MEV Explained: Best, Exclusive Guide from Sandwich to PBS

Maximal Extractable Value (MEV) is the extra profit captured by those who control transaction ordering in a block. It shapes user experience, gas markets, and even protocol design. Learning how it works—and how Ethereum’s Proposer/Builder Separation (PBS) changes the game—helps you trade smarter and reason about where blockspace markets are heading.

MEV in one sentence

MEV is the profit from picking, ordering, and including transactions in a block in a way that benefits the block’s creator or their partners, often at the expense of regular users.

Why MEV exists

When a transaction hits the mempool, it’s public. Anyone can simulate outcomes and decide which other transactions to place before or after it. If a DEX swap will move the price, that knowledge itself is valuable. Block producers and specialized searchers compete to capture that value by reorganizing transactions. The outcome feels like high-frequency trading with public orderflow and no central exchange.

The classic MEV playbook: frontrun, sandwich, backrun

Three patterns dominate. They exploit public information and deterministic settlement. Consider a small trader swapping 25 ETH to a volatile token on a popular AMM. The mere presence of the order signals a price impact, and bots react within milliseconds.

Here’s what typically happens:

1. A bot detects the pending swap and submits a buy right before it at a slightly higher gas price (frontrun).
2. The trader’s swap executes at a worse price because the pool’s reserves shifted. Slip hurts them; the bot profits.
3. The bot immediately sells after the trader (backrun) to lock in the price difference, completing a “sandwich.”

On-chain traces often reveal this as three consecutive trades: bot buy, victim trade, bot sell. Tiny scenario: a user sets 1% slippage but trades during an NFT mint spike; the bot’s sandwich pushes the price just enough to skim 0.8% and exits one transaction later.

Not all MEV is equal

Some MEV degrades user outcomes; some can be neutral or even useful. The nuance matters when evaluating solutions and policy proposals.

To anchor the taxonomy, consider these common categories and what they do to users and protocols.

• Harmful: Sandwiching swaps, reordering liquidations to extract premiums, time-bandit reorgs on unstable chains.
• Neutral-ish: Backrunning price updates after an oracle posts, arbitrage that restores cross-DEX prices.
• Constructive: Liquidations that keep lending markets solvent, blockspace auctions that pay validators without chain reorgs.

The boundary isn’t always neat. A backrun can both refloat prices and profit from a user’s visible intent. That ambiguity is why builders talk about “MEV minimization” versus “MEV redistribution.”

From miners to validators: who extracts MEV?

On Ethereum, miners used to propose and build blocks. After the Merge, validators propose blocks, but specialized “builders” often assemble them. Searchers generate bundles—pre-ordered transaction sets—that express a strategy (like a sandwich). Builders compare bundles, simulate profits, and construct the most valuable block. Finally, the proposer (a validator) picks the best-paying block. This separation creates a market for blockspace, not just a queue.

Without guardrails, the same actor could propose and build, enabling censorship and opaque orderflow capture. That’s where PBS enters.

The PBS idea

Proposer/Builder Separation splits responsibility: builders compete to assemble blocks; proposers select the highest-paying bid without seeing the block’s contents in advance. On Ethereum today, this runs via MEV-Boost, an off-chain market connecting proposers to multiple relays and builders.

A simplified PBS flow looks like this in practice:

1. Searchers submit bundles to builders (often via private channels or relays).
2. Builders simulate, pack transactions, and produce a block plus a bid (the payment they’ll make to the proposer).
3. Relays attest to the block’s validity and forward bids to proposers.
4. The proposer picks the highest bid, signs the header, and the relay reveals the block.
5. The network finalizes the block the proposer chose, and the payment flows to the validator.

This reduces incentives for proposers to censor or reorder for personal gain, since they can simply accept the best economic offer. It also moves the MEV “fight” into a specialized, competitive builder market.

What actually changes with PBS?

PBS alters who sees what, and when. It introduces private communication paths, sealed bids, and separation of duties. These changes mitigate some risks but open new ones, like relay trust and builder centralization.

In short, PBS makes block supply chain-like. Each link is optimized; the network benefits from higher proposer revenue and, often, better UX via private orderflow, but it must monitor new chokepoints.

Side effects: censorship, relays, and builder power

Relays sit between builders and proposers. They protect proposers from seeing full contents pre-commitment and ensure delivery. But they are entities with policies. During regulatory stress, some relays filtered transactions, fragmenting block supply. This risk is visible and measurable, which is progress, but it’s still a risk.

Builder concentration is the other concern. Superior simulation hardware, orderflow deals, and latency edges can create winner-take-most dynamics. If three builders win 80% of blocks, they can shape inclusion norms. The counterweight is pluralism: multiple relays, builder diversity, and open-source clients plus frequent monitoring by researchers.

What users can do today

Users don’t control PBS, but they can choose how their orders hit the chain. A few practical habits reduce extractable value from your transactions.

• Use private RPCs or protect endpoints that submit your transaction directly to builders, skipping the public mempool.
• Set realistic slippage and avoid peak times; small trades during gas spikes are sandwich magnets.
• Prefer batch auction DEXs or RFQ-style trades when size is meaningful; auctions blunt sandwiches by clearing many orders at once.
• For liquidations or NFT mints, consider tools that simulate final cost and use time or price guards.
• If running infrastructure, connect to multiple relays and keep software updated to avoid stale or censored views.

Tiny example: swapping a volatile token? A private route can land your order directly in a builder’s block. The trade clears at your set price without broadcasting bait to mempool bots.

The road to enshrined PBS and beyond

Today’s PBS runs off-chain via MEV-Boost. The research path points to “enshrined PBS,” where Ethereum itself enforces separation and reduces reliance on external relays. That move would standardize interfaces, curb trust in intermediaries, and make block sourcing more neutral.

Several adjacent ideas are on the table:

To map the territory, these are the proposals and what they target.

1. Inclusion lists: Proposers publish transactions that must be included unless invalid, reducing censorship levers.
2. MEV smoothing or “MEV burn”: Spread builder payments across validators or burn a portion to blunt centralization and random luck.
3. Encrypted mempools and delay reveals: Hide order details until inclusion to prevent pre-trade simulations.
4. Orderflow auctions: Wallets or dApps sell the right to route private orderflow, ideally rebating users.
5. Cross-domain solutions: SUAVE-like coordination or shared sequencers for L2s to handle MEV across rollups with consistent rules.

Each proposal trades latency, complexity, and neutrality against user protection. Expect multiple layers: some protections enshrined in protocol, others at the wallet, RPC, and dApp level.

Micro-examples that make it concrete

Two quick sketches help ground the abstractions. First, a DEX trade: a trader uses a batch auction DEX on a Sunday evening. Their 50,000 USDC swap settles alongside dozens of orders in the same block; no single order can be sandwiched, and the clearing price lands near the oracle median.

Second, a liquidation: a borrower’s health factor flips. Multiple searchers compete to repay and seize collateral. The winning bundle lands via a builder with the highest bid to the validator. The borrower loses collateral, but the market remains solvent—a classic case of “constructive” MEV under a PBS pipeline.

Practical mindset for readers

Assume public orderflow is adversarial. Favor routes that either hide intent until inclusion or aggregate demand into auctions. If you operate infrastructure, diversify relay connections and track builder share metrics. If you build apps, adopt intents, RFQs, and batch clears where possible.

MEV isn’t going away. With PBS, it’s being compartmentalized and priced in a more open market. The challenge—and opportunity—is to push value back to users while keeping block production decentralized and credibly neutral.