Everything

Crypto trading bot borrows $200M for a $3 gain

1 Mins read



A crypto trading bot programmed to perform arbitrage trades made various complex moves within the Ethereum blockchain — including taking a $200 million flash loan — to secure a mere $3.24 profit. 

On June 14, blockchain analysis firm Arkham Intelligence shared a breakdown of the bot’s movements. According to the firm, the transaction was made by an arbitrage bot that uses flash loans.

The analysis firm explained that the reason the Dai (DAI) amount borrowed was high is that MakerDAO’s “DssFlash” contract allows zero-fee borrowing on any amount of DAI, with a limit of $500 million. This means that the bot can make uncollateralized loans as long as the assets loaned will be paid within the same block.

According to Arkham, the bot borrowed 200 million DAI from the decentralized finance (DeFi) protocol MakerDAO and supplied the funds to the Aave DAI market. Following this, 1.349 Wrapped Ether (WETH) was borrowed against the funds. The WETH was then used to purchase Threshold Network (T) tokens on the Curve Finance exchange and sold on the liquidity protocol, Balancer.

With these transactions, the bot gained a total of 0.019 Ether (ETH), valued at around $33 at the time of the trade. However, the transaction fees for the trade were roughly $28.76, with an additional $1 sent to the block builder. This left the bot with only $3.24 in profit.

Related: Sandwich trading bots lose bread and butter in $25M exploit

While the risk seemed high compared to the amount gained, a community member applauded the bot saying that “profit is profit.“ One Twitter user said that doing all this for a small profit shows how bad the bear market is.

Meanwhile, not all bots are working on low-profit trades. On April 20, a bot operator gained over $1 million using sandwich attacks on memecoin traders. Most of the profits came from targeting trading activity from memecoins like Pepe coin (PEPE) and Wojak (WOJAK).

Magazine: ‘Moral responsibility’ — Can blockchain really improve trust in AI?