$86 billion is about to get vacuumed out of this market in one print. Where it lands when it comes back is the trade — and it isn't aerospace.
Demand is not the problem. Building fast enough is.
Tomorrow, the biggest IPO ever soaks up as much as $86 billion of investors' cash in one day. To buy in, people sell other stocks first — that's this week's pressure. In two or three weeks the money flows back into the market, but it won't chase defense stocks that already doubled; it'll hunt for the cheap thing — software companies that actually charge money for AI. And SpaceX itself? Its early investors get to start cashing out in August. History says that rarely ends well for whoever bought the top.
$250B of demand chasing $75B of paper. Every allocated dollar got sold out of something else first.
| Buyer | Evidence | Read |
|---|---|---|
| Institutions | BlackRock reportedly ≥$5B order · $1B+ family-office bids · book 3.5–4x covered | Funds liquidating existing AI/megacap positions to fund allocations — the week's tape pressure |
| Retail | $70B+ in retail orders vs ~$22–26B available · Fidelity cut minimums $500K → $2,000 · Robinhood/Schwab/E*TRADE live | Maximum FOMO — Bloomberg-covered reports of people taking out loans to subscribe. Most get partial fills or nothing |
| Passive (forced) | Nasdaq-100 fast-entry after just 15 trading days · CRSP/Vanguard within weeks · est. 17–25% of float bought T+5 to T+15 | Index funds are the engineered exit bid — demand arrives before insider supply unlocks |
| Day-1 sellers | Directed-share program: up to 5% of offering (~$3.75B) to hand-picked employees/persons — no lockup at all | The only unrestricted supply on day one · caps the squeeze |
$75–86B is roughly 17% of one full day's total US equity volume concentrated into a single print — bigger than Aramco's record raise ($29.4B with greenshoe) and Alibaba's ($25B) combined, more than twice over. Strategists are already mapping the flows: Empower's Marta Norton said this week's AI-complex selling "could be making way for mega-IPOs like SpaceX"; Prof G's Galloway estimates SpaceX + OpenAI + Anthropic could demand ~$400B of fresh equity from this market. The counter (Allianz): the raise is <1% of US market cap and $8T sits in money funds — the cash exists. Both are right: the drain is real but transient. The money comes back. The question this report answers is where it goes when it does.
So many people want SpaceX shares that orders total more than three times what's for sale. To pay for them, buyers sold other stocks this week — that's part of why the market's been red. The selling stops once the IPO is funded, and within a few weeks that ocean of cash flows back into the market looking for the next thing to buy.
The supply is pre-scheduled. SPCX's most dangerous window is August through October — right after the VCs can finally sell.
This is not a standard 180-day lockup. SpaceX engineered a rolling release valve — and a performance trigger that accelerates insider supply if the stock rips. Mark the calendar:
| Precedent | At unlock |
|---|---|
| Facebook (2012) | Down 40%+ from IPO price by final lockup expiry (later recovered) |
| Uber (2019) | Hit its all-time low on the expiration date — down 40% from IPO |
| Rivian (2021) | −20% in one session when Ford disclosed it would sell at the 180-day mark |
| Palantir (2020) | −13% in a single day as Thiel and insiders sold tens of millions of shares into the retail premium |
| Snowflake (2020) | Used a staggered schedule like SpaceX — still dropped ~11% into the final expiration week |
The pattern, per Trefis: "the higher the speculative premium at IPO, the more painful the expiration." SPCX prices at ~95x trailing sales — the highest speculative premium of any mega-IPO ever. Retail allocation buyers face their own friction: no legal lockup, but Fidelity's anti-flip window is 15 days (selling earlier risks an SSN-tied lifetime ban from future IPOs; FINRA's flipping benchmark is 30 days — check your broker). Translation: retail is soft-locked through late June while insiders' calendar marches toward August.
SpaceX's early investors agreed to wait before selling — but not long. The first big batch (20% of everything insiders own) can hit the market two days after the company's first earnings report in August, with more unlocking every few weeks after. Every famous hyped IPO — Facebook, Uber, Rivian, Palantir — dropped hard when those gates opened. And here's the twist: if SpaceX's stock jumps 30% and stays there, an extra batch unlocks early. A big pop literally creates its own sellers. Elon himself can't sell for a full year — he's the only one truly locked in.
$135 deal · ~$75 fair value. You're paying for the rocket; the price only works if it's software.
Stack the deal price against what the business supports. Sum-of-the-parts on 2025 numbers (revenue $18.67B · net loss $4.94B · adj EBITDA $6.58B) lands at $58–92/share, midpoint ~$75:
The deal is priced 1.5–2.3x above any reasonable framework — and the mark-up is fresh: the Feb 2026 SpaceX–xAI merger valued the combined entity at $1.25T; the IPO asks $1.77T — +42% in four months with no operational catalyst. The only lens where $1.77T computes is treating it as a software company: Starlink is 61% of revenue ($11.39B) at 63% gross margin — genuine recurring-subscription economics — while launch is the capital-furnace loss leader and xAI burned $(6.4B) on $3.2B of revenue. The biggest aerospace IPO in history is being priced on its software segment. That's not a bug in this report's thesis — it's the proof of it.
With research coverage now live, forward numbers exist for the first time. Circulating Street models cluster around $29.7B revenue in 2026 and $47.8B in 2027 (+59% and +61%) — and the bull case goes much further: Goldman, the lead underwriter, models the AI segment alone at $15.6B in 2026 → $34.5B in 2027, building to a moonshot $474B revenue / $352B EBITDA by 2030. One caution on the numbers making the rounds: the "$4.8B 2025 EBITDA" figure is wrong — the S-1's audited 2025 adjusted EBITDA is $6.58B ($4.8B appears to be a garble of the $4.94B net loss).
| 2025 (S-1 audited) | 2026E | 2027E | |
|---|---|---|---|
| Revenue | $18.7B | ~$29.7B (+59%) | ~$47.8B (+61%) |
| Adj EBITDA | $6.58B | ~$13.2B est (unverified) | $14.5–22.2B (initiation spread) |
| $1.77T multiple | 95x rev | ~60x rev | ~37x rev · ~80–122x EBITDA |
That last row is the entire valuation debate in one line. Grant the insane growth — a clean double-double in revenue — and at $135 you are still paying ~37x 2027 sales and somewhere between 80x and 122x 2027 EBITDA. Seeking Alpha's June 9 initiation put it bluntly: "priced for 2032, not 2026" — underweight, $80 base target, $91 probability-weighted fair value, landing within dollars of this report's $75 SOTP midpoint. And the Q1 2026 actuals explain the skepticism: revenue $4.69B and adj EBITDA $1.13B against $10.1B of capex in the single quarter ($7.7B of it AI), debt up $24.7B → $29.1B, cash down to $11.4B — a ~$3B/quarter burn with a $41.3B accumulated deficit. The growth is real. The price already spent it.
| Day-1 close above | Implied move | Odds |
|---|---|---|
| $1.8T | ~+2% · positive close | 62% |
| $2.0T | ~+13% | 41% |
| $2.4T | ~+36% · pop trigger zone | 22% |
| $3.0T | ~+70% squeeze | 11% |
Real money handicaps a 62% chance of a green day one and ~41% odds of a double-digit pop — but note the 22% bucket: a +36% close puts SPCX deep into the $175.50 performance-trigger zone, where the early unlock fires. The market is pricing the pop and its own undoing simultaneously.
We're watching SPCX price discovery on Hyperliquid → xyz:SPCX. Not sponsored — it's simply the #1 venue trading IPOs right now, which means the deepest liquidity and the cleanest real-time price signal before and around the listing. Pre-market, after-hours, through the unlock dates: when in doubt, trust the price where the most money trades.
By the math, SpaceX shares are worth maybe $75. They're selling for $135. The only way $135 makes sense is if you value it like a software company — because the part that makes money, Starlink, really is one: ten million subscribers paying monthly. Betting markets expect a small first-day gain. But a huge first-day gain would actually backfire — it triggers early insider selling.
Aerospace already paid. Software hasn't. The rotation charts — this is the whole trade in three pictures.
| Signal | Number | Read |
|---|---|---|
| A&D ETF inflows | $8.2B in 3 quarters of 2025 · +573% YoY (FactSet/Kobeissi) | The money already arrived |
| New A&D ETF launches | 17 in 2025 vs 2 in 2024 | Product factories chase tops, not bottoms — the classic late-cycle tell |
| State Street 2026 outlook | "Defense was the standout theme of 2025" | When the biggest custodian calls it the theme, it's priced |
| Spending backdrop | Global defense $2.63T in 2025 · NATO Europe +20% YoY · every ally at 2% for the first time | Real — but known, owned, and in the multiple |
| Software, mirror image | Worst quarter since 2008 → 10/50-day golden cross Apr 23 → CNBC "mini bull market" flagged May 19 | Bottomed, turning, under-owned — the rotation already started quietly |
| Metric | Software (IGV-type) | A&D primes (ITA-type) |
|---|---|---|
| Gross margins | 70–80% | ~20–30% |
| Operating margins | ~25–35% at scale | ~8–13% |
| Revenue growth | Low-to-mid teens, AI-monetization inflecting | Mid-to-high single digits, backlog-gated |
| Capital intensity | Minimal — code ships free | Heavy — factories, primes, supply chains |
| Valuation vs own history | Compressed — multi-year lows after the 2026 reset | Premium — priced for a "new super-cycle" |
| Positioning | Under-owned · outflow fatigue | Crowded · +573% inflows |
Software grows faster, at triple the margins, with no factories — and it's the side trading at a discount to its own history while A&D trades at a premium to its own. The returns gap isn't a verdict on business quality; it's a positioning artifact. Positioning artifacts mean-revert when the marginal dollar moves — and the marginal $86B moves this week.
Defense stocks already had their huge run — up 50–70% in a year, with record money flooding in and 17 brand-new defense funds launched to catch the hype. Software companies — better profit margins, faster growth — went down over the same stretch and just hit their cheapest levels in years. When the SpaceX money comes back into the market, it hunts for what's cheap and ignored, not what already doubled. That's software.
The 2–3 week round trip. Liquidity leaves June 11–12 · comes back by early July · lands somewhere new.
Mega-IPO digestion follows a known arc: forced selling into the print, a vacuum week, then the cash recycles as allocations settle and index demand arrives. The calendar stacks three liquidity-return events inside three weeks:
| Window | Event | Flow effect |
|---|---|---|
| Jun 11–12 | Pricing tonight · debut tomorrow | Peak drain. Final allocation funding · AI-complex selling crescendos · the worst tape of the cycle |
| Jun 13–17 | Digestion · FOMC Jun 16–17 (Warsh's first meeting) | Vacuum week — unallocated retail cash ($40B+ of unfilled orders) refunds back to brokerage accounts looking for a home |
| Jun 26–27 | Russell rebalance (CRWV, IREN et al. enter Russell 3000) | Mechanical passive bid across the market · first organized re-deployment of cash |
| ~Jul 2–6 | SPCX Nasdaq-100 fast-entry (~15 trading days) · CRSP/Vanguard adds | Index funds finish buying SPCX (est. 17–25% of float) — the drain officially ends · rotation window opens |
| Mid-Jul | Q2 earnings season begins — software reports | The catalyst that converts the rotation from flow story to fundamental story |
| ~Mid-Aug | SPCX first earnings + 2 days → 20% insider unlock | SpaceX's supply era begins · capital that chased the halo trade gets a reason to leave |
The historical base rate backs the arc: after Alibaba's then-record 2014 IPO drained $25B, the S&P digested within weeks and leadership rotated away from the IPO-adjacent momentum complex. After Aramco (2019), energy — the IPO's own sector — lagged global equities for the following two quarters. The pattern: the record IPO marks the sentiment top for its own sector, and the recycled liquidity leads somewhere else. Space pure-plays carry a second, unique problem: for a decade, "you can't buy SpaceX" was the entire bull case for buying RKLB, ASTS, LUNR and PL as proxies. As of tomorrow, you can just buy SpaceX. The proxy premium dies with the IPO bell.
Think of the market like a pool. Tomorrow, SpaceX drains $86 billion out of it. Over the next three weeks the water flows back in — refunded orders, index funds finishing their buying, a major index reshuffle on June 26. By early July the pool is full again. History says the refill doesn't go back where it came from — and the small space stocks people bought as "SpaceX substitutes" just lost their whole reason to exist.
The AI-monetization layer. Where the recycled dollar compounds.
"Software" here means one specific thing: the companies that charge money for AI — not the ones buying GPUs. Three years of capex flowed into compute (the hardware supercycle: semis, neoclouds, power). The monetization layer that sits on top — PLTR, NOW, CRM, MSFT — collects recurring, 70-80%-gross-margin revenue from that infrastructure, and it's the side that lagged into multi-year-low relative valuations. The bridge is already visible: Palantir sits as a top-five holding inside SHLD, a defense ETF — the market quietly pays defense-money for software economics where it can find both. The rotation just makes that trade explicit.
| Name | The AI-monetization engine | Position |
|---|---|---|
| PLTR · Palantir | AIP — the defense/software crossover · already owned by defense ETFs · the rotation's natural first stop | Watchlist |
| NOW · ServiceNow | Now Assist / agentic workflow monetization across the enterprise install base · RPO compounding | Held — common + Jan $100 calls |
| CRM · Salesforce | Agentforce — per-conversation AI pricing on the largest CRM footprint on earth | Watchlist |
| MSFT · Microsoft | Copilot seats + Azure AI — the toll booth on both layers at once | Watchlist |
For three years, the AI money went to the companies building the machines. The next leg goes to the companies charging rent on what the machines do — software firms billing monthly for AI assistants and agents. They're cheaper than they've been in years, and they're the natural landing spot for the SpaceX cash when it returns. I already own ServiceNow.
The case for the rotation — and what breaks it.
▲ For The Rotation
- $86B drains and returns inside 3 weeks — refunded retail orders + Russell rebalance Jun 26–27 + SPCX index buying done by ~Jul 3.
- A&D is maximally crowded: +573% ETF inflows · 17 new funds in a year · State Street's "standout theme" — the marginal buyer is already in.
- Software is maximally washed out: worst quarter since 2008 → golden cross Apr 23 → CNBC-flagged mini bull market since mid-May. The turn predates the IPO.
- Fundamental gap favors software: 70–80% gross margins vs 20–30% · triple the operating margin · no factories · valuation at a discount to history vs A&D at a premium.
- Space proxies lose their reason to exist — "can't buy SpaceX" dies June 12 · RKLB +171% in 2025 on borrowed halo.
- Mega-IPO base rate: the record IPO marks its own sector's sentiment top (Aramco → energy lagged 2 quarters; BABA → momentum rotated).
- SPCX's own prospectus is the tell — 61% Starlink subscription revenue at 63% margin · priced at 95x sales · even SpaceX is valued as software.
- Q2 software earnings mid-July land exactly as the rotation window opens.
▼ What Breaks It
- FOMC Jun 16–17: Warsh's first meeting with Dec hike odds ~70% — software is the most rate-sensitive sector on the board. A hawkish shock delays the rotation regardless of flows.
- Geopolitics doesn't care about positioning — Iran war headlines, NATO escalation, a new conflict re-fuels the A&D bid instantly.
- The AI-compute cannibalization is real: Q1's software wreck came from IT budgets diverting to GPU spend — if that persists, software's earnings disappoint in July and the rotation dies on contact.
- SPCX could defy gravity: 5% float + forced index buying + retail soft-lock = a squeeze that keeps the aerospace halo alive for months (see: day-1 odds, 22% chance of +36%).
- Crowded can stay crowded — defense outperformed for two straight years while everyone called it crowded.
- Money funds at $8T — the IPO may be funded from cash, not stock sales, muting the whole drain-and-return mechanism (the Allianz view).
The rotation case: the money is leaving the expensive crowded thing and the cheap hated thing already started turning before the IPO. The risk case: the Fed meets in four days and could spook the exact stocks this trade buys — and world events can always re-light defense stocks. Manage it with timing, not faith.
Sector report card. Software 8.7 · Semiconductors 7.2 · Aerospace & Defense (incl. space) 5.6.
Six categories, three sectors, graded on the next 6–12 months from June 11, 2026. Space pure-plays are folded into A&D where they belong — same flows, same halo. Semiconductors added as the third lane: the hardware layer of the AI trade that already paid. Grades are mine; inputs are everything above.
| Category | Software (AI-monetization) | Semiconductors | A&D + Space |
|---|---|---|---|
| Revenue growth | A− · teens, AI inflecting | A · AVGO AI +143% · still the steepest line | B− · primes single digits · space fast but tiny |
| Margin / FCF quality | A+ · 70–80% GM, capital-light | A− · 50–75% GM at the leaders, capex rising | C · primes 8–13% op margins · space mostly pre-profit |
| Valuation vs history | A · multi-year relative lows post worst-Q-since-2008 | C+ · June 5 reset helped, still elevated vs history | D+ · "super-cycle" premium + halo-inflated story multiples |
| Momentum / trend | B+ · golden cross Apr 23, mini bull underway | C+ · SOX −10% in the crash · choppy repair | B− · still trending but extended · space peaked with IPO hype |
| Positioning / crowding | A · under-owned, outflow fatigue spent | C− · the most-owned trade on earth, partially flushed Jun 5 | D · +573% inflows, 17 new ETFs · proxy premium dies Jun 12 |
| Catalyst path | A− · rotation window + Q2 earnings mid-July | B− · demand intact (AVGO record) but priced · rotation headwind | C− · budgets known and priced · faces SPCX unlock supply Aug–Oct |
The ordering is the thesis. Semis sit in the middle on purpose: the demand is real (Broadcom just printed a record AI quarter), the businesses are spectacular — but it's the most-owned trade in the market, the June 5 crash showed how crowded the exits are, and the marginal dollar now gets more for its money one layer up the stack. Hold the leaders you own; the fresh capital goes to software. A&D plus its space satellites grade last on the only test that matters: a premium multiple, record crowding, a +52–69% year already banked — and as of tomorrow, the halo names lose their whole reason to exist. You can just buy SpaceX now.
Report card, three lanes: software gets the A — best margins, cheapest prices, money about to flow its way. Chip stocks get the B — amazing companies, but everyone already owns them and they already had their giant run. Defense and space stocks get the D — expensive, crowded, and the small space names just lost their main selling point: being the only way to bet on SpaceX.
John's read. I'm skipping the IPO — and positioned for what comes after.
- I'm skipping the IPO. Requesting shares and flipping them a few days in is a legitimate play if you want it — the day-1 odds favor a small pop. But know the price tag: selling inside your broker's flip window (Fidelity 15 days, FINRA's benchmark is 30) can get you banned from future IPO allocations — at some brokers permanently.
- Why I'm sitting this one out: I'm saving my account. There are other listings on my radar in the coming months, and burning my IPO eligibility to scalp a few hundred dollars on SPCX is terrible expected value. One flip can cost every future allocation. The lottery ticket isn't free — it's priced in access.
- SPCX fades when the VCs can sell. 20% of insider holdings free two days after the August print, supply waves through October, and a pop above $175.50 accelerates it. Every hyped-IPO precedent — Facebook, Uber, Rivian, Palantir — says the same thing.
- I'm already positioned for the other side. Software and the AI names I hold — NOW (common + calls), with PLTR/CRM/MSFT on the rotation watchlist — betting the money bleeds out of aerospace over the coming months as the halo trade loses its sponsor. Liquidity is back by early July (Russell Jun 26–27, index fast-entry done); I want to be sitting where it lands, not chasing it there.
- The kicker I keep coming back to: SpaceX priced itself at 95x sales because Starlink is software. The biggest aerospace listing ever just endorsed the rotation thesis in its own prospectus.
Want to see what price I bought in at — or the other stocks I'm in? Click the button below to join the Discord.
My exact entry, targets and stop. Locked — they live in the Discord.
This is my actual rotation plan — the software names I'm buying, entries, sizing, and the SPCX day-1 tree. It's blank here on purpose.
Want my exact rotation buy list — names, entries, sizing?
Join the Discord to find out! →SpaceX IPO prospectus coverage — Morningstar (tiered lockup structure), Trefis, Motley Fool, Yahoo Finance (Jun 2026) · CNBC IPO terms (Jun 3) · X/Twitter retail-demand scrape Jun 11 (oversubscription, BlackRock order, retail $70B, broker minimums) · Fidelity flip-window reporting (Yahoo Finance) · Polymarket day-1 markets ($905K vol) · IGV performance — FinanceCharts/Yahoo (YTD −11%, TTM −8.5%, 2025 +5.6%) · ITA/XAR returns — 24/7 Wall St (Jan-25→May-26 +49%; 1Y +52%/+69%), State Street 2026 Global ETF Outlook · A&D flows — FactSet via Kobeissi (+573%, $8.2B) · Kiplinger/Zacks (17 ETF launches, SHLD/PLTR) · CNBC software mini-bull coverage (May 19) · NFRS Vol. 36 SpaceX SOTP framework · S-1 segment data via Yahoo Finance / Gate (Q1 2026 actuals, capex, deficit) · Goldman Sachs initiation via 24/7 Wall St (Jun 4) · Seeking Alpha initiation (Jun 9, underweight $80) · IPO lifecycle chart: NFRS illustration. Prices as reported by sources, Jun 2–11, 2026.