Here's the number that should stop you: $87.525. That was the last reported price on Trade.xyz for Unitree's pre-IPO perpetual contract before the official STAR Market subscription opens. Extrapolate that price across the post-issuance share count of about 404 million shares, and the market assigns Unitree a $35.4 billion valuation. The official IPO price is 150.8 yuan per share, which means the pre-IPO contract is trading at roughly 3.91 times the official issuance price. No public share has traded yet. The headline math says a single lot of 500 shares carries a "potential profit" of about 219,600 yuan. That's a 291% return on the subscription payment. Headlines like that create their own gravity.
Let's dissect what this number really is. The subscription lot costs 75,400 yuan. At the latest Trade.xyz price, that same notional 500 shares would be worth 295,000 yuan. Subtract the subscription cost, and you get 219,600 yuan. The arithmetic is fine. The assumption is not. The pre-IPO perpetual price is not a bank quote; it's a derivative print from an unregulated market with shallow depth. It embeds hopes, inventory risk, and exchange mechanics. Before we discuss profits, we have to discuss what exactly is being bought.
Context: A Derivative Wearing an IPO Costume
Unitree, the robotics company, is listing on the STAR Market in an IPO that will issue 40,446,400 shares, exactly 10% of post-issuance total capital. The share count after listing will be about 404 million. At the official issue price of 150.8 yuan, that means the issuer is selling at a valuation of roughly $9 billion. The pre-IPO contract values the same company at $35.4 billion. The difference between $9 billion and $35.4 billion is the pre-IPO market's risk premium on a robotics narrative, not a structured fact.
The product on Trade.xyz is a perpetual swap, not a forward share sale. There is no custodian holding shares behind the contract. There is a settlement index. That index could be generated from a single exchange's order book or a composite of private markets. When you buy this pre-IPO perp, you are not a shareholder. You are a counterparty in a binary bet on the future listing price. The entire "profit per lot" calculation depends on the assumption that the derivative price at expiration is the same as the cash share price. That is the first trap.
Core: The Three Assumptions Behind the 291% "Profit"
The first assumption is price convergence. A perpetual contract is supposed to converge to its underlying asset via funding and arbitrage. But for pre-IPO perps, the underlying does not exist yet. There is no spot market to arbitrage against. The exchange decides what the index is, and the funding rate is a monthly game of guess-the-listing. When the reference index is constructed in-house, it is not an independent oracle. It is an internal ledger that can be gamed by a small volume print.
The second assumption is allocation. In STAR Market IPOs, retail subscriptions are heavily oversubscribed. The 219,600 yuan "profit per lot" ignores the probability of winning the lottery. If the offering is 100 times oversubscribed, the expected profit per subscription effort is closer to 2,196 yuan, and only for those who actually get the lot. The pre-IPO perp price does not reflect the chance of allocation. It reflects a synthetic claim to shares that most subscribers will never hold.
The third assumption is liquidity. The latest price in a thin pre-IPO book is not a reliable market cap input. I have spent years querying wallet clusters on Dune, and I can tell you with precision: a single order can mark a derivative market like a pen writes a contract. When you look at a $35.4 billion valuation derived from $87.525, the first SQL you should write is a volume-by-top-wallet query. If 70% of the volume in that market comes from two addresses, the last print is not a consensus. It is a whisper.
In my own 2020 analysis of Compound and Aave, I mapped five hundred unique addresses and found that most of the "yield" was generated by three arbitrage clusters recycling the same capital. That pattern repeats in every synthetic marketplace. The actors are different, but the mechanics are identical: price is set at the margin, and the margin is controlled by the few. The 291% return is the arithmetic of the margin. The average participant's return is the arithmetic of probability and fees, and that is far less exciting.
Where the Funding Rate Tells the Truth
The real forensic evidence appears in the funding rate. A perpetual market requires longs and shorts to pay each other at regular intervals based on the difference between the contract price and the index. If the pre-IPO perp is trading at 3.91x the IPO price, the funding rate will measure how much long-side fear is subsidizing the position. If longs are paying a high positive funding rate through every interval, the premium is not free. It is a cost of carry. Every day the listing is delayed, that cost erodes the potential profit.
Let me put real numbers on that. The notional value is roughly 295,000 yuan per 500 shares. If the funding rate is 0.1% per 8-hour interval, holding for five days costs about 4,425 yuan. Not catastrophic. But in a frenzy, funding can spike to 0.5% per interval. That same five-day hold becomes a 55,000 yuan drag. The paper profit of 219,600 yuan starts to look like an upper bound, not a realistic outcome.
Watch open interest too. If the pre-IPO contract has 20% of its open interest in one wallet, the "market" is a house of cards. A single large trade can liquidate or inflate the price by double-digit percentages. In pre-IPO perp markets, price discovery is noisy before the anchor event exists. The only thing you can reliably measure is the cost of holding, not the eventual settlement price.
Another detail rarely mentioned: the timing of the cash refund. If you subscribe but do not win a lot, your 75,400 yuan is locked for days. That capital cannot be deployed elsewhere. If you win the lot, your shares are subject to trading mechanics, volatility pauses, and order-book gaps. The flawless bridge between subscription price and pre-IPO terminal price does not exist.
Contrarian: The 291% Is a Lottery Ticket With a Traded Premium
Here is the contrarian truth. The 3.91x premium is not a prediction. It is a compensation structure. Traders who hold this perpetual before an IPO are providing liquidity to an uncertain event. They demand a high discount because the event may not happen on time, the valuation may gap down, or the exchange may change its settlement rules. The premium is the rent for that risk. Conflating that rent with a realized profit is exactly the mistake that eats retail traders in every bull narrative.
There is also a structural feedback loop. The pre-IPO perp price itself reinforces the retail subscription hype. Seeing 295,000 yuan per lot on a terminal is an advertisement. It makes the IPO price look cheap. That cheapness creates more subscription demand. More demand means an even smaller chance of allocation. The 291% number is used to sell a lottery ticket, not to reveal a probable return.
I have seen this before. In 2017, I traced fourteen wallet clusters attempting to hide governance control in a supposedly transparent project. The lesson still applies: the most visible number is often the one manufactured to be visible. "Trust the hash, not the headline" is not a slogan in my workflow; it's a methodology. The hash of a subscription allocation, the hash of a block, and the hash of a transaction are more trustworthy than a pre-IPO terminal print.
What Data Will Decide the Outcome
Do not ask whether the IPO will raise 20% or 200% on day one. Ask three questions. First, does the pre-IPO book retain its open interest after the subscription date? If the smart money starts closing the derivative before the listing, they are signaling that the premium is unsustainable. Second, what is the funding rate? If the rate is deeply positive, longs are paying for a crowded trade. Third, where is the volume aggregated? If a single market maker controls both the derivative and the index, the 3.91x multiple is not a market, it is a one-line entry in a spreadsheet.
I have written before that chaos is just data waiting for the right query. The chaos of a 291% potential profit is loud. The correct query is quiet: order book depth, funding rate, open interest distribution. When you run that query, the $35.4 billion valuation either stands up or melts. Most of the time it melts.
Takeaway: The Number to Watch After Subscription
Tomorrow, when subscription opens, the only meaningful signal is not the printed perp price. It is the first funding rate after the subscription closes. If the funding rate remains positive and open interest rises, the pre-IPO premium may be real enough to survive the listing day. If the funding rate flips negative or open interest falls by 30% before the shares hit the terminal, your "profit per lot" is a memory.
Yields don't materialize from arithmetic; they materialize from settlement. Settlement is what the market does with real shares, real constraints, and real sellers. The blocks will remember the actual close, not the derivative's mid-green line. Trust the hash, not the headline.