The Pre-IPO Market, On-Chain

A market overview and product comparison of crypto's three pre-IPO rails — spot tokenization, perpetual futures, and prediction markets — for pricing the world's largest private companies before they list.

Published2026-06-10

Data as of2026-06-10

AuthorCoinMarketCap Research

Download PDF

$135

SpaceX fixed IPO price — $1.75–1.8T, June 12 Nasdaq debut; books 3.5–4× oversubscribed

$2.94B

cumulative pre-IPO perp volume across 10 venues

product categories — spot tokenization · perpetual futures · prediction markets

$7.9M

prediction-market open interest — an all-time high, 75% of it SpaceX

Chapter01

The three routes

Overview — three ways in

Spot, perpetual, prediction — legal nature · price discovery · regulatory exposure all differ

Before the detail, the map. "Trading SpaceX before it lists" is not one product — it is three distinct routes, each answering a different question: do you want to own a claim, take a leveraged view, or bet on an event? They share a headline and almost nothing else.

	Spot	Perpetuals	Prediction
You own	Claim on equity (SPV / note / native / broker)	Synthetic position; no shares	Binary / threshold event
Price discovery	Strike + AMM / matching / note secondary	Mark + funding convergence	Resolution probability
Key risk	Issuer "void"; SPV nesting	No clear US synthetic exemption	Discrete events only
Representatives	Jarsy · PreStocks · Republic · Robinhood EU	Binance · OKX · Coinbase · Kraken · TradeXYZ · Ventuals · Helix	Polymarket · Kalshi

The three routes at a glance

Route 1 · Own a claim

Spot

A token that represents a claim on the underlying — real shares (via SPV), a contingent note, an issuer-minted security, or a broker contract.

Best for: Durable, ownership-like exposure with no funding cost
Payoff: Linear; tracks the company 1:1 in principle
Leverage: None (you post the full value)
Time horizon: Hold to IPO / note maturity
Core hazard: Legal wrapper fragility — the May "void" statements

Verdict: the only route that can give true ownership — but also the only one exposed to the issuer voiding the transfer.

Route 2 · Take a view

Perpetual

A cash-settled synthetic that tracks an implied valuation via a mark price and a funding rate. No claim on the company at all.

Best for: Traders wanting leverage, 24/7 liquidity, easy shorts
Payoff: Linear & leveraged; both directions
Leverage: Up to ~5× on most venues
Time horizon: Any — no expiry; funding paid continuously
Core hazard: Funding tail + opaque marks; convergence only at IPO Rebase

Verdict: where the volume is. Deepest liquidity and the cleanest way to short — but you never own anything and bear the mark/funding mechanics.

Route 3 · Bet an event

Prediction

A contract that resolves on a discrete outcome — "does SpaceX list before date X," "does valuation cross $Y" — priced as a probability between 0 and 1.

Best for: Expressing a clean yes/no on timing or a threshold
Payoff: Binary; $0 or $1 at resolution
Leverage: Implicit (cheap long-shot odds), no liquidation
Time horizon: To the event's resolution date
Core hazard: No continuous-valuation exposure; resolution-source risk

Verdict: the cleanest legally — no equity, no funding — but it can only answer discrete questions, never "how much is it worth right now."

Compared on the axes that actually decide outcomes

Dimension	Spot	Perpetual	Prediction
Ownership / rights	Claim (varies by wrapper)	None — synthetic	None — wager
Leverage	1× only	Up to ~20×, liquidation risk	Implicit via odds; no liquidation
Carry cost	None	Funding (Ventuals ~40–56% median, 415% peak)	None (pay the spread once)
Can you short?	No	Yes, natively	Yes (buy the "No")
Liquidity depth	Thin, fragmented pools	Deepest (CEX books)	Moderate; event-specific
Price-discovery quality	Lags (monthly marks)	Real-time but anchor-free	Probability, not valuation
Headline failure mode	Issuer "void"; FTX-style custody	Funding blow-up; Mirror-style oracle	Bad/ambiguous resolution
US-accessible?	Accredited (Reg D) or not at all	No (offshore; not for US persons)	Kalshi yes (CFTC DCM); Polymarket limited
Regulatory path	Securities (fragile for SPV)	No clear US synthetic exemption	Event-contract / CFTC

Critical read — what each route is really selling

Spot's promise and its catch. Only spot can plausibly deliver ownership — but the wrapper decides everything. A custodial SPV nests your claim behind a third party that needs the issuer's cooperation; the May "void" statements proved that cooperation can be withdrawn overnight (PreStocks ANTHRO −27% on day one). A structured note (Republic) sidesteps the equity-transfer ban by being debt, and native issuance (Aktionariat) is immune because the issuer mints it — but neither is the "shares" the marketing implies. Read the wrapper, not the ticker.

Perpetuals win flow for reasons unrelated to quality. They dominate volume because CEXs convert an existing retail base with leverage and 24/7 liquidity — not because the price is more "real." With no tradeable spot to converge to, the mark is whatever the venue's oracle says, and funding is the only honest signal of one-sidedness. A fixed 0% funding (OKX) defers all of that tension to a single IPO-Rebase moment. You are trading the venue's mechanics as much as the company.

Prediction markets are honest about being bets. That honesty is the point: no equity, no "void" exposure, no funding tail, and the cleanest US path (Kalshi as a CFTC DCM). The limitation is structural — a binary can tell you the probability SpaceX lists by a date, but not what a share is worth today. Resolution-source risk replaces oracle risk. Great for timing and thresholds, useless for continuous valuation.

The cross-cutting trap: implied valuations don't compare. The same per-share view produces wildly different "implied valuations" across venues because each assumes a different share count — a unit mismatch, not a market disagreement. Layer on the fact that a thin spot pool can post a trillion-dollar implied cap on a few hundred thousand of real depth. Anchor on per-contract price and real liquidity, never the headline market cap.

What's actually being priced — the underlying valuations, June 2026

Company	Latest primary round / deal	Secondary-market mark	Pre-IPO presence
SpaceX incl. xAI since Feb 2	IPO fixed $135 · $1.75–1.8T (Jun 3); merger valued the combo ~$1.25T in Feb	Hiive $146.40 · Notice $116.17 /sh (Jun 9)	All three tracks, every major venue
OpenAI	$122B at $852B post (Mar 31) — largest private round ever	~$830B implied · Notice $676.15/sh (Jun 10)	Spot · perps · prediction; confidential S-1 May 22
Anthropic	$65B Series H at $965B (May 28) — 2.5× its Feb $380B round	Forge $589.01/sh ≈ ~$1T (2.6× Feb primary)	Spot · perps · prediction; confidential IPO filing Jun 1
Anduril	$5B at $61B (May 13) — 2× in a year; 2025 rev $2.2B	n/d	Spot tokens (PreStocks · Jarsy)
Neuralink	$650M at ~$9B pre (Jun 2025)	~$9.7B tracker marks	Spot — the hottest PreStocks listing by run-rate
Stripe	Tender at $159B (Feb 24) — +74% YoY	n/d	Prediction (17% IPO-by-2027); near-zero spot ($10.5K on Jarsy)
Databricks	$5B at $134B (completed Feb 9)	n/d	Jarsy spot; Polymarket NPM market set
Kalshi · Polymarket	Kalshi $22B (Coatue, Mar) · Polymarket $15B talks (Apr)	n/d	Listed as PreStocks spot tokens — venues as assets

One company, six prices — SpaceX per share, June 8–10: traditional marks at or below the deal; crypto perps 15–26% above, two days before convergence

Notice.co algorithmic mark · Jun 9

$116.17

IPO — fixed Jun 3, the deal itself

$135.00

Hiive order-book estimate · Jun 9

$146.40

Perp VWAP cross-venue · Jun 8

~$155

Binance SPCX vs BitMEX ~$160

~$170

TradeXYZ peak May 18 launch spike

$216

$110$165$220

Per-share normalization is approximate — venues assume different share counts (OKX divides by 1B notional shares; Coinbase quotes valuation units where 1,735 = $1.735T), so cross-venue comparisons are indicative. The ordering is the point.

The same trade, three tracks — SpaceX flow compared

Track	Instruments	Cumulative volume	Open positions / backing	Signal vs the $135 deal
Spot	PreStocks SPACEX · Bitget preSPAX · Bybit IPO Express · Jarsy	PreStocks $319M; preSPAX $177M committed	~$6.2M TVL (PreStocks, Jun 10)	Clustered at/below $135–145 — custody, not discovery
Perpetual	13+ venues, SPCX-style contracts	$1.97B SpaceX · $2.94B all names; peak day $391M	OI ~$155M (Binance, Jun 10)	Marks $160–170 = +15–26% premium, two days out
Prediction	Polymarket (×NPM) · Kalshi IPO events	$20.9M volume (Polymarket + Kalshi)	$7.9M open interest (our data)	Jun-12 leg 97%; modal valuation bracket sits exactly on the deal

Perps discover, predictions calibrate, spot custodies. The perp track carries ~95× the prediction track's volume on the same name ($1.97B vs $20.9M on SpaceX), yet it is the prediction bracket that sits precisely on the priced deal while perps hold a stubborn +15–26% premium into a fixed-price conversion. Volume buys liquidity and leverage; it does not automatically buy calibration. Spot trails both on flow because its product is ownership, not price — which is exactly why its failure mode (the "void" shock) is legal, not market.

One sentence to carry through the whole report: these three tracks share a headline ("trade SpaceX before it lists") but almost nothing else — different legal claims, different price-discovery engines, different regulators.

Chapter02

Spot tokenization

The spot pre-IPO market

Custodial SPV · structured note · native issuance · MiFID broker

Spot is the track where you hold a claim — on shares, on a note, on a broker contract. The legal wrapper is everything: the May "void" statements graded the four models in a single overnight move. Nine months of venue-level data show a market that grew 500× in volume, repriced 22.5% lower in 48 hours when that wrapper risk fired, and still backs trillion-dollar implied caps with $33M of real TVL.

Spot model	Token =	Needs issuer cooperation?	"Void" impact
Custodial SPV Jarsy · PreStocks · Bitget IPO Prime	Beneficial interest in shares	Yes — secondary transfer	Direct hit: Anthropic −27% day one; −34% to −46% on the week (by source)
Structured Note Republic rSPAX	10-yr contingent debt instrument	No	Near-zero (note ≠ equity transfer)
Native Issuance Aktionariat (Swiss DLT Act)	Issuer-minted security	Is the issuer	Immune by construction
MiFID Broker Robinhood EU	Contractual claim, non-transferable	Broker compliance	Outside frame; Lithuania probe

Venue registry — what you actually buy, and for how much

Venue	Wrapper	Min ticket	Fees	Verified scale	Fine print
PreStocks Solana · Aug 2025 · Republic Capital-backed	SPV beneficial interest (SPL token)	DEX-level	Spread + mint/redeem fee (% undisclosed)	$22.1M TVL · $1.36B cum vol (our data); self-reported "$1.25B+ vol, 20K holders"	Never published the attestation reports promised at launch; no third-party audit; no public response to the void statements
Jarsy Base · Delaware LLC SPVs · $5M pre-seed (Breyer)	1:1 token per share held	$10	Platform fee + carry (% undisclosed)	$11.2M TVL across 28 names (Jun 9)	"Settlement is not guaranteed and is subject to available liquidity"; 6–12-month lock-ups possible
Bitget IPO Prime Solana · tokens issued by Republic	Tokenized debt, 1:1 post-IPO economics	100 USDT	No mgmt fee / no carry (claimed)	preSPAX: 14.4K users, $177M committed vs ~$61M allocation (~3×); preOPAI ($725/tok): $119M vs $21.08M (~5×)	Redemption ~6 months after the underlying IPO; preOPAI launched into the void-statement week anyway
Bybit IPO Express via Kraken xStocks / Backed (Jersey)	Tracker certificate, claimed 1:1 share-backed	100 USDC	5% underwriting fee	$9.1M committed by Jun 9 (window Jun 7–11)	Indicative 135 USDC = the IPO price + fee; max 50 orders/user; excl. US/UK/CA/AU

Meanwhile, TradFi is buying the traditional rails. Forge Global — $17B lifetime secondary volume — sold to Charles Schwab for $660M (about a third of its SPAC-era price); EquityZen is now a Morgan Stanley subsidiary; and Linqto, once a top retail pre-IPO platform, went through Chapter 11 after new management found securities-law violations (13,000+ customers, >$460M of historical transactions; plan confirmed Feb 2026 (~95% customer approval) with Forge as liquidating trustee). The crypto venues in this chapter are rebuilding, with thinner disclosure, exactly the access layer that just consolidated into broker-dealers — Linqto is the cautionary case for what happens when an SPV-claim platform gets the paperwork wrong.

The void statements, verbatim. Anthropic: SPV transfers are "void under our transfer restrictions"; third parties selling exposure via "direct sales, forward contracts, tokenized securities, or other mechanisms" are "likely either engaged in fraud or offering an investment that may have no value." OpenAI: unauthorized transactions "may result in the invalidation of the underlying equity" and "carry no economic value to you." These were separate statements days apart, not a joint action — which makes the synchronized market response more telling, not less. PreStocks, the most exposed venue, has issued no public response; Bitget's Republic-issued preOPAI subscription closed >$100M the same week.

267 days of spot flow

The numbers below are built from venue-level daily series: PreStocks (the largest custodial-SPV venue on Solana) per-ticker TVL and traded volume from 2025-09-16 through 2026-06-09, plus a same-day Jarsy TVL snapshot across its 28 listed names. Together they are the best available proxy for the custodial-SPV spot track.

$1.36B

cumulative spot volume on PreStocks over 267 days — from a $752K first month to a $385M April

10.3×

TVL growth Sept 16 → May 10 peak ($2.78M → $28.7M) before the “void” shock

−22.5%

sector TVL wiped in 48h after the May 12 Anthropic/OpenAI “void” statements — never recovered

$33.3M

total tracked spot TVL today (PreStocks $22.1M + Jarsy $11.2M) — the real backing behind trillion-dollar implied caps

01 · Total spot pre-IPO volume & TVL — PreStocks, daily

Daily volume7-day avg volumeTotal TVL

Volume in bars (blue) with 7-day average (black); total TVL as the amber area on the same $ axis. Monthly volume scaled ~500× in eight months: $0.75M (Sep ’25) → $310M (Mar) → $385M (Apr) → $362M (May). June is running at $6.9M/day, about half April’s pace — the post-void, post-fixed-price cooldown.

02 · TVL by ticker — daily, stacked (the AUM picture)

ANTHROPICSPACEXOPENAIANDURILNEURALINKKALSHIPOLYMARKETXAI

Anthropic is the venue’s anchor asset — $8.3M today, 38% of venue TVL — and was the epicenter of the May 12 shock: −15.8% in one day, −34.5% over three. AI names (Anthropic, OpenAI, xAI, Neuralink) are 57% of TVL.

03 · Spot volume by ticker — weekly, stacked

ANTHROPICSPACEXOPENAIANDURILNEURALINKKALSHIPOLYMARKETXAI

Flow rotates to whatever is newest or closest to an event: Kalshi (listed Jan 23) and Neuralink (Apr 17) took over a third of recent flow; Neuralink’s $2.1M/day run-rate since listing is the venue’s highest. SpaceX did $96.6M in the last 30 days — then went quiet once the $135 price was fixed. The record week — $109M — was the week of May 11: the “void” panic was also the venue’s biggest trading event.

Per-ticker scoreboard — PreStocks, as of 2026-06-09

Ticker	TVL now	vs peak	Cum. volume	30d volume	Avg $/day	Read
ANTHROPIC listed 2025-09-16	$8.34M	-37%	$197M	$40.3M	$738K	−37% off peak — epicenter of the May 12 "void" shock; still #1 by TVL
SPACEX listed 2025-09-16	$6.18M	-17%	$319M	$96.6M	$1,195K	Volume −98% after the Jun 3 fixed price; awaiting Jun 12 conversion
OPENAI listed 2025-09-16	$2.61M	-34%	$261M	$15.3M	$977K	Followed Anthropic down in the void repricing (−34% off peak)
ANDURIL listed 2025-09-16	$1.73M	-30%	$208M	$26.2M	$780K	Steady mid-pack; defense exposure diversifies the AI tilt
NEURALINK listed 2026-04-17	$1.45M	-21%	$111M	$24.1M	$2,063K	Hottest listing — highest per-day volume run-rate on the venue
KALSHI listed 2026-01-23	$0.75M	-8%	$165M	$73.5M	$1,197K	#2 in 30-day flow; prediction-market operator as a spot asset
XAI listed 2025-10-02	$0.41M	-81%	$93M	$0.4M	$372K	Peaked in January; no trade printed since May 27 — a dead listing

Second venue check — Jarsy TVL snapshot (2026-06-09)

Jarsy · $11.2M across 28 names — top-3 hold 67%

SPAX SpaceX Tech. Corp.

$3.91M

ANTH Anthropic, PBC.

$2.02M

XAI X.AI Corp.

$1.56M

RFLX Reflection AI Inc.

$0.61M

ANDL Anduril Industries, Inc

$0.51M

PSIQ PsiQuantum Corp.

$0.48M

POLY Polymarket (Blockratize Inc.)

$0.42M

KRAK Payward, Inc. (Kraken)

$0.22M

ANY Anysphere Inc. (Cursor)

$0.21M

DATA Databricks Inc.

$0.19M

18 others incl. Stripe $10.5K · Sesame AI $1

$1.10M

Jarsy lists more than 3× the names PreStocks does, but the long tail is near-empty: 18 tickers share $1.5M, Stripe holds $10.5K and Sesame AI literally $1. Breadth of listings ≠ breadth of liquidity.

The “void” shock is now quantified — and it left a permanent mark. Sector TVL fell 22.5% in the 48 hours after May 12 ($28.7M → $21.0M) on record exit volume (Anthropic alone traded $4.7M on May 12, ~6× its daily average). Four weeks later TVL sits 23% below peak. The legal-wrapper risk in the table above is not theoretical: it repriced the entire venue overnight, and capital has not come back.

The SpaceX paradox: a fixed price kills a spot market. SPACEX traded $8.2–8.6M/day around the June 1–2 pricing run-up, then collapsed ~98% to ~$110–300K/day once the $135 price was fixed on June 3. With the answer published, there is nothing left to discover — TVL is bleeding out ahead of the June 12 conversion. Watch whether that capital recycles into the next names (Anthropic, OpenAI, Neuralink) or leaves the track.

Listings are the growth engine. Each new ticker reset the venue’s volume highs — Kalshi’s January listing preceded the February jump ($20M → $133M monthly), Neuralink’s April listing the April peak ($385M). New supply, not deepening books in old names, is what grew this market 500×.

Watch the gap, not the headline. All tracked spot venues combined hold $33.3M of TVL against trillion-dollar implied caps — a liquidity-to-cap ratio in 10⁻⁵ territory. And listings can simply die: xAI peaked in January, lost 81% of its TVL, and has not printed a trade since May 27. The AUM/TVL series, not the implied market cap, is what separates genuine backing from a quote a single mid-size order could break.

Chapter03

Perpetual futures

The perpetual pre-IPO market

No credible spot → funding either runs away or is suppressed

This is where the volume is. One synthetic "SpaceX perp" is really five different products once you look at how each one sets its mark — and by June 2026 the centralized venues had multiplied from a handful to nearly a dozen, with Binance setting the pace.

Fully market-endogenousFully external-oracle driven

Hyperp

TradeXYZ

No external oracle; EMA of its own marks; 0.005× damped funding

Hybrid

Ventuals

⅓ Notice + ⅔ 2h EMA — failed May 28

Off-chain push

Helix

SEDA + Caplight, every 4 hours

Black box

Binance/OKX/Bybit

No formula; Rebase on S-1

Internal DEX

Aster / Gate

Own mark logic

Read funding as the tell. With no real spot to converge to, funding is the only honest signal of how one-sided the book is — and 0% funding could be the most dangerous design, not the cheapest: it defers all convergence pressure to the single IPO-Rebase moment.

The five products, compared in detail

Design	Venue(s)	Mark / oracle	Funding mechanism	14-day median APR	IPO conversion	Key trade-off
Hyperp market-endogenous	TradeXYZ (Hyperliquid HIP-3)	No external oracle; EMA of its own book (mark capped vs the trailing EMA)	0.005× damped multiplier	+1.67%	Auto-converts to listed-equity perp	Most transparent & manipulation-resistant; but liquidity-sensitive and unproven at $1T scale
Hybrid blended	Ventuals (Hyperliquid HIP-3)	⅓ Notice secondary data + ⅔ 2-hour EMA	Standard (uncapped)	+43% / +39% / +56% OpenAI / SpaceX / Anthropic medians; peaks to 415%	First-day close; undisclosed if no IPO	Real external anchor, but runaway funding = LPs carry an IPO-settled tail — and the anchor itself failed May 28 (−43% in 7 min on a bad Notice print)
Off-chain push single external feed	Helix (Injective)	SEDA + Caplight, pushed every 4 hours	Hard-capped ~2%	≤2.4%	Standard perp conversion	Cap protects LPs, but a single 4-hour feed could create risk
Black box CEX discretionary	Binance · OKX · Bybit · Bitget · Coinbase · Kraken · BitMEX	Mostly undisclosed marks; Rebase on S-1. Exception: Kraken discloses a smoothed "PreMarket Synthetic" index clamped ±0.25%/interval	Discretionary (Binance manual, params changed mid-cycle; OKX & BitMEX fixed 0%)	0% – manual	Auto-converts to live equity perp on listing	Deepest liquidity & retail reach; least verifiable, 0% funding concentrates tail at Rebase
Internal DEX own mark logic	Aster · Gate (Lighter/Ostium: no pre-IPO product verified)	Protocol-internal mark	Varies by venue	n/d	Varies	On-chain & composable; smaller books, mechanics vary and are less documented

"n/d" = not disclosed. Funding APRs are 30-day medians from our cross-venue dataset, window 2026-05-11 → 06-10.

The CEX field doubled in three weeks — Binance led

The launch calendar tells the story of a category racing one IPO date.

Venue	Launched	First / notable contracts	Note
TradeXYZ DEX · HIP-3	May 1	CBRS → SPCX (May 18) → QNT (Jun 4)	Opened the category; CBRS final-hour VWAP within 1.3% of the $350 open; SPCX: $150 initial ref, ±20% discovery bands, $150M OI cap
OKX	May 7	SPACEX · OPENAI · ANTHROPIC	Mark = market cap ÷ 1B notional shares; fixed 0% funding, settled 8-hourly
Binance	May 21	SPCXUSDT → OPENAIUSDT (May 26)	>$280M in first 5 days; OPENAI >$41M at launch; 1.1× contract-size rebase June 10; funding capped ±2% post-listing
Bybit	May / Jun 7	SPCXUSDT perp + "IPO Express" tokenized allocation	IPO Express via Kraken xStocks / Backed: 135 USDC + 5% fee, $9.1M committed by Jun 9
Coinbase Intl	June 3	SPCX-PERP (pipeline: tech/AI/energy/space)	Valuation-index units (1,735 = $1.735T); P&L-neutral rebase with 5-min TWAP bridge at IPO; excl. US/CA/UK/SG/IN/AU
Kraken	June 7	SPCX	"PreMarket Synthetic" index — exponentially smoothed, clamped ±0.25%/interval; the most disclosed CEX mark; excl. US/EEA/CA/AU/NZ
BitMEX	May–Jun	SPCXUSDT	Fixed 0% funding; quoted ~$160 vs Binance ~$170 in early June — a ~1.5% rebase-convergence arb it published itself
Aster	May 18	SPCX → OPENAI (May 26)	Synthetic perps with explicit no-equity disclaimers
Crypto.com	2026	OpenAI · SpaceX · Anthropic	Institutional only, by request; external private-market data feeds
Helix · Ventuals · Gate	2025–26	SpaceX · OpenAI · Anthropic · Perplexity	On-chain venues; oracle designs differ (see above); Ventuals up to 10×

Spotlight · Binance pre-IPO perps

Binance — product facts
First contract	SPCXUSDT · May 21, 2026
Second contract	OPENAIUSDT · May 26, 2026
First-5-day volume (SPCX)	>$280M
Settlement	USDT-margined & settled
Rebase	June 10, 2026 — a 1.1× contract-size adjustment to the actual ~13.08B share count in the June 3 S-1/A; positions auto-convert to the live equity perp on the June 12 listing
Mark, June 8	~$160–170 (Castle Labs); cumulative SPCX volume >$1B by Jun 8
Funding	Capped to a ±2% band after the listing window (no disclosed manual schedule)
Positioning	Part of Binance "financial super app" strategy (Abu Dhabi / ADGM HQ)

Why it matters
Reach	Converts an existing global retail base with no new KYC — the structural reason CEXs win flow
Liquidity	Deepest book of any pre-IPO venue; continuous pricing pre- and post-listing
The catch	No disclosed mark formula; the rebase + funding cap are the only published mechanics. Users bear unverifiable platform discretion
Disclaimer (Binance's own)	Price may stay below final IPO price; no guarantee the IPO proceeds

The funding picture — measured

30-day median funding APR — managed (CEX) vs market-driven (DEX) · our data

Managed / fixed — CEX

OKX all 3 · pinned

Bybit SPCX · pinned

~0%

Bitget mixed · mostly flat

~0% (occas. ±)

Binance all 3 · fixed 0.005%/8h

5.5%

Market-driven — DEX (longs pay the carry)

TradeXYZ SPCX · damped 0.005×

~0% med · ▲82% peak

Ventuals vntl:OPENAI

+43%

Ventuals vntl:SPACEX

+39% · ▲162%

Ventuals vntl:ANTHROPIC

+56% · ▲415%

Medians over the 30-day window 2026-05-11 → 06-10 (▲ = single-day peak, true value labelled). The split is the whole story: the four CEXs express no long-crowding signal — Binance is a fixed 5.5%, the rest pinned to zero — so the only honest read of how one-sided the book is comes from Ventuals, where Anthropic longs paid a 56% median and a 415% spike on May 27. Zero funding is a design choice that defers all convergence pressure to the IPO Rebase, not evidence the book is balanced.

Perp flow — measured across 10 venues

Built from venue-level daily notional volume for every tracked pre-IPO perp (SpaceX, OpenAI, Anthropic, Quantinuum) across Binance, OKX, Hyperliquid HIP-3, Bitget, MEXC, Aster, Lighter, Apex, GRVT and Pacifica — $2.94B cumulative, Nov 2025 → Jun 2026.

01 · Monthly perp volume — log scale (every month legible)

Log scale, so the standing-start months read alongside the explosion. Hyperliquid alone carried the market at $5–23M/month for half a year (green); once the CEXs arrived the market jumped two orders of magnitude — $1.18B in May, $1.67B in the first eight days of June (Jun* = partial month, to Jun 8). Peak day: $391M on June 8.

01b · Who owned each month — venue share of volume

BinanceOKXHyperliquidBitgetMEXCOther

The same data normalized to 100%: Hyperliquid (TradeXYZ + Ventuals) was effectively the entire pre-IPO perp market through April — then Binance and OKX launched in May and took ~88% of flow in their first month. The on-chain venue invented the category and was out-distributed by the CEXs within weeks of their entry.

Chapter04

Flow analysis

The flow war: a CEX duopoly

Binance + OKX = 82% of all pre-IPO perp volume; Hyperliquid (8%) is the only DEX with real share — our cross-venue data, Nov 2025 → Jun 2026

7.1_×

Binance $1.68B ÷ Hyperliquid $237M — cumulative pre-IPO perp volume, all names

Cumulative pre-IPO perp volume share by venue ($2.94B across 10 venues · our data)

Binance 56.9%

OKX 24.9%

Binance $1.68B OKX $734M Hyperliquid · TradeXYZ + Ventuals $237M Bitget $147M MEXC $107M · Aster/Lighter/Apex/GRVT/Pacifica $43M

02 · Cumulative volume by underlying

SpaceX

$1967M · 67%

Quantinuum

$385M · 13%

OpenAI

$311M · 11%

Anthropic

$281M · 10%

SpaceX is two-thirds of all pre-IPO perp flow. Quantinuum (QNT) — perp launched on TradeXYZ May 29, company IPO'd June 4, also live on Binance — is already the #2 name, ahead of both OpenAI and Anthropic: flow follows the freshest listing, not the biggest company.

03 · Volume by venue — exact

Venue	Cumulative	Share	Last 30d
Binance CEX	$1,675M	56.9%	$1,675M
OKX CEX	$734M	24.9%	$641M
Hyperliquid DEX	$237M	8.1%	$136M
Bitget CEX	$147M	5.0%	$147M
MEXC CEX	$107M	3.6%	$100M
Other mixed	$43M	1.5%	$42M

Last-30-day column shows the duopoly tightening, not loosening: Binance + OKX held ~85% of the recent window as the IPO approached.

This is a two-CEX market. Binance and OKX together took 82% of $2.94B in cumulative pre-IPO perp volume; Hyperliquid is a distant third at 8%, just ahead of Bitget (5%) and MEXC (3.6%). The entire long tail of on-chain venues (Aster, Lighter, Apex, GRVT, Pacifica) is 1.4% combined. CEXs win flow for structural reasons: an existing retail base, no new KYC, fiat rails, leverage.

Chapter05

Prediction markets

The prediction market

No equity-law risk · discrete events only · the calibration layer

The third track sidesteps the legal question entirely: you trade the event — will the IPO happen, will valuation cross a threshold — not a claim on the company. Two years of data across Polymarket (96% of volume) and Kalshi (4%) show a market that was a rounding error until late 2025 and has since become a $40.6M venue with $7.9M of live open interest — overwhelmingly concentrated in SpaceX, and increasingly about valuation rather than timing.

$40.6M

cumulative prediction volume (Polymarket + Kalshi), Jun 2024 → Jun 2026, across 25 private names

$7.9M

open interest — an all-time high on Jun 10, having risen every month since launch and never drawn down

75%

SpaceX share of open interest (51% of volume) — the most concentrated of all three tracks

65%

of open interest sits in "closing market cap" bets — the open question is at what valuation, not if

The category went mainstream in six months — monthly volume

01 · Monthly pre-IPO prediction volume — log scale

Log scale, so the $5K months of 2024 read next to the $8M months of 2026. Dormant — a few thousand dollars a month, almost all on Kraken — until Polymarket's September 2025 expansion to seven AI/tech names, then compounding: $0.38M (Oct '25) → $1.4M (Nov) → $4.0M (Jan '26) → $8.3M (Mar) → $8.6M in the first ten days of June (Jun* partial). By year: 2024 $162K, 2025 $3.7M, 2026 $37M.

This is the calibration layer, sized — and now measurable in conviction. At $40.6M cumulative volume and $7.9M of open interest it is ~75× smaller than the perp track ($2.94B) and ~35× below cumulative PreStocks spot ($1.36B) — yet it is the only track whose central estimate sits on the actual deal: Polymarket's modal SpaceX bracket lands on the priced $135 ≈ $1.75T. Small money, sharp signal, rising conviction. Kalshi is now a real participant (4% of volume but 9% of open interest) — the regulated venue punches above its flow weight in positions actually held. Treat the track as the independent probability anchor against the perp and spot marks.

Open interest — the conviction signal the volume chart hides

02 · Open interest by ticker — daily, from January 2026

SpaceXOpenAIAnthropicStarlinkOther

Volume is spiky and event-driven; open interest is a one-way climb — and it is overwhelmingly one name. From $0.9M in early January to $7.9M on Jun 10, an all-time high with no monthly drawdown; SpaceX holds $5.9M (75%) of it, with OpenAI ($0.7M), Anthropic ($0.6M) and Starlink ($0.1M) far behind. Daily volume is now only ~4% of standing OI: traders are holding positions into the IPO, not churning them.

03 · What gets traded — category mix by month

IPO TimingIPO Closing Market CapCross-Company IPO Comparison

As a name nears its IPO, betting rotates from timing ("will it list") to closing market cap ("at what valuation"). Dec 2025 was 70% timing; June 2026 is ~80% market-cap — and 65% of all open interest now sits in valuation brackets. Once the date is near-certain, the open risk moves to price.

Live markets — fetched June 10, 2026

Platform	Market	Latest pricing	Volume	Resolution basis
Polymarket	SpaceX IPO by June 15 / 30 / Dec 31	99% / ~100% / ~100%	$3.96M (event)	Official announcements + "credible news sources"; acquisition → No
Polymarket	SpaceX IPO date = June 12	97%	n/d	Exact-date leg
Polymarket	SpaceX first-day closing market cap	>$2.0T 64% · $1.8–2.0T 21%	$2.72M	Shares outstanding × first-day close (NPM data)
Polymarket	SpaceX IPO valuation $1.75–2.00T	93%	n/d	Sits exactly on the $135 ≈ $1.75T deal
Polymarket	OpenAI IPO by Sep 30 / Dec 31	30% / 73%	$1.94M	Confidential S-1 filed May 22, confirmed Jun 8 (GS/MS/JPM)
Polymarket	Anthropic or OpenAI to IPO first?	Anthropic 70.5%	n/d	Anthropic filed confidentially Jun 1
Polymarket	IPOs before 2027 — Stripe · xAI	17% · <1%	n/d	xAI leg dead since the Feb 2 SpaceX merger
Kalshi	KXIPOSPACEX — SpaceX IPO announcement	effectively resolved	$1.76M cum.	S-1 effective OR priced OR exchange ticker assigned
Kalshi	KXIPO-26 — who announces an IPO in 2026	OpenAI "a couple of months behind Anthropic's"	n/d	Per-company legs (Discord, OpenAI, Anthropic…)

Live odds fetched Jun 10. Kalshi is the smaller venue by flow ($1.76M cumulative, 4% of the track) but holds 9% of open interest; neither platform publishes per-leg OI, so the figures above are event-level. Volume and OI series are from our cross-platform dataset.

04 · Cumulative volume by underlying

SpaceX

$20.9M · 51%

OpenAI

$5.9M · 15%

Anthropic

$3.5M · 9%

Kraken

$2.3M · 6%

Discord

$1.5M · 4%

Databricks

$1.2M · 3%

Cerebras

$949K · 2%

Consensys

$752K · 2%

Fannie Mae

$579K · 1%

Stripe

$548K · 1%

22 others

$2.5M · 6%

SpaceX alone is 51% of volume and 75% of open interest — even more concentrated in held positions than in flow. The same hot-name skew as the spot and perp tracks; Starlink and Kraken are the only non-AI names of any size.

Chapter06

Regulation

The regulatory fork

One door closing, another opening

Closing · SEC

Innovation Exemption

Jan 28: Corp Fin statement on tokenized securities; Atkins "Token Safe Harbor" speech Mar 17
May 21 (Peirce): exemption will be "narrowly tailored" — on-chain versions of existing public equities with voting + dividends preserved; synthetics & price-trackers excluded
May 22: exemption delayed after Nasdaq/NYSE/Cboe object (Reg NMS bypass, CAT reporting, surveillance, issuer-consent gaps) — no new timeline
Net: US tokenized-equity launches slip toward 2027; DTCC tokenized pilot still on (Jul 2026)

⇋

Opening · CFTC updated

Regulated crypto perps

Chair Selig (sworn in Dec 2025); Mar 3: will "clear path for US perpetual futures in coming weeks"
May 28: KalshiEX BTC perp approved under Reg 40.3 — order holds perps on digital commodities are futures under existing law, open to every DCM
Same day: Coinbase no-action treats its Bermuda perps as "foreign futures" for US-adjacent access
ETH perp live Jun 4; 12 altcoins + HYPE filed. Equities are the open question — regulators are publicly silent on contracts referencing SPCX, and silence is the finding

IPO-cancellation protection

What happens to your position if the listing never comes:

Platform	On cancel / delay	Downside protection
TradeXYZ	Lifecycle TWAP (only formula)	★★★★★
Republic rSPAX	Note-maturity formula	★★★★★
Ventuals	First-day close; undisclosed if no IPO	★★★
OKX / Binance / Bitget / Coinbase / Kraken	Sole discretion / fair-value	★
Jarsy	"Redemption not guaranteed"	★

Chapter07

Timeline

How fast it formed

Twelve months, zero to contested battleground — the full timeline

The market for trading the world's largest private companies before they list went from a single tokenized note to a thirteen-venue battlefield — and a $1.75–1.8 trillion IPO test — in one year.

2025-06

Republic rSPAX note launches; Robinhood EU lists OpenAI/SpaceX tokens — OpenAI immediately calls them unauthorized

2025-10

Injective + Helix launch the first pre-IPO perpetual (SEDA + Caplight oracle)

2026-02-02

SpaceX acquires xAI — ~$1.25T combined entity, the largest merger ever; the IPO numerator now includes the AI bet (~$20B/yr spend)

2026-03

Hyperliquid HIP-3 makes perp-market creation permissionless — the base layer for TradeXYZ & Ventuals

2026-03-31

OpenAI closes $122B at $852B post-money — the largest private round ever (Amazon $50B, Nvidia $30B, SoftBank $30B)

2026-05-01

TradeXYZ opens the category with a Cerebras perp — 13 days before listing, no external oracle

2026-05-07

OKX lists the first CEX pre-IPO trio — SPACEX / OPENAI / ANTHROPIC at fixed 0% funding, mark = market cap ÷ 1B notional shares

2026-05-12/13

Anthropic & OpenAI publish separate "void" statements days apart — SPV transfers "void under our transfer restrictions"; tokens plunge ~40% (ANTHRO −38%, OPENAI −46% on the week)

2026-05-14

Cerebras debuts (CBRS): priced $185, opens $350, closes +68% at $311 — the hyperp's final-hour VWAP ($354.54) lands within 1.3% of the open; the IPO window reopens

2026-05-19/20

Polymarket × Nasdaq Private Market launch private-company prediction markets (OpenAI, Anthropic, SpaceX, Stripe, Anduril, Databricks); SpaceX S-1 goes public May 20

2026-05-21

Binance lists SPCXUSDT ($280M in 5 days); Bybit follows — SEC's Peirce says the innovation exemption excludes synthetics; UK FCA puts Hyperliquid on its warning list the same day. Binance adds OPENAIUSDT May 26 (>$41M at launch)

2026-05-22

OpenAI files a confidential S-1 (confirmed June 8); the SEC delays the innovation exemption after Nasdaq/NYSE/Cboe object; ICE licenses Brent/WTI to OKX for perps

2026-05-28

CFTC clears the first US perpetual (KalshiEX BTC; Reg 40.3: perps are futures under the CEA) + a Coinbase no-action treating its Deribit FZE perps as foreign futures. Same day: Anthropic raises $65B at $965B, and Ventuals' SpaceX perp flash-crashes −45% on a bad Notice print after SpaceX's 5-for-1 split (405 wallets / 1,393 positions, $1.51M)

2026-06-01

Anthropic files confidentially for an IPO — three days after its $965B round; Polymarket has it 70.5% likely to beat OpenAI to listing

2026-06-03

SpaceX sets a fixed $135/share ($1.75–1.8T, 555.6M shares, ~$75B); roadshow opens June 4. Coinbase launches pre-IPO perps the same day; Kraken follows June 7 with a disclosed, clamped index

2026-06-08/10

Books go 3.5–4× oversubscribed (>$250B demand); institutional books close June 10. Perp marks still run a premium over the fixed price; Binance applies a 1.1× contract-size rebase June 10

2026-06-12

SpaceX targeted Nasdaq debut (SPCX) — pre-IPO perp positions auto-convert to the live equity perp. The first trillion-dollar test of on-chain price discovery