What Is MEV? The Invisible Tax on Every Chain

You send a swap. It confirms. The tokens arrive, the price looks about right, and you close the tab. What you never see is that somewhere between broadcast and confirmation, another party read your trade, calculated exactly how to profit from it, and took a cut. You paid it. You just never saw the line item.

That cut has a name: MEV, or Maximal Extractable Value. It is one of the most important concepts in crypto that most casual users have never heard of, and it touches nearly every transaction you make on-chain. This piece explains what it is, why it exists, who pays it, and how to shrink your exposure.

What MEV Actually Is

MEV is the value a block producer or a searcher can extract by choosing the order, inclusion, or exclusion of transactions in a block.

Start from a fact most people never think about: transactions do not settle the instant you send them. They wait. Someone (a validator, a builder, a sequencer) collects the pending transactions and decides which ones go into the next block, and in what sequence. That sequencing is not neutral. In a world of automated markets, the order in which trades execute changes who gets what price. And whoever controls the order can insert, reorder, or drop transactions to capture the difference.

Put plainly: tx ordering is a scarce resource, and MEV is the rent it commands. Blockspace is finite, and the right to arrange that blockspace is worth money.

The term started as "Miner Extractable Value," coined in Ethereum's proof-of-work days when miners produced the blocks (the idea was formalized in the 2019 "Flash Boys 2.0" paper). After the shift to proof-of-stake the "M" was generalized to "Maximal," because the extractor is no longer just a miner; it is an entire supply chain of searchers, builders, and validators competing for the same rent.

Why It Exists

MEV is not a bug someone forgot to patch. It falls directly out of how public blockchains are built. Three properties, together, make it inevitable:

1. Execution is transparent. On most chains, pending transactions are visible before they settle, broadcast in cleartext to a public mempool where anyone can read them. Your swap announces what you want to buy, how much, and the worst price you will accept, all before it executes.
2. On-chain pricing is deterministic. An automated market maker (AMM) prices a trade with a fixed formula over its reserves. Anyone who can read the pool can compute, to the token, how much your trade will move the price. There is no guesswork, only arithmetic.
3. Someone controls ordering. A block producer, or a searcher bidding for placement inside a block, decides the sequence. That is the lever.

Combine a transparent view of your intent, a pricing curve anyone can solve, and a party who controls the order, and profit extraction stops being clever and becomes mechanical. Bots do it millions of times a day. The transparency that makes blockchains trustless is the same transparency that makes your trades targetable by predators.

The Main Forms

Not all MEV is theft. This is the part most explainers get wrong. Some MEV is the market working as intended; some is a direct tax on you. Four forms, from most benign to most extractive:

Arbitrage. Two exchanges on the network price an identical asset differently, say WETH trades slightly higher on one DEX than another. A searcher buys low, sells high, and pockets the spread. This is usually beneficial: it drags the two prices back into line, which is exactly what keeps markets across a fragmented ecosystem honest. The searcher gets paid for a service the market needs. Nobody is harmed, and prices stay accurate. Most MEV, by volume, is arbitrage.

Liquidations. When a lending position falls below its collateral threshold, protocols allow anyone to liquidate it for a bonus. Searchers compete to be first. This is largely healthy (liquidations protect lenders and keep the protocol solvent), though the competition for them is fierce and the borrower being liquidated rarely enjoys it.

Backrunning. A searcher places a transaction immediately after yours to capture a state change your trade created, for example, the arbitrage your large swap just opened up between two pools. Backrunning is often neutral-to-beneficial: it does not worsen your price, it just harvests the imbalance you left behind. Some systems even refund part of it to you.

Sandwiching and frontrunning. This is the extractive one. A searcher sees your pending swap, buys the same asset just before you (pushing the price up), lets your trade execute at that worsened price, then sells just after (the backrun). You get filled (at the worst price your slippage tolerance allowed) and the difference lands in the attacker's pocket. You were never tricked into a bad contract. You got exactly the trade you asked for, drained through the slack you authorized. This is a direct tax on ordinary users, and it is where MEV earns its reputation.

The honest summary: MEV is not synonymous with theft. Arbitrage and liquidations keep on-chain markets efficient and solvent. But sandwiching is a straight skim off the top of your swap, and it is common enough that every active trader has paid it, usually without noticing.

Where It Lives

The economics are universal: wherever ordering can be sold, someone will pay for it. What differs by chain is how much there is to extract and the machinery built to sell it. Here is the arc, from the simplest case to the most developed.

Bitcoin, the primitive case. Miners have always ordered transactions the obvious way: pack in the highest-fee-paying ones first to earn the most from each block. That is the most basic form of profiting from ordering, and on Bitcoin it is roughly where the story ends. Its transactions are mostly plain payments, with no on-chain markets to exploit, so there is little to extract beyond the fees themselves. This is exactly why MEV grew into a named phenomenon on Ethereum rather than Bitcoin: you first need programmable, deterministic markets (AMMs, lending pools) before the right to arrange transactions becomes worth far more than the fees attached to them.

Ethereum. The classic surface is the public mempool: pending transactions propagate openly, giving searchers a window to read and react. After The Merge, Ethereum adopted proposer-builder separation (PBS). Specialized builders assemble blocks (ordering transactions to maximize value) and sell them to proposers (the validators) through relays, a pipeline most people know as MEV-Boost. Searchers construct ordered bundles and bid builders for exact placement. It is a full, competitive marketplace for ordering, running under every Ethereum block.

Solana. Solana has no global mempool in Ethereum's sense. Transactions are forwarded straight to the current leader (the validator producing the slot), so there is no open pool of pending trades to snipe from. That does not eliminate MEV; it relocates it. Jito's block engine runs an off-chain auction where searchers submit bundles and bid tips for placement in the leader's block, much like Ethereum's builder market. And orderflow still leaks: searchers gain early sight of trades through bundle simulation, forwarding services, and validators willing to share what they see. Same primitive (ordering as a sellable resource), different plumbing.

The general rule across both: MEV needs (a) some pre-execution view of your trade and (b) control over the ordering around it. The chains differ only in how those two ingredients get supplied.

Who Pays, and How Much

You do. MEV is skimmed from ordinary swaps, lending positions, and mints, rarely as a visible fee, almost always as a slightly worse price than you should have gotten. That is why "invisible tax" fits: there is no receipt.

The scale is not small. On Ethereum, Flashbots' conservative tracking put cumulative extracted MEV past $1.38 billion by mid-2023, and that figure covers only a handful of protocols and undercounts by design; real totals run higher. On Solana, sandwich bots alone are estimated to have extracted somewhere between $370 million and $500 million in roughly sixteen months through mid-2025. A single Solana bot known as "arsc" reportedly pulled in around $30 million in two months in mid-2024, entirely from sandwiching. And the defensive side has its own price tag: Solana users have at times paid millions per week in Jito tips specifically to route around sandwich bots.

These numbers move and the methodologies differ, so treat them as orders of magnitude, not precise ledgers. The point stands: MEV is a nine-figure-and-climbing transfer from ordinary users and traders to a sophisticated, automated extraction industry.

How to Reduce Your Exposure

You cannot abolish MEV; it is a property of transparent chains, not a fixable defect. But as a normal user you can deny the extractive kind most of what it needs. Briefly:

• Use private orderflow. Private RPC endpoints (Flashbots Protect on Ethereum and similar services) route your transaction to builders without ever broadcasting it to the public mempool. If a searcher cannot see your swap before it executes, they cannot sandwich it. This is the single strongest lever.
• Set slippage deliberately. Your slippage tolerance is the attacker's budget: the looser it is, the more they can extract before your trade reverts. Set it per-pool and tight enough to matter, not a lazy global default you never revisit.
• Let a smart router handle big trades. A good aggregator (1inch, CoW Swap, or Jupiter on Solana) splits a large order into smaller pieces across several pools, so no single fill moves the price much. The less your trade moves the price, the less a sandwich bot can skim from it. Many of these routers also send your order through private channels by default.

Each of these attacks one of MEV's two prerequisites: visibility or ordering slack. For the full mechanism of how a sandwich is built, sized, and defended against (the arithmetic that turns your slippage into someone else's profit), read the next piece on this track, Anatomy of a Sandwich Attack.

Securing the Unseen

MEV is honest about what a blockchain really is: an ordered ledger where the right to decide the order is worth money. Some of that value keeps markets efficient. Some of it is skimmed straight from your swap. The difference between the two is not visible on your screen, which is exactly why it deserves your attention.

At 0xhades we investigate mempools and orderflow at the byte level (the way an extractor does) because that is the only vantage from which the tax you are paying becomes visible, and the only one from which a defense can be proven rather than assumed.