By - MEV is value extracted by choosing the order of transactions in a block, not by breaking anything.
- The common forms are arbitrage, liquidations and sandwich attacks. Only the last directly harms the user.
- It is a structural consequence of public mempools, and it exists on every chain with them.
MEV, or maximal extractable value, is the money that can be made purely by deciding which transactions go into a block and in what order. No exploit, no stolen key, no bug. Just sequencing.
It is one of the least visible costs in crypto and one of the largest, and most people paying it never learn that they did.
Where does the opportunity come from?
From the fact that pending transactions are public before they settle. When you submit a trade, it sits in a waiting area called the mempool, visible to anyone watching, until a validator includes it.
That gap between broadcast and confirmation is the whole opportunity. Anyone who can see your intended trade, and who influences ordering, can position around it. The mechanics of block construction are documented in Ethereum developer documentation.
What are the main forms?
Arbitrage. The same asset trades at different prices on two venues. A bot buys low, sells high, in a single transaction. This is genuinely useful; it is what keeps prices consistent across exchanges, and nobody is worse off.
Liquidations. A lending position falls below its collateral requirement and anyone can close it for a fee. Competing to be first is a race, but the liquidation was going to happen regardless. The protocol needs this to stay solvent.
Sandwich attacks. This is the one that costs you. A bot sees your pending buy, buys immediately before you, lets your order push the price up, then sells into it. You receive a worse price and the difference goes to the bot.
How much does a sandwich actually cost?
It depends almost entirely on your slippage tolerance, which is the setting most people never touch.
Slippage tolerance tells the trade how much price movement you will accept before it fails. Set it to 5% on a thin pool and you have announced that you will tolerate losing up to 5%. A bot will take close to all of it, because you authorised exactly that.
Set it to 0.5% and the same attack becomes unprofitable, because there is not enough room to extract. The trade may fail more often in volatile conditions, which is the trade-off, and it is usually the right one. This interacts directly with how gas fees work, since a failed transaction still costs a fee.
Can you avoid it?
Partly, and the methods are practical rather than technical.
Tighten slippage. Trade in deep pools, where moving the price enough to profit costs the attacker more than they gain. Split large orders. And for anything sizeable, use a private transaction relay, which submits directly to block builders without exposing the trade to a public mempool first.
Understanding pool depth matters more than any single setting, and our Uniswap review covers how automated market makers price trades against available liquidity.
Is MEV going away?
No, and any claim otherwise is worth treating carefully.
Ordering has value wherever transactions are public before they settle. Research on mitigations, including encrypted mempools and fair-ordering protocols, is active and collected in public cryptography research, but nothing has eliminated the underlying incentive. Firms such as Chainalysis now track extraction as a distinct category of on-chain flow.
What changes is who captures it and how much reaches ordinary users rather than a small set of sophisticated operators. Treat MEV as a cost of trading on a public ledger, price it into your settings, and stop leaving 5% slippage on.
Who actually receives the money?
It splits between searchers, builders and validators, and the proportions have shifted over time.
Searchers run the bots that spot opportunities and construct the profitable bundle. Builders assemble bundles into complete blocks. Validators choose which completed block to propose. Each layer takes a cut, and competition between searchers pushes most of the value up the chain toward validators, who are the scarce resource.
The practical consequence is that a meaningful share of validator income on some networks is not block rewards or ordinary fees. It is extraction. That changes the economics of running a validator, and it means the security budget of the chain is partly funded by users receiving worse prices than they expected.
Does this exist outside crypto?
Yes, and the parallel is exact enough to be useful.
Traditional markets have spent decades on the same problem under the name front-running. Seeing an order before it executes and trading ahead of it is illegal in regulated equity markets, and enforcement exists precisely because the incentive is permanent wherever order flow is visible.
Crypto rebuilt the same structure without the prohibition, then made the order book public by design. The difference is not that the behaviour is novel. It is that here it is permissionless, automated, and nobody is going to be prosecuted for it.
What's your reaction?
