Author: Erik Johansson | MEV Researcher | Former Flashbots Contributor

Every day, roughly $2-5 million is extracted from Ethereum users through MEV. Most people have no idea it's happening. They just notice their swap gave them less than expected, or their transaction took longer than it should. Welcome to the invisible tax of blockchain.

What is MEV, really?

MEV stands for Maximal Extractable Value. In plain terms: the profit that block producers can make by manipulating the order of transactions in a block.

Here's a simple example. You submit a swap on Uniswap — buying $10,000 worth of tokens. A bot sees your pending transaction in the mempool. It front-runs you, buying the same token first, pushing the price up. Your transaction executes at the higher price. The bot then sells immediately after, pocketing the difference.

You paid more. The bot profited. This happens thousands of times per day.

Sandwich attacks are even worse. The bot places one transaction before yours and one after — squeezing profit from both sides. I've seen single sandwiches extract $50,000 from unsuspecting users.

The mempool problem

The root issue is transparency. When you broadcast a transaction, it sits in the public mempool waiting to be included in a block. Everyone can see it. Searchers — the bots hunting for MEV — monitor this pool constantly, looking for profitable opportunities.

This creates a dark forest. Every transaction is prey. The mempool becomes a battlefield where milliseconds matter and ordinary users always lose.

Some argue this is just arbitrage — market efficiency in action. I disagree. There's a difference between correcting price discrepancies across venues and literally stealing value from users by manipulating transaction ordering. One improves markets. The other is extraction.

Private mempools: Solution or new problem?

The industry's response has been private mempools. Instead of broadcasting your transaction publicly, you send it directly to a block builder who promises not to exploit it.

Flashbots Protect is the most popular example. You submit transactions through their RPC, and they route it to builders who commit to not front-running. Your swap executes without being sandwiched.

Sounds good, right? It is — for individual users. But zoom out and the picture gets complicated.

Private mempools concentrate power. A handful of builders now produce most Ethereum blocks. They see transaction flow that others don't. They have information advantages that compound over time.

We solved one centralization problem by creating another. The cure might be as dangerous as the disease.

The current landscape

As of early 2025, roughly 70% of Ethereum blocks come through MEV-Boost — the system that connects validators to builders. A small number of builders dominate: Flashbots, BeaverBuild, Titan, and a few others control the majority of block production.

This isn't what decentralization looks like.

Order flow auctions have emerged as another model. Instead of hiding transactions, you auction the right to execute them. MEV still gets extracted, but some value flows back to users. Platforms like MEV Blocker and BackRunMe offer this approach.

The economics are better for users, but the centralization concerns remain.

What actually fixes this?

I've spent three years researching MEV. Here's my honest take on solutions.

Encrypted mempools could help. Transactions stay encrypted until block ordering is finalized, so no one can front-run what they can't see. Shutter Network and similar projects are building this. The challenge is latency — encryption adds overhead that traders hate.

Fair ordering protocols attempt to guarantee first-come-first-served execution. Chainlink's Fair Sequencing Services is the most prominent effort. Technically complex, but promising.

Application-level design matters too. DEXs can implement batch auctions, where all orders in a time window execute at the same price. No ordering advantage means no MEV. CoW Protocol does this effectively.

Ultimately, we need multiple approaches. No single solution handles every MEV type. The goal isn't eliminating extraction entirely — some MEV, like arbitrage, genuinely improves markets. The goal is preventing predatory extraction that harms users.

Why this matters beyond crypto

MEV isn't just a blockchain problem. It's a preview of algorithmic markets everywhere. As AI trading systems proliferate, the same dynamics will appear in traditional finance — faster, smarter bots extracting value from slower participants.

The solutions we build here will matter beyond crypto. Fair ordering, encrypted execution, value redistribution — these concepts apply anywhere algorithms compete.

We're not just fighting for fairer DEX swaps. We're establishing precedents for algorithmic fairness itself.

Erik Johansson is an independent MEV researcher based in Stockholm. He previously contributed to Flashbots and advises protocols on MEV mitigation strategies.