The 370% Gap: Unitree’s Pre-IPO Perpetuals Are a Sentiment Gauge, Not a Valuation Signal

On August 6, 2025, a crypto derivatives protocol called Serenity opened a perpetual futures market on a company that has not yet completed its initial public offering. The implied market capitalization of Unitree, the Chinese humanoid robot maker, jumped to $29.3 billion overnight. The company’s own IPO prospectus targets a valuation between $5.7 billion and $6.2 billion. The gap is not a rounding error. It is a 370% to 414% mismatch between two markets that are supposedly pricing the same asset. That divergence deserves a more skeptical treatment than it has received. Crypto-native pre-IPO perpetuals are an interesting piece of financial infrastructure, but they are not a price discovery mechanism in any meaningful sense. They are a sentiment monument. The robots will eventually march to the stock exchange, but the derivative market has already priced in a parade. I have spent twenty years in macro strategy and the last eight mapping crypto derivatives onto traditional asset classes. I know a market that is one crowd away from a stampede when I see one. The most dangerous number in any liquidity cycle is the one that cannot be arbitraged against, because it means the price is a narrative rather than a settlement. Unitree’s $29.3 billion implied valuation is exactly such a number. Context: What Serenity Is Actually Doing Serenity operates in the thin boundary layer between crypto derivatives and private equity. Traditional pre-IPO trading is closed, slow, and reserved for accredited investors with legal teams. Serenity opens that trade to anyone with a wallet. Traders deposit collateral, take long or short positions on a private company’s valuation, and pay or receive funding rates. When the company eventually lists, the contract is settled against the market price of the newly public shares. The mechanism has precedent. Cerebras, the AI chip company, and SpaceX both had pre-IPO perpetuals on other platforms. In both cases, per Serenity’s own commentary, the contract price eventually landed close to the actual opening price. That evidence is often cited as proof that the model works. In statistical terms, it is proof of nothing. The sample size is two, the selection bias is enormous, and the underlying assets are not comparable to a Chinese humanoid robot maker with a convoluted shareholder structure and a supply chain that spans two contested trade regimes. I do not dismiss the innovation. Standardized derivatives for pre-IPO exposure is a genuine improvement over phone calls and paper documents. It democratizes access to a market that was only open to friends of late-stage funds. It also creates a paper trail, a transparent order book, and a real-time funding rate. For that alone, the model deserves attention. But democratization of access does not mean validity of price. The old system was slow because institutional money required diligence, lockups, and negotiated discounts. The new system is fast because crypto traders want leverage and immediate gratification. Speed is not a proxy for correctness. Core: The Price Discovery Failure at the Heart of $29.3 Billion The essential question is not whether Unitree is worth $29.3 billion. The essential question is whether Serenity’s order book is capable of producing a fair price for a company that has never traded in a regulated market. The answer is no. In traditional IPO pricing, the underwriter collects bids from institutional investors, runs a book-building process, and sets a price that clears demand. The price is anchored by audited financials, comparable company analysis, and a discounted cash flow model. None of that exists inside a perpetual contract. The perpetual price is the aggregate of leveraged traders with no fundamental anchor. It is pure supply and demand in a market where the total float of available exposure is determined by the platform’s open interest caps, not by the company’s actual shares. This is why the $29.3 billion number is dangerous. It looks like a market price, but it is actually a margin squeeze. When a perpetual market first launches, the long side tends to dominate because shorting private companies is operationally difficult. The funding rate becomes a penalty paid by shorts, which pushes more traders to the long side, which pushes the price higher. The process is reflexive, and it does not stop until the funding rate becomes so expensive that even the most confident bull refuses to pay. There is no evidence that any algorithmic market maker stepped in to arbitrage the 370% gap. There is no evidence that a serious short seller could hold a position without being liquidated by funding payments. There is, in fact, a structural reason why the gap cannot close quickly: the perpetual market cannot deliver the underlying shares. It is a cash-settled derivative, so the traditional arbitrage model of shorting the expensive future and buying the cheap spot is unavailable. You cannot short a private company until it lists. You cannot buy the pre-IPO shares unless you are in the actual financing round. The only people who can arbitrage the gap are the very people who already hold Unitree equity, and they are not allowed to sell until well after the IPO. That is the core insight I want readers to take from this article: the 370% gap does not mean Unitree is undervalued. It means the pre-IPO perpetual market has no redemption mechanism. It is an opinion market with a funding rate. Code is law, but man is the loophole. Let me be more specific about the pricing mechanics. If Unitree lists at the top of its targeted range, $6.2 billion, and the derivative price currently implies $29.3 billion, then the contract is pricing a first-day gain of 372% above the high end of the offering. For context, the most famous recent tech IPO, Arm Holdings, rose about 25% on its first day. A 372% first-day gain would be so far outside the historical distribution that every institutional investor who sees it should ask one question: why would the underwriters leave that much money on the table? The answer, almost always, is that they did not. In the rare cases where an IPO is deliberately underpriced, the discount is a marketing tool, not a multi-billion-dollar gift. Discounts of three to four times the offer price are not a feature of a functioning IPO process. They are a contradiction of the term “price discovery.” The professional underwriters who set the $5.7 billion to $6.2 billion range have access to audited financials, customer contracts, order books, and channel checks. The anonymous traders in the perpetual market have access to Twitter, a chart, and a dream. I know which one I trust. The scenario table in my mind is the following. In an extreme scenario, the opening price actually does approach $29.3 billion. That would constitute an entirely new valuation regime for humanoid robotics, one in which a company with a relatively small revenue base is priced as if it were already a diversified industrial conglomerate. The probability is below 5%. In a moderate scenario, the opening price is 100% to 200% above the IPO target, which would be a strong pop but far below the derivative market’s implied level. That would mean the perpetual market was directionally right but quantitatively wrong. In the realistic scenario, the opening price settles 20% to 80% above the IPO target. The derivative market then rapidly converges to the actual stock price, and everyone who bought the perpetual at $29.3 billion learns the difference between a narrative and a balance sheet. There is also a non-trivial chance, perhaps 20%, that the stock opens up only slightly or even breaks below the offering price. In that case, the robot sector narratives will shift from “exponential growth” to “commercialization timeline.” The more important transmission effect is not the opening day print. It is the capital expenditure cycle that Unitree’s listing will trigger either way. A company that raises up to $6.2 billion from an IPO will need to spend money on manufacturing capacity, supply chain contracts, and inventory. In a humanoid robot, speed reducers alone account for roughly 30% to 40% of the bill of materials. The major suppliers—Leaderdrive, Harmonic Drive, and Ouster in the sensor stack—are therefore direct beneficiaries of any significant capacity expansion, regardless of whether Unitree’s stock opens at $30 billion or $8 billion. That is the real sector transmission story: not the ephemeral valuation of one equity, but the durable allocation of capital along the entire industrial chain. The historical parallel is the electric vehicle supply chain. Tesla did not become valuable overnight because its first-day pop was enormous. It became valuable because its capital expenditures forced suppliers to build factories, hire engineers, and sign long-term contracts. Those suppliers then re-rated exponentially. The same pattern is likely to play out in robotics. Even if Unitree’s public market valuation disappoints relative to the crypto perpetual market, the very existence of a successful IPO sends a signal to every harmonic reducer manufacturer and lidar company that the demand schedule is real. That signal is worth more than any single price print. I also want to flag the selection bias inside Serenity’s supply chain list. Ouster is not on the list because Unitree is its largest customer. Ouster is on the list because Ouster is a highly liquid, actively traded stock with a large options market and a crypto-friendly shareholder base. Serenity is running a derivatives business, not a research department. It will list the names that produce volume. That is not a conspiracy. It is a business model. But when you read Serenity’s commentary about which suppliers will benefit from a high Unitree valuation, you should discount it accordingly. Contrarian: The Gap Is Not a Sign That the IPO Is Undervalued The prevailing interpretation among crypto media is that the $29.3 billion implied valuation is evidence that Unitree’s IPO range is too low. The logic is simple: the crowd knows better than the underwriters. I think that logic is backwards. The pre-IPO perpetual market is populated by traders who are structurally biased to the long side. There is no short side capacity in any meaningful sense. The funding rate mechanism taxes shorts, and the lack of a deliverable asset means there is no arbitrage mechanism to correct a mispricing. In such a market, the price is not a consensus estimate of fair value. It is a measure of how much leverage the longs are willing to carry. When the IPO finally happens, the actual stock becomes tradable, the perpetual contract becomes settleable, and the theoretical link between the derivative and the underlying asset snaps into place. That is when the gap closes. It will likely close violently. I admit there is a contrarian alternative: that the IPO underwriters are deliberately issuing at a low valuation to guarantee a massive first-day pop, creating an impression of momentum for the broader robotics sector. But the magnitude required here is historically absurd. A 370% first-day gain would not be a pop. It would be a regulatory scandal. The Securities and Exchange Commission and the Hong Kong Stock Exchange would both ask why the underwriters left billions of dollars on the table. That is not a sustainable game. A more subtle, and more interesting, possibility is that the derivative market is not wrong about the narrative but wrong about the timing. Humanoid robotics will eventually be a multi-trillion-dollar market. Unitree may eventually be worth $29.3 billion. But the perpetual market is a spot price, not a discounted cash flow. It is saying that the company is worth that number today, with no discount for the execution risk, the supply chain fragility, or the regulatory uncertainty that the company will face over the next decade. A market that cannot discount the future is not a price discovery mechanism. It is a pressure gauge on the collective nervous system of the crypto community. I have seen this movie before. During DeFi Summer, Aave and Compound interest rates were treated as pure markets, but they were actually arbitrary parameters that had nothing to do with real supply and demand. During the NFT boom, OpenSea’s royalty enforcement was treated as immutable, but it was just a social convention that could be disabled at any time. The pattern is always the same: a new form of market infrastructure appears, and the crypto community mistakes intrinsic fragility for technical truth. Pre-IPO perpetuals are the latest example. The market is not failing because the protocol is badly designed. It is failing because the underlying asset has not yet been born. Takeaway: Ignore the Number, Watch the CapEx Cycle So what should an investor actually do with this information? The $29.3 billion implied valuation is a curiosity, not a trade. Do not base a position on it. Do not short the perpetual contract either, because the funding rate will bleed you dry before the IPO date. The only reliable signal will come after the opening bell: the ratio between the IPO price and the first-day close. That ratio will reset the entire valuation frame for every robotics company in the sector. If the first-day close is only 20% to 50% above the IPO range, the entire ecosystem gets a reality check. If it is 100% or more, expect a wave of capital into supply chain names. If it is anywhere near the $29.3 billion implied level, then we are no longer in a rational market, and the subsequent volatility will be severe. My recommendation is to watch the supply chain, not the celebrity unicorn. The companies making harmonic reducers, lidar modules, and precision actuators have a clearer revenue path than any single humanoid OEM. Their financial statements are already public, their customer concentration is already disclosed, and their earnings reports will show the actual order flow. That is the real data. The perpetual contract is just an emotion with a funding rate. The robots will arrive. The question is whether the crypto market can wait for them. Code is law, but man is the loophole, and the loophole in pre-IPO perpetuals is the absence of delivery. The safest position is to be a bystander. Let the leverage meet the opening bell. When the gap closes, it will tell you more about the future price of the entire robotics sector than any chart, any tweet, or any pre-IPO order book. Wait, measure, and then act. That is how you build a macro position that survives contact with a new asset class.