Sandwich attacks are a form of transaction-level manipulation where an attacker inserts one transaction before and one after a target trade to capture price movement. Slippage is the difference between the expected execution price and the actual execution price; in a sandwich scenario the attacker deliberately increases slippage for the victim. Research on miner and maximal extractable value by Phil Daian Cornell Tech documents how transaction reordering and selective inclusion create predictable opportunities for value extraction and worsen execution quality for ordinary traders.
Mechanism and causes
A sandwich attacker first front-runs by buying into the same liquidity pool, pushing the asset price up. The victim’s trade then executes at the inflated price, suffering greater slippage. Finally the attacker sells into that higher price, capturing the spread. This sequence directly reduces the window in which independent arbitrageurs can profitably restore price parity between venues. Vitalik Buterin Ethereum Foundation has discussed how these forms of extractable value arise naturally from public transaction pools and the ability of block producers to reorder transactions, creating structural incentives for repeated exploitation.
Consequences for arbitrage outcomes
Slippage induced by sandwiches affects arbitrage outcomes in three connected ways. First, reduced arbitrage margins make certain cross-market trades uneconomic because the attacker has already captured the delta that arbitrageurs rely on. Second, increased execution risk raises the probability that arbitrage attempts revert when on-chain slippage protection triggers, producing wasted gas costs and adversarial loss. Third, the presence of persistent sandwich activity shifts market behavior: arbitrageurs may demand larger safety buffers, use private transaction channels, or withdraw from thinner markets, degrading liquidity and price discovery. Flashbots research team Flashbots has promoted private relay models precisely to limit harmful public reordering and to quantify these effects.
The human and cultural consequences matter: retail traders bear the immediate cost of worse prices, while professional traders and infrastructure providers adapt by centralizing flows or paying for protection, altering the decentralized ethos of many markets. Environmentally and territorially, higher rates of failed or duplicated transactions increase chain load and fees, affecting users disproportionately in regions sensitive to transaction costs. Addressing slippage from sandwich attacks requires both protocol design changes and market infrastructure that reduce incentives for reordering rather than merely shifting the problem.