The stranded population

What the launchpad lets a launch choose

Pons lets whoever creates a launch pick the asset the curve is quoted in. Pick ether and the result is the ordinary thing: buyers pay ether in, the reserve accrues in ether, the displayed price is a price in ether. Pick one of the tokenised Robinhood equity tokens circulating on Robinhood Chain and the curve is identical except in denomination — reserves accrue in that token, the price is expressed in it, the graduation threshold is measured in it.

None of that is a defect. The launchpad does what it documents and the choice is visible at creation. The problem appears two steps later, at a boundary neither system owns.

What the curve then does, which is work perfectly

Every curve opens with a virtual reserve of 1.68 units of its quote asset standing against a fixed supply of 1,000,000,000 tokens, so the constant product it trades along is 1,680,000,000. Buys walk the price up it, sells walk it back down, and when 4.2 units of the quote asset have been raised the curve graduates: 71.43% of supply has been sold, the curve stops accepting sells entirely, and the position becomes a concentrated-liquidity problem instead of a curve one.

All of that executes. The curve is not broken and does not refuse to buy — a holder can sell back into it at any point before graduation and will receive exactly what it says. What they receive is the quote asset. For 115,094 of the 240,350 live curves, that is where the road ends.

What no route precisely means

A route means a pair. Converting one of these equity tokens into ether requires a contract that takes the token on one side and gives back the chain asset on the other, which on this chain means a pool on the canonical automated market maker factory, against WETH or against USDG. For the tokens these curves are denominated in, the factory has neither. The lookup returns nothing, because nothing was deployed.

The distinction that matters is between a bad price and no price. An illiquid pair has a price and it is terrible: you can measure the depth, compute the slippage, and take the exit next week when somebody has added liquidity. A pair that does not exist has no depth to measure and no number to refuse. The position is not expensive to exit; it is unexitable until somebody deploys a contract.

This is also the mechanism behind the most misleading number on the chain. The deepest curve in the survey is denominated in AMC Entertainment · Robinhood Token and reports a reserve of 1625. Every screener that reads a reserve assumes the units are ether, so it prints $3,996,184 of implied capitalisation at 2459.19 USD per ether. The reserve is 1625 AMC, and its ether value is undefined, because the conversion the printed number assumes has no venue at which to happen.

How the count is constructed

Of 240,350 curves still holding a reserve at block 55,223,440, 125,256 are quoted in ether. The stranded population is defined by exclusion: everything alive that is not in that set, 115,094 curves, 47.89% of the live chain. The construction is deliberate — the figure does not depend on assembling a list of equity tokens and hoping the list is complete, which makes the ether-quoted count the load-bearing measurement. It was taken as described in Methodology.

What would falsify this

The claim asserts that something does not exist, so one transaction kills it. Deploy a pool on the canonical factory pairing one of these equity tokens against WETH or USDG, with any depth at all, and for every curve denominated in that token this page is wrong from that block onward. The position leaves the Equity-quoted · no venue tier, where the desk holds it outright because nothing offsets it, and becomes either Equity-quoted · borrowable or an asset the desk declines the way it declines everything that already has a market.

Two weaker falsifiers deserve the same directness. The survey checked the canonical factory; if a second factory on Robinhood Chain already carries these pairs, the conclusion is wrong today and the error is in the survey rather than in the world. And if the issuer of one of these tokens operates a redemption path a holder can actually reach, an exit exists that is not a pair — the pair-level claim would be right while the conclusion drawn from it was wrong. Nobid has not tested that path and does not claim to have.

The incentive should be stated plainly: the desk would rather the first of those happened. A real venue is a better outcome for a holder than a counterparty warehousing the risk, and it would remove most of the reason for this desk to exist. The case is not that a venue should not appear. It is that across 47.89% of a chain it has not.

Why the clock makes it worse

Even where a stranded position can be referenced against the listed equity behind the token, that reference is live for 32.5 hours of the 168 in a week, and the curve trades through all of them. For 80.65% of the time a holder who wants out and a desk that would quote them are both looking at a last price rather than a price.