The court took until September. LIBRA's $57.6M was swapped to SOL and gone by June.

01 · The rulingDismissed, 104 days too late to matter

The LIBRA lawsuit ended on September 29, 2026. A federal judge in New York threw it out for good. The investors who sued cannot bring the same claims again.

LIBRA is the Solana memecoin that Javier Milei, Argentina's president, posted about in February 2025. It climbed to a value in the billions and fell by more than 90% within hours. Investors sued Hayden Davis, his company Kelsier Ventures and Benjamin Chow, the former chief of Meteora, the exchange the token launched on. They said insiders had taken the money out. The defendants denied wrongdoing.

Most coverage of the ruling treats it as the end of the story. On-chain, the story ended earlier. In 2025 the same court froze two Solana wallets holding USDC, then released them. Bitquery indexes every transfer on Solana, so we followed those two wallets from the day their money first moved to the day of the ruling.

By then the money was long gone. It was swapped into SOL, sold again, and sent off the visible trail through two exchanges and a bridge. The last transfer came in June.

02 · The courtWhat the judge decided, and what she did not

The case is Hurlock v. Kelsier Ventures. It covered LIBRA and a second Meteora token, M3M3. Judge Jennifer Rochon's opinion runs to 81 pages, and it turns on law rather than on what happened at the launch.

The plaintiffs had sued under RICO, a racketeering law that needs a pattern of conduct over time. The judge found that about six months of launches was too short. She also found that Meteora could not be sued as an entity, and that the complaint did not show Chow meant to defraud anyone. Reports of the ruling are in Protos and The Crypto Times, both published on October 1.

The case ended before any evidence was exchanged. So the court never ruled on whether insiders took money out of LIBRA, or how much.

A separate inquiry in Argentina is still open. In July 2026 the judge running it, Marcelo Martínez de Giorgi, ordered 25 wallets tied to LIBRA identified and frozen at Binance, Bybit, OKX and Bitfinex. Press reports at the time said the exchanges had not yet acted on it. We do not know whether any wallet we traced is among them.

03 · The swapNinety minutes out of USDC

The freeze in 2025 worked because of what the money was. USDC has an issuer, Circle, and Circle can lock the USDC at any address when a court tells it to. We have counted how often stablecoin issuers use that power. SOL has no issuer, and nobody can lock it that way.

For 91 days after the freeze was lifted, the two wallets did nothing. Then, on November 18, 2025, both of them swapped all of their USDC into SOL. Onchain Lens reported it on X that day, citing Nansen data: "2 of their wallets bought 456,401 $SOL for $61.59M $USDC at a price of $135." The first wallet is labelled Libra Deployer. It had also pulled a few million dollars more out of LIBRA's own trading pool minutes earlier, so the two wallets swapped more than the court had frozen.

That works out to about $135 per SOL. From that morning on, the money was in a form no issuer could freeze. Published coverage of these wallets stops here.

04 · The saleSold for less than it cost

The SOL did not stay where it was bought. The first wallet passed its share to a new address half an hour after the swap. A week later, on November 25, that address split it between two more wallets. Nine minutes after that, the second frozen wallet sent its whole share to a new address of its own.

We call these three the selling wallets, because selling is what they did. The two on the first branch sold their SOL for USDC between November 27 and December 4. The one on the second branch sold in three bursts: December 2025, one afternoon in May 2026, and two weeks in June. It sold on-chain, against trading pools and market makers. Its largest source of stablecoins is an address our address labels mark as a Wintermute liquidity bot.

SOL fell while the second branch waited. The result was a loss.

The first branch held SOL for under three weeks and came out about a million dollars ahead. The second held some of it for seven months and came out six million behind. Together they turned $61.6 million into $56.6 million, a loss of about 8%. The third selling wallet now holds no SOL and no stablecoins. Everything it received has been sent on.

05 · The exitsThree ways out

Selling the SOL was only a step. The stablecoins moved on from the selling wallets too. They left by three routes, and both branches used all three.

Binance. In early December four pass-through wallets took between five and ten million dollars each from the selling wallets and split it between two addresses. One of the two forwards everything it receives to a Binance hot wallet, the address an exchange pays withdrawals from. That is how a deposit address behaves: it is the address an exchange gives one customer to pay into. Binance lists that hot wallet in its proof of reserves.

Bybit. The other address paid most of what it received straight into a Bybit hot wallet, which Bybit also lists in its proof of reserves. About a fifth of its payments went to a busy address we could not identify.

deBridge. The rest never touched an exchange on Solana. It went out in chunks of about half a million dollars, each to a wallet that had never been used. A few minutes later each wallet paid the full amount to deBridge, a bridge that moves funds to other blockchains, less a fee of 0.04%. Then the wallet went quiet for good. There were 54 of them. Our labels mark the receiving address as deBridge. Bubblemaps named the same address in a post on X in January 2025, writing that wallets "got SOL from deBridge Finance" and giving this address.

One detail about the two exchange-bound addresses matters. In the same week they also took about $17 million from four wallets we could not connect to LIBRA. So they may belong to a broker who handles money for several clients, and not to the people who held the frozen wallets. The chain shows the payments. It does not show whose account they landed in.

06 · The timingGone before the judge ruled

Put the exits on a calendar and the ruling sits far to the right of them.

The first stablecoins went to deBridge nine days after the swap. By December 10, about three quarters of the money was out. The second branch then sat on its remaining SOL for more than four months before selling the rest in May and June.

The last transfer we can find is from June 17, 2026. Nothing in any of the wallets named here has moved since, and that includes the week of the ruling. LIBRA itself did not react either. The token trades a few hundred dollars a day, at a price that values all of it at about three million dollars.

The freeze was lifted 302 days before that last transfer. A court that wanted to reach the money again would have had to act inside that window, and against SOL it would have had no issuer to send an order to.

07 · The stakesSo what if the money is gone?

For the investors who sued, the money was the point. A lawsuit like this one ends, if it is won, in an order to pay. An order to pay is only worth what can be found.

In May 2025 all of it could be found. It sat in two wallets, in a stablecoin whose issuer answers to courts. When the judge lifted the freeze that August, she found the investors faced no harm that a payment of money could not repair later.

That finding assumed there would be something to collect from. Ninety-one days later the money was SOL. Within a month most of it had left the wallets the court knew about. By the day of the ruling, those two wallets held almost none of it.

The dismissal means there is nothing to collect in this case. The story may not end there. Argentine press reports say the investors are weighing an appeal, and prosecutors in Argentina are still at work. If either gets as far as asking where the money is, the answer used to be two addresses on Solana. Now the answer is a Binance deposit address, payments into Bybit, and a bridge.

So is it out of reach? Partly. About thirty million dollars went to exchanges, and exchanges know their customers. A court can order one to name an account holder and to hold whatever is still in the account. Whether anything is still there, ten months on, only the exchanges know. The bridge share is the hard part. Nobody can freeze it from Solana, and we cannot see where it landed.

Nothing here broke a rule. From August 2025 to the last transfer, no order we know of stopped the holders from moving the money, and no court has found that they had no right to it.

08 · The moneyWhere it is now

We can say where the trail stops. We cannot say what happened after it.

Money paid into an exchange deposit address is credited to an account. Binance knows which customer owns the address that took $15.1 million. Bybit knows who was credited with the payments into its hot wallet. From there the owner can trade it, hold it or withdraw it to a bank. None of that is visible on-chain. Stolen funds often take the same road. We followed it in the TradeWiz drain and in the Bitget hack. This money is different in one respect: no court has found that it was stolen.

The bridge is harder. deBridge releases funds on another blockchain, and the Solana side of the transfer does not show which one or to whom. The payments we looked up are not in deBridge's public order records, so we stop at the bridge. For $26.3 million, we do not know the next address.

09 · MethodHow we traced it

We started from the two wallets named in the freeze and read every transfer out of them from November 18, 2025. At each step we followed the SOL, then the stablecoins it was sold for, until the money reached an address that belongs to an exchange or a bridge. A route counts only if we followed every wallet on it. All 54 single-use wallets were checked one by one: each received USDC from a selling wallet and paid it to deBridge.

Exchange and bridge names come from our own labels. The two exchange hot wallets are also in each exchange's published proof of reserves. Dollar figures are stablecoin amounts at face value. All times are UTC.

There are limits. We do not know who controls any wallet after the two the court named. The difference between the money in and the money out includes trading costs, and the second branch bought some SOL back before selling it again. We did not trace the other insider wallets from the launch in February 2025. And we did not follow anything past the bridge.