Ethics & The Sandwich Question

Constructive MEV vs extractive sandwich attacks. MEV-Blocker, CowSwap, UniswapX as user-facing defenses. Jito and Solana sandwich economics. Vitalik's Endgame and trustless block-building.

25 min · expert · part of MEV Searcher Strategies

Not All MEV Is Equal

Several lessons in this module have used phrases like "constructive MEV" and "extractive MEV" interchangeably. The distinction is the central ethical question of the entire field, and it deserves direct treatment. **Constructive MEV** is value capture that comes from doing socially useful work. Arbitrage closes price gaps between markets. Liquidations keep lending protocols solvent. JIT liquidity provision (debatable, mildly extractive from LPs but slightly helpful to swappers) sits on the border. The capture mechanism is competitive — many actors race for the same opportunity — and the user who triggered the opportunity is no worse off than they would have been in a world without the searcher. Often they are better off (tighter spreads, faster price reconciliation). **Extractive MEV** is value capture that comes from harming a specific identifiable user. The canonical example is the sandwich attack: a searcher sees a user's pending DEX swap, front-runs it to push the price against the user, lets the user's swap execute at the worse price, and back-runs to capture the price they pushed. The user receives a worse fill than they would have without the searcher's interference. Pure value transfer from a less-sophisticated party to a more-sophisticated one. No social work done. Whether MEV is "good" or "bad" depends entirely on which category a given strategy falls into. A searcher who runs only arbitrage and liquidation strategies is doing economically valuable work and deserves the resulting profit. A searcher who runs sandwich bots against retail Solana users is, ethically, picking pockets at scale. The infrastructure is the same. The choice of strategy is the ethical choice. This lesson is candid about both ends and about the protocol-level work being done to make sandwich attacks harder.

Also in this lesson

How Sandwich Attacks Work
MEV-Blocker, CowSwap, UniswapX
Solana and Jito: A Different Equilibrium
Vitalik's Endgame and Trustless Block Building

Key terms

Constructive MEV: Value capture from socially useful work — arbitrage closing price gaps, liquidations preserving lending solvency, JIT liquidity (borderline). The user is no worse off than in a no-searcher world and often better off.
Extractive MEV: Value capture that comes from harming an identifiable user. The canonical example is the sandwich attack. Pure value transfer from less-sophisticated to more-sophisticated parties without social work done.
Sandwich attack: A three-transaction MEV strategy: front-run a user's pending swap to worsen its price, let the swap execute, then back-run to capture the price impact. Solana sandwich bots earned $370-500M between January 2024 and May 2025 per Eigenphi.
MEV-Blocker: A private-RPC service launched 2023 by Cow Protocol, Beaverbuild, and others. Routes user transactions privately, prevents front-running, and refunds a share of any back-run profit. Single-digit-billions monthly volume by 2025-2026.
CoW Protocol (CowSwap): An intent-based DEX where users sign trade intents and solvers compete in periodic batch auctions to find best execution. Eliminates per-transaction mempool exposure that enables sandwiches. Multi-billion monthly volume.
UniswapX: Uniswap Labs' intent-based protocol launched 2023. Users sign Dutch-auction orders; fillers compete to fulfill. No mempool transaction; no sandwich exposure. Uniswap's answer to the threat of intent-based protocols eating AMM volume.
Jito Block Engine: Solana's primary MEV bundle relay. Jito-Solana client runs on 70%+ of Solana stake by 2025. Tip flow exceeds $50-80M/month. Both enables sandwich attacks and offers private-flow protections; commercial incentives create tension.
Shutter Network: Threshold-cryptography-based encrypted mempool. Transactions are encrypted in the mempool and decrypted only after block inclusion is committed. Live on Gnosis Chain; Ethereum mainnet integration discussed for 2026-2027.
Inclusion list: Vitalik's proposed censorship-resistance backstop: proposers mandate inclusion of specific transactions even if builders excluded them. EIP-7547 was a 2024 candidate; refined design is on the Glamsterdam fork roadmap.
Endgame (Vitalik, 2021): The thesis that block production will centralize because MEV economics make it economies-of-scale, but block validation will remain decentralized with protocol-level censorship-resistance backstops. The dominant conceptual frame for current MEV research.

Continue this lesson — 4 more sections in the CryptoBipto app.

Open lesson

Educational only — not financial advice.