On 28 October, at 22:15 UTC, a new pool opened on Solana for a token called FLOAT. Eight days later it was worth $3.32 million.
Those eight days were rough. On 4 November it fell 37.3% to a market cap of $2.28 million — and traded $3.48 million that same day, more than the entire value of the coin.
I did not start paying attention because of the swing. I started because of what sits behind it: a platform that makes money and hands seven-tenths of it to other people. You do not see that very often.
It is a launchpad
Think of a shopping mall. The mall sells nothing itself. It provides the counters, the power and the cash registers, and anyone can rent a stall. Float is the mall, and anyone can launch a coin inside it.
Most launchpads take a cut and stop there. Float adds one thing: a coin does not have to be priced against dollars. It can be priced against a company’s stock.
It supports well over a hundred stock and ETF tokens — Nvidia, Tesla, McDonald’s, and pre-IPO names such as OpenAI and Anthropic. You pay in SOL on the front end; the platform swaps it into the stock on the back end.
Those tokens are issued by third parties such as xStocks and Sunrise. They track a share price. They are not shares. You do not own a single Nvidia share, and you have no shareholder rights. The token is only the unit your coin is quoted in.
A new coin starts at a market cap of about $20,000. At roughly $150,000 — 7.5 times — it graduates into a proper pool, and the liquidity in that pool is locked for good. Nobody can pull it.
It gives away 70% of its fees
The person launching a coin sets the fee: 1%, 2% or 2.5%. Take a day when the fees from every coin on the platform, after Raydium takes its cut, add up to $100. Here is where the $100 goes:
- $35 buys FLOAT on the open market and burns it. Gone permanently.
- $35 is converted to USDC and airdropped to the wallets with the most trading volume (100 of them in the first round).
- $15 goes to the coin’s holders, paid pro rata in the stock token that coin is paired with.
- $5 goes to the creator.
- $10 goes to the platform treasury.
The platform keeps a tenth. The first time I saw that split, I assumed I had misread it.
The second detail matters more. That $35 goes to the people who traded the most, not the people who held the longest. It rewards trading, not conviction.
Giving away 70% is not generosity. It is the price of being late. Raydium opened custom quote assets on 6 September; Pump.fun followed on 11 September; StonkFun had been running since 23 July and peaked at a $300 million market cap. Float opened on 28 October. When you arrive last, you buy volume with margin.
74.6 million tokens burned in eight days, and you can check it
FLOAT started with a supply of one billion. On 4 November at 13:25 UTC there were 939.32 million left. On 6 November at 06:20 UTC there were 925.36 million. That is 13.96 million gone in forty hours, or about 8.4 million a day. Since the first pool opened, 74.64 million have been burned — 7.46% of the original supply.
At the prices of the past few days — $0.0033 to $0.0036 — those 8.4 million tokens are worth $28,000 to $30,000.
Thirty thousand dollars a day against a $3.32 million market cap means 0.9% of the market cap is bought back and destroyed every day. That is not a large number. But it is real, it updates daily, and anyone can check it without trusting me or the team.
That money is not its own
Here is the part that decides everything else. When FLOAT itself is bought and sold on Raydium, the fees do not enter that $100 split. They go to the developer wallet.
So the $35 that funds the buyback comes entirely from coins other people launched on Float.
Which means FLOAT’s price depends on whether anyone still wants to launch here, and whether anyone wants to trade what they launch.
The mall comparison goes one step further. What you are buying is the mall’s foot traffic, not any single shop. The busier the mall, the more it is worth. If the tenants leave, the rent stops.
One coin, one table, two opposite answers
Suppose the platform does $3 million a day in volume, not counting FLOAT itself. At a 2% fee, minus Raydium’s 0.25%, that is $52,500 a day to split. The buyback’s 35% is $18,375 a day, or $6.71 million a year.
Against a $3.32 million market cap, that is 0.49 times. The market pays 49 cents for a dollar of annual buyback. Cheap, on paper.
Now run the same table at $500,000 a day. The split is $8,750. The buyback’s 35% is $3,062, or $1.12 million a year. Against the same $3.32 million, that is 3 times — three dollars for a dollar of buyback.
Same coin, same table, opposite conclusions. One number sits between them: how much the platform trades each day. Whether FLOAT goes up has little to do with how well its story is told, and a lot to do with how well everyone else’s business is doing.
I think there is one more leg up within two weeks
Three reasons.
First, on 3 November MEXC and XT both opened FLOAT/USDT within two hours of each other. Two exchanges doing the same thing on the same day suggests demand that is not confined to on-chain traders.
Second, the 4 November selloff took it from $0.005293 down to $0.001221, a high-to-low spread of more than 4x in two days. The first wave of momentum buyers has been washed out. There is a detail underneath: over those two days, tracked smart-money wallets made 11,671 buys and 7,176 sells. More buys than sells by count, roughly even by value. Small orders coming in, large orders going out.
Third, the first USDC airdrop settled on 4 November. On 1 November the team removed 14 bot wallets from the leaderboard and returned their share to real traders. That act is itself an announcement: there is real money in the pot, and someone is watching it. Once the payout lands, some of those wallets will come back to farm the next round.
If all three hold, $3.32 million to $10 million is 3x, or $0.0108 a token — roughly double the $0.005293 high.
This is the best of three outcomes, not the middle one. For it to happen, three things have to line up: platform volume reaching $3 million a day, no repeat of a 37% single-day drop before December, and at least one wave of new launches that makes on-chain trading busy again. Lose any one of them and FLOAT ranges between $2 million and $5 million, which is where I think it is more likely to settle.
The water line is the first week of December. If daily volume cannot hold $1 million, the buyback stops meaning anything.
I watch three numbers
The burn page: how much was destroyed today. It updates daily, and it is the one part of this mechanism that does not require trusting anybody.
Platform volume: how much the launchpad trades each day. That single number decides whether the coin is worth 0.49 times its annual buyback or 3 times.
Volume in the first week of December: still above $1 million a day, or not. If it is, the story survives. If it is not, there is nothing left to watch.
What worries me
No audit. The project’s own documentation says the program has not been through a third-party audit, and that upgrade authority still sits with the deployer’s key, with a move to a multisig planned. The rules are what they are today. Tomorrow they can be something else.
Ownership is concentrated. The top ten addresses hold 28.3%, the largest 5.53%. There is no unlock schedule, but there is no spread either.
Holders arrived fast: 22,182 wallets in eight days. They can leave at the same speed.
The “stocks” are not stocks. The issuer can pause transfers or move tokens out of any wallet. The Nvidia you are paired with has nothing to do with Nvidia.
The leaderboard was farmed on day one. The fact that fourteen bot wallets had to be removed on 1 November tells you that manufacturing volume cost less than the airdrop paid out. The second round is on 11 November, with the list widened from 100 wallets to 250 — the leaderboard needs fresh entrants to keep volume up.
Same name, many contracts. Several tokens are called Float. Only one of them is in this article.
I treat it as a machine, not a coin
This is not a memecoin living on a story. It is a buyback machine that happens to be issued as a token. Buying it is a bet that, on Solana, people will keep launching coins and keep trading each other’s launches for another year.
For the past eight days that has been true. Whether it is true in week twelve, I do not know. What I do know is that if it stops being true, it will show up in the platform’s volume first, and in the price second.
About me and Tronsell.io
I am Louise Lee, a researcher at Tronsell.io, working on on-chain data and memecoin research. Tronsell.io runs its own energy pool with roughly 400 million TRX staked, supplying about 3.7 billion energy and 35 million bandwidth to more than ten institutional clients. What we care about is what actually happens on-chain, not how loudly a story is told.
Prediction fact card
| Item | Data (as of 06:20 UTC, 6 November 2026) |
| Token / chain | FLOAT / Solana (SPL) |
| Contract | Bv1a7cEj5aRnJQhMw4kRQqAFesMSd5LwcDnUzq52BMmr |
| First pool | 28 October 2026, 22:15 UTC, Raydium, FLOAT/SOL |
| Price / market cap | about $0.00359 / about $3.32 million (on supply of 925.36 million) |
| Supply | 925,363,977 (from an original 1 billion; 74.64 million burned, 7.46%) |
| 24-hour volume | about $1.3 million across eight FLOAT-quoted pools |
| Liquidity | about $563,000 base-side, $367,000 in the main pool |
| Holders | 22,182 (RugCheck) |
| Top ten | 28.3%, largest 5.53% |
| Authorities | mint revoked, freeze revoked, metadata immutable |
| LP lock | main pool 100% (Burn & Earn); other pools unlocked |
| Audit | none; upgrade authority with the deployer key, multisig planned |
| Fees and split | 1% / 2% / 2.5%; after Raydium’s 0.25%: 35% buyback and burn, 35% USDC to top-volume wallets (100 in round one), 15% to holders in the paired stock token, 5% to the creator, 10% to the treasury |
| Listings | MEXC (3 November 2026, 02:40 UTC), XT (3 November 2026, 12:30 UTC+8) |
| Site / X | floatlaunch.xyz / @floatdotxyz |
| Name clash | several contracts share the name; only the address above is FLOAT |
Sources
- Price, market cap, volume and liquidity: DexScreener aggregate data, as of 06:20 UTC on 6 November 2026. Eight FLOAT-quoted pools, $563,489 of base-side liquidity and $1,300,090 of 24-hour volume. DexScreener computes market cap on the original one-billion supply and shows about $3.58 million; this article uses the on-chain supply of 925.36 million throughout, which gives $3.32 million.
- Supply, holders, top-ten share, authorities and LP lock: RugCheck report, as of 06:20 UTC on 6 November 2026. Supply 925,363,977, 22,182 holders, top ten 28.28%, main pool 100% locked, other pools unlocked.
- Earlier supply figure: DexTrading recorded 939,323,984 as of 13:25 UTC on 4 November 2026. The 13.96 million difference from item 02 is the burn over those forty hours.
- First pool: DexScreener pair creation timestamp, 2026-10-28 22:15:43 UTC.
- Fee split, fee tiers, the $20,000 starting market cap, the $150,000 (7.5x) graduation line and the permanent Raydium lock: project documentation at floatlaunch.xyz/docs, consistent with write-ups by Bittime and CoinGabbar.
- FLOAT’s own trading fees going to the developer wallet rather than into the split: project documentation, verbatim.
- Airdrop rules — first round top 100 wallets by cumulative volume since launch, settled 17:00 UTC on 4 November; second round on 11 November with the top 250; bot wallets and wash volume excluded: project documentation, as of 6 November 2026.
- 14 bot wallets removed on 1 November, and the first USDC payout on 4 November: project announcement on X, as reported by CoinGabbar and Coinokc.
- Listing times: MEXC announcement (02:40 UTC on 3 November 2026, Meme+ zone) and XT announcement (12:30 UTC+8 on 3 November 2026).
- 4 November market detail — price $0.002424, market cap $2.42 million, 24-hour change -37.3%, 48-hour range $0.001221 to $0.005293, 24-hour volume $3.48 million: DexTrading, “Why Is FLOAT Trending?”, as of 13:25 UTC on 4 November 2026. Its market cap uses the one-billion supply; against the 939.32 million tokens in existence that day it comes to $2.28 million, which is the figure used in the text.
- Smart-money buy and sell counts and values — 11,671 buys worth $1.14 million and 7,176 sells worth $986,000 over 48 hours: same source as item 10, covering its tracked wallet sample, not the whole market.
- Third-party issuance of stock tokens, the “not a share” characterisation, and the issuer’s power to pause transfers or move balances: project documentation, risks section.
- No third-party audit, upgrade authority still with the deployer key: project documentation.
Disclaimer
This is a personal research note, not investment advice. Memecoin prices are extremely volatile and can go to zero in a short time. The token discussed here launched eight days ago, the program has not been audited, the contract’s upgrade authority is still held by the deployer, and the tokenized stocks are issued by third parties that can pause or freeze their own tokens. Any one of those can wipe out the entire principal. FLOAT has no relationship with Nvidia, Tesla, McDonald’s or any other company named. Use only the contract address listed in this article, and do not put in anything you cannot afford to lose completely.