17-ai-mega-ipo-lane-desk-pass-3-post-ipo-drain-vs-halo-tell (2026-06-17 ~T+3 intraday snapshot). Both headline legs tested null post-listing. Primary confirms-sell-the-news: the space halo basket bled through the listing (RDW −10%, LUNR −9.7%, ASTS −7.4%, FLY −5.1%, IRDM −3%, SATS/RKLB ~−1.4%) while SPCX itself ran +41% from the $135 IPO — desk pass #2's sell-the-rumor unwind continued past the print, the anticipatory bid was the trade. Secondaries confirms-nothingburger on the drain leg (Mag7 flat-to-up: NVDA +2.7%, AAPL +1.3%, MSFT −2.3% — no funding-pocket drain, exactly the macro-buffer base case) + confirms-dispersion (AI-infra/power ran independently — ARM +24%, MRVL +15%, APLD +8%, IREN +7%) + a new confirms-narrative-monopoly read (the only-way-to-play scarcity premium concentrated into the single liquid vehicle, starving adjacencies; the halo went inward, not outward — ARKX's +2.9% is mechanical SPCX exposure, not a comp re-rate). Graduation: #1 explicitly does NOT fire (SPCX moved neither the AI cohort nor the space basket — explains less cross-market behavior than hypothesized); #4 is numerically reached (3rd desk pass) but the evidence argues keep-as-thread — all three passes find the cross-market linkage thin, so the classifier has earned its keep as a falsifier, not a tradeable regime. Flagged for reviewer as a decision; not auto-promoting. Trade read unchanged: no halo trade in the comps, no drain to fade, the dated lockup overhang stays the only high-confidence SPCX signal. Space-comp weakness is sector-specific, NOT SpaceX contagion. Next: the dated passive-flow gates (FTSE Russell ~T+5 06-18, Nasdaq-100 Fast Entry ~T+15 07-03) test forced index demand — a separate flow question.

Cross-Links

• Parent-lane investigation: 2026-06-02-ai-mega-ipo-fast-entry-liquidity-regime-parent-lane
• Desk pass #1 (methodology + pre-event tape): 2026-06-02-ai-mega-ipo-lane-desk-pass-methodology-and-pre-event-tape
• SpaceX child track: 2026-05-26-spacex-ipo-liquidity-event
• Anthropic child track: 2026-05-27-anthropic-ipo-compute-financing
• OpenAI / Anthropic S-1 trigger: TASKS-RESEARCH "Anthropic + OpenAI S-1 filing watch"
• S-1 fundamental-transmission instrument (cross-cutting, validated vs Cerebras): 2026-06-02-ai-lab-s1-thesis-crosswalk-playbook — turns each lab's public S-1 into thesis-routed datapoints; Hypotheses D2 (Capital-Formation Accelerant / Burn Reveal) are its lane-side outputs.
• Evidence captures (desk pass #1): Dunn/AlgorithmicFIRE forced-selling model (2026-05-18), SpotGamma per-index flow (2026-05-27), Friedman/InvestmentNews float-math nothingburger (2026-05-28), moomoo book-build liquidity-shock (2026-06-02), Anthropic confidential S-1 (2026-06-01), SpaceX S-1/A offering-blank + staggered-lockup (2026-06-01).