Ninety-five percent of the float sat in one room. The chart did the rest.

Marco bought LAB on his lunch break.

He is thirty-four. He runs a warehouse shift in Tampa and he keeps a second phone in his locker because his wife does not like the trading. The second phone has Bitget on it and a Telegram folder with sixteen channels and a YouTube tab where a man with a ring light explains which AI tokens are about to run. On Wednesday, May 6, the man with the ring light said LAB. Marco watched the candle for nine minutes during his break. Green. Green. Green. He bought at $4.10.

By the next morning the position was up. By Thursday night it was a number he did not want to say out loud because saying it out loud felt like jinxing it. He told his wife the warehouse was hiring a second shift lead. He did not tell her about the phone.

On Friday the chart did something he had seen before but had not yet learned to read. It went flat. Then it went down. Then it went down faster. By 3:14 a.m. on Saturday he was sitting on the edge of his bed with the second phone in his hand and the position was at minus seventy-one percent. His daughter was asleep in the next room. The Telegram channel had gone quiet. The last message in it was a rocket emoji from two days earlier.

He did not yet know that almost every chair in the room had been held by the house.

I.

The token was called LAB. It is the native asset of LABtrade, a project that sold itself as a multi-chain trading execution layer. An AI terminal. A place where, the marketing said, the platform would not extract fees and value from its users. The founder is listed as Vova Sadkov. There is a handle, vsadkovv, and not much else. ZachXBT has tried to reach him. As of the bounty post on May 7, there has been no response.

LAB launched in late 2025 and traded below a dollar for months. Then, in the first week of May 2026, the candle did something a token does when the float is thin and the holders are not strangers to each other. It moved up more than 537 percent in one week. Some analyses traced a 3,700 percent move over the month. The peak print came in above $6. Market cap reportedly touched $1.61 billion.

There was no product announcement that explained any of it.

II.

Here is the part that matters. Here is the part Marco did not know when he hit buy on his lunch break.

ZachXBT, the pseudonymous investigator who has spent years tracing crypto fraud on-chain, alleges that insiders control approximately 95 percent of LAB's supply. On-chain analytics firm Bubblemaps put the figure higher, closer to 99 percent when you include the market maker. Other reports cite roughly 98 percent of the circulating supply sitting with the team.

Read that slowly. Ninety-five percent. Ninety-eight. Ninety-nine.

That is not a public token. That is a private holding with a public ticker. The float, meaning the actual portion of supply available for ordinary buyers like Marco to trade, was the empty chairs in a room where almost every seat was already taken.

ZachXBT's post documented the movements that preceded the surge. On April 8, a team-linked wallet deposited 40 million LAB to Bitget. At the prices then, that was around $13.6 million worth of token positioned at an exchange where it could be sold. Days before the May 1 price increase, additional team-affiliated wallets transferred 96 million LAB to Bitget. At the prices then, $63 million worth. The developer team, by ZachXBT's count, distributed roughly 100 million LAB across ten wallets coming out of Bitget, a figure valued at $480.33 million and representing over 32 percent of circulating supply.

The exchange reportedly still holds another 159 million LAB.

I worked a metals room in Chicago for a few years when I was younger. A hundred desks, a hundred phones. The room had a rule that nobody ever wrote down. You do not sell the customer what you are about to sell. You sell the customer what you have already bought and are looking to get out of. That is the whole game. The pre-positioning is the trade. The pitch is the exit.

What the wallets did in April is the pre-positioning. What happened in May is the exit.

III.

Marco did not see the wallets. Marco saw the candle.

This is what the float room counts on. The retail buyer is not watching the on-chain ledger. The retail buyer is watching the chart and the channel and the YouTube clip. He sees momentum and reads it as demand. He sees volume and reads it as discovery. He does not see that the volume is largely the same hands trading with each other, walking the price up the staircase until the stairs go away.

The futures market told its own story. Open interest in LAB futures exceeded $163 million. Over 74 percent of the open positions were shorts. That looks, on the surface, like a setup for a squeeze, and indeed about $15 million in long positions were liquidated when the price crashed. But it also means a lot of people were betting against this token without knowing how thin the float really was. When you are short a token whose supply is in ten wallets, you are not betting against the market. You are betting against the room.

ZachXBT posted on May 7. He offered a $10,000 bounty for information on the market-making activity behind LAB across Bitget, Bybit, Binance, and OKX. He asked for clues on the identity of Sadkov. He named Bitget directly and named its leadership, Shawn Liu and Gracy Chen, citing what he called a pattern, including an earlier alleged manipulation involving a token called RAVE that he said wiped out roughly $6 billion in value.

Within hours of the post, LAB's market cap fell about 42 percent. Within hours of the original peak, the token had already crashed approximately 84 percent. Reports put the wipeout at more than $250 million in value.

Bitget had not publicly responded to the specific allegations as of publication. Neither had Sadkov. Allegation is not adjudication. No charges have been filed.

IV.

By Sunday Marco had stopped opening the second phone.

He left it in the locker. He worked his shift. He came home and ate dinner with his daughter and did not check the price. He told himself he was going to wait it out. That is what the people in the channels were saying. Wait it out. The fundamentals are strong. The AI terminal is real. The market overreacted. The whales will defend the level.

There were no whales defending the level. The whales were the ones who sold.

I have watched this exact movie maybe forty times in forty years, under forty different names. In a Chicago metals room in the late 80s I watched a guy named Tony walk a small blimp company's stock up an entire afternoon and dump the whole inventory at 3:55 p.m. The blimp company did not make blimps. It made stock certificates. The certificates and the tokens are the same product. The chart is the pitch. The float is the trap.

What is different now is the speed. What is different now is the geography. The room in 1988 was a hundred desks in the Loop. The room in 2026 is a Telegram channel, a YouTube thumbnail, an exchange in a jurisdiction that may or may not answer the phone, and a wallet cluster that anyone can see if they know where to look.

Marco did not know where to look. Marco was looking at the candle.

V.

Here is the renaming, and it is the part I want you to keep.

A 95 percent insider hold is not a token. It is a position with marketing. The chart is not a market. It is the inside of a single trade walked across many fake trades. The exchange is not a venue. It is the door the position walks through on the way to your account. The Telegram rocket emoji is not a community. It is a man with a ring light who got paid to say a word.

LAB may or may not become a court case. Vova Sadkov may or may not surface. Bitget may or may not answer. Those cases, when they exist at all, often take years and rarely return the money. What you can do today is read the structure. Look for the float. Look for the wallet cluster. Look for the news that does not exist behind the price that moves. Look for the room.

Marco still has the second phone. He has not deleted the app. He told himself last week that he is going to be smarter next time, that the next one he will check the on-chain data, that the next one he will wait for an actual product. He believes this. I believed it too, after the blimp.

The machine is not the token. The machine is the room with most of the chairs already taken.

The empty chair is what they sold to Marco. He thought he was buying a seat at the table.

He was the seat.