Ethereum sandwich bot generated $295M, later lost $7.5M

On the Ethereum mainnet a prominent sandwich bot accumulated roughly $295 million in gross profit through repeated MEV trades and later recorded about $7.5 million in net losses after failed executions and adverse market moves.

The bot executed sandwich trades by detecting pending large swaps on decentralized exchanges, submitting a buy transaction just before the target swap to raise the price, and then selling immediately after to capture slippage. These operations relied on fast transaction submission, precise gas price bidding and, in many cases, private relays or bundled transactions to secure ordering.

The $295 million figure represents gross gains collected across many trades during the bot’s active period. The subsequent $7.5 million loss figure reflects net losses from mispriced transactions, failed backrun executions, aggressive competition for block inclusion and rapid price reversals that turned expected profits into losses.

Activity occurred on on-chain decentralized exchanges where trade settlement and ordering are visible to mempool-monitoring systems. Sandwiching is a form of MEV, the profits available to actors who can influence transaction ordering. Since Ethereum’s move to proof-of-stake, validators, block builders and MEV relays have been able to capture these ordering profits.

Ordinary traders executing large swaps experienced higher effective slippage and, at times, higher transaction costs when bots outbid user transactions for block space. Liquidity providers and other traders saw short-term price moves around sizable transactions linked to these strategies.

Exchanges and protocol teams have adopted technical measures to limit front-running and reduce harmful extraction. Changes include tighter slippage controls, transaction batching, alternative order matching, private relays and the rise of block-building services. Research and development efforts continue on protocol-level options such as fair ordering, encrypted transaction pools and auction-based inclusion to change how transaction ordering is allocated.

Operational risks cited in on-chain records include canceled or reduced-size target trades, failures of the bot’s follow-up transactions to execute in time, volatile gas costs that exceeded anticipated gains, and competition from other automated strategies and block builders.