Skip to content
Crypto.world

The Ledger

Bitquery Says One Bot Drove 22.7% of Raydium CLMM Volume

One Solana wallet generated $1.40 billion of Raydium CLMM’s $6.17 billion in trading volume over 30 days by repeatedly swapping USDC for fake tokens and back. Bitquery’s investigation says the pools held about $19 in USDC each and found no sale by the wallet cluster to an outsider at any price.

Bitquery Says One Bot Drove 22.7% of Raydium CLMM Volume

One Solana wallet accounted for 22.7% of Raydium CLMM’s trading volume over 30 days, according to an investigation Bitquery published on October 21. The wallet generated $1.40 billion of the protocol’s $6.17 billion total by repeatedly trading tokens against USDC and reversing the trades, Bitquery says. The activity creates real USDC-denominated volume, but the investigation describes the tokens in the pools as fake and the trading as wash trading.

Bitquery traced the wallet’s activity across 438 Raydium concentrated-liquidity pools between September 21 and October 20. Its figures value each swap at the lower of the two priced sides, which was the USDC side in these pools, rather than using the displayed prices of the tokens. The company says that method prevents inflated fake-token prices from inflating its volume estimate.

A $19 USDC reserve behind a $579 million display

One pool illustrates the gap between displayed liquidity and the assets available to trade. DEXScreener showed the fake LSK/USDC pool with about $579.3 million in liquidity, while Bitquery’s records showed 19.27 USDC and nearly one billion tokens named LSK in the pool. Lisk has no official Solana token, according to the investigation.

The displayed figure comes from valuing both sides of the pool at its own quoted price. With nearly a billion fake tokens priced close to the real Lisk market price, the screen’s arithmetic suggests hundreds of millions of dollars in liquidity, even though the USDC reserve is less than $20. Bitquery says its Trading API measures the quote-side reserve behind the price instead.

One round trip costs about $1.13

In a transaction example from October 20, the wallet put 11,102 USDC into the LSK pool and received 19,155.59 LSK. It sold the tokens back 22 seconds later for 11,100.87 USDC, leaving the pool with its original 19.27 USDC reserve. Bitquery calculates that the round trip cost $1.13, plus two network fees of about 0.00002 SOL each, while recording roughly $22,200 in volume.

Across the 30-day period, the wallet made 108,871 swaps and completed 22,685 round trips. Its median time between a buy and matching sell was 20 seconds, and Bitquery estimates it used about 12,000 USDC and 1.3 SOL as working capital. The report says the wallet typically carried out about 3,000 swaps a day.

Fake majors fill 438 pools

The pools used tickers associated with assets including XMR, LTC, XRP, HBAR and TAO, as well as stablecoin names such as TRC20-USDT and ERC20-USDT. Among 175 fake tokens whose tickers matched a Binance market, Bitquery found only 20 priced within 5% of the matching asset; the fake XMR token was priced at $547 against a reported Binance spot price of $119. The investigation also traced shared gas funding and transfers among trading, token-creation and helper wallets, but did not identify who controlled the cluster.

Bitquery found that four wallets outside the cluster traded one of the pools in the week covered, for about $27 combined. It says it found no sale by the cluster to an outsider at any price and no outsider purchase at size. The report describes the cluster as an inference based on shared pools, funding and direct wallet transfers—not a confirmed identification of an operator.

Bitquery’s investigation states:

“It does not yet show anyone taking it, and we did not find a sale by the cluster to an outsider at any price.”

Bitquery investigation

Bitquery’s 22.7% estimate covers swaps on Raydium’s CLMM program from September 21 through October 20, valued at the lower side of each swap. The firm says the pools and wallets extend beyond that measurement window, while the volume comparison uses the same 30-day period for the full program.

Sources

  1. Original source Bitquery bitquery.io

Related stories

Explore

More articles