# MEV Sandwich Attacks by Chain: Monthly Losses

> A monthly report on MEV sandwich attacks by chain: attacks found from swap order, value taken from traders, bot profits, the pools hit most and what changed.

Markdown copy of https://voltfincrypto.com/apps/mev-sandwich-attacks-by-chain

**Question:** How many sandwich attacks happened this month on each chain, and how much value did they take from traders?

- Kind: Report in Voltfin Folio
- Topic: [DeFi](https://voltfincrypto.com/topics/defi)
- Chains: Ethereum, Solana, BNB Chain, Base, Arbitrum
- Status: Planned. The data is not published yet, so this page states no figure; each figure will open its receipt when it is.

A sandwich attack wraps a victim's swap between two trades by the same bot: a buy just before it that pushes the price up, and a sell just after it that takes the difference. This report counts those attacks each month on each covered chain, from the order of swaps inside each block, and states how much value moved from traders to bots.

Sandwiching depends on who can see a swap before it lands. Public mempools, private order flow, sequencer design and validator behaviour differ from chain to chain, so the same detection method gives very different results. Each edition compares the chains on one definition and shows which pools, tokens and bots drove the month.

For a single swap, the sandwich attack checker answers whether it was hit. This report is the monthly record across chains, dated so that reclassified attacks appear as corrections in the next edition.

## Blueprint

### The month in one paragraph

How many sandwiches were found, the value taken from traders and how the month compared with the last, chain by chain.

Cites: Sandwich attacks this month; Value taken from traders; Change from last month.

### Sandwiches by chain

Attacks, victims and value taken on each covered chain, with the share of each chain's DEX volume that was sandwiched.

Cites: Attacks per chain; Value taken per chain; Share of DEX volume sandwiched.

### The pools and tokens hit hardest

The pools with the most attacks and the largest losses, and how much of the month's loss came from new or thinly traded tokens.

Cites: Most sandwiched pools; Losses on new tokens.

### Who ran the sandwiches

The bot contracts behind the month's attacks, their profit after costs and what they paid builders or validators to be ordered.

Cites: Profit of the largest bots; Paid to block producers.

### What victims had in common

Trade sizes, slippage limits and routes of sandwiched swaps, and how many of them went through private transaction routes.

Cites: Median loss per victim; Victims with a wide slippage limit.

### Method and corrections

The detection rules this edition used, attacks reclassified since the last edition and the known gaps on each chain.

## How it will be measured

- A sandwich is a front-run and a back-run by the same bot, in opposite directions in the same pool, around a victim's swap in that pool inside one block; on Solana the search also spans consecutive slots of one leader.
- The victim's loss is what the swap would have received at the pool state before the front-run minus what it did receive, valued at oracle prices at that block.
- Bot profit is the back-run's output minus the front-run's input, net of gas, priority fees and any tips paid to builders or validators.
- Counts and amounts cover the calendar month in UTC and each attack links to its transactions; chains are compared only on this shared definition.

## Data it needs

- `mev_daily` (to build): Blocks by builder, MEV-Boost relay share, and detected sandwich attacks with the value taken from users, per day.
- `dex_swaps_daily` (to build): Swap counts and volume per DEX, pool and chain per day, decoded from Swap events, valued with stablecoin legs and oracle prices.
- `address_labels` (to build): Labels for known addresses: exchanges, bridges, protocols, funds, treasuries, sanctioned addresses and known attackers.
- `native_prices_daily` (to build): Reference prices of major assets read from on-chain oracle feeds (Chainlink latestRoundData) at each day's last block, with each round's update time.
- Source: DEX Swap logs in each block, ordered by transaction index and log index
- Source: Solana DEX program swap instructions, ordered by slot and position
- Source: Pool reserves or tick state before each front-run, to replay the victim's swap
- Source: Gas, priority fees and builder or validator tips paid by the bot transactions

## Questions

### What is a sandwich attack?

A bot sees a pending swap, buys the same token just before it so the victim pays a worse price, then sells right after the victim's trade. The victim's loss is capped by the slippage limit it set, and the bot keeps the difference minus its costs.

### Which chain has the most sandwich attacks?

It depends on how transactions reach block producers. Chains with a public mempool, or with validators willing to reorder trades, see more, while sequencers that keep pending transactions private see fewer. The report ranks the chains each month on one detection method rather than assuming.

### How can traders reduce the risk of being sandwiched?

A tight slippage limit caps what a sandwich can take, and private transaction routes keep a swap out of public view until it is included. Splitting a large swap and avoiding thin pools also shrink the gain an attacker can make.

### Why do sandwich estimates differ between sources?

Methods differ. Some count only attacks inside one block while others include attacks spread over several slots, and losses can be measured against the pre-attack price or the victim's limit. This report states its method and links every counted attack to its transactions.

### Is every front-run a sandwich?

No. Arbitrage that follows a trade, or a liquidation that races others, is MEV but takes nothing extra from one victim's swap. The report counts only the pattern of a front-run and a back-run by the same bot around a victim in one pool.

## Glossary

- **MEV**: Maximal extractable value: the profit a block producer, or bots that pay it, can make by including, excluding or ordering transactions.
- **Slippage limit**: The worst price a trader accepts for a swap. A sandwich can move the price no further than this limit without making the swap fail.
- **Private order flow**: Transactions sent straight to builders or validators instead of a public mempool, so other bots cannot see them first.

## Related apps

- [Sandwich Attack Checker: Was My Swap Front-Run?](https://voltfincrypto.com/apps/sandwich-attack-checker)
- [Ethereum Block Builders: Market Share of Blocks](https://voltfincrypto.com/apps/ethereum-block-builders)
- [DEX Volume by Chain and Protocol](https://voltfincrypto.com/apps/dex-volume-by-chain)
- [Solana REV: Base Fees, Priority Fees and Jito Tips](https://voltfincrypto.com/apps/solana-rev-jito-tips)
- [DeFi Monthly Report: TVL, DEX Volume, Lending and Perps](https://voltfincrypto.com/apps/defi-monthly-report)
- [DeFi TVL by Chain: Total Value Locked Rankings](https://voltfincrypto.com/apps/defi-tvl-by-chain)

Build it in Voltfin Studio: https://app.voltfincrypto.com/signup?next=%2F