Sandwich attacks are a symptom of a broken transaction ordering system. The public mempool broadcasts user intent, creating a predictable profit opportunity for searchers using MEV bots.
mev-the-hidden-tax-of-crypto
Blog
Why Sandwich Attacks Are a Symptom, Not the Disease
Sandwich attacks are not the core problem; they are the visible symptom of a deeper architectural flaw: the transparent, first-price auction mempool. The real disease is naive transaction exposure, and the cure is a shift to private order flow and intent-based architectures.
introduction
THE DATA
Introduction: The $1 Billion Symptom
Sandwich attacks are a $1B+ annual tax on DeFi, revealing a fundamental flaw in the public mempool's design.
The real disease is transparency. Protocols like Uniswap and Curve operate on-chain, but their pre-execution state is fully visible. This creates a predictable profit opportunity for any actor who can front-run a trade.
This is not a DEX-specific flaw. It is a universal mempool problem that impacts every transaction with slippage. The $1B+ extracted annually proves the economic incentive to exploit this structural weakness.
Evidence: Flashbots' mevboost.pics data shows searchers extracted over $120M from Ethereum alone in Q1 2024, with sandwich attacks constituting a dominant share. This is a systemic cost borne by end-users.
key-insights
SYSTEMIC FRAGILITY
Executive Summary: The CTO's Diagnosis
Sandwich attacks are not a bug of individual protocols but a structural failure of the public mempool model, exposing a fundamental misalignment between user intent and execution.
01
The Public Mempool is a Toxic Asset
Broadcasting a plaintext transaction to thousands of nodes is an invitation for frontrunning. This design, inherited from Bitcoin, is fundamentally incompatible with a multi-billion dollar DeFi ecosystem where latency is profit.
- Creates a zero-sum game where user slippage becomes MEV revenue.
- Forces protocols like Uniswap to operate on a hostile network layer.
- Centralizes block building as searchers and builders consolidate to win the latency war.
$1B+
Annual MEV Extracted
~12s
Public Exposure Window
02
Intent-Based Architectures (UniswapX, CowSwap)
The solution is to separate declaration of intent from on-chain execution. Users sign a desired outcome, not a specific transaction, delegating pathfinding and execution to a competitive network of solvers.
- Eliminates frontrunning surface by removing the public transaction.
- Introduces off-chain competition among solvers, improving price execution.
- Enables cross-chain swaps natively, as seen with Across and LayerZero integrations.
~100%
Attack Reduction
Best Execution
Guarantee
03
Private Transaction Channels
For transactions that must follow a specific path, direct submission to builders via private RPCs (e.g., Flashbots Protect, bloXroute) is the current pragmatic fix. This bypasses the public mempool entirely.
- Immediate mitigation for existing dApps without a full architectural overhaul.
- Shifts trust from the network to a few relay/block builder entities.
- Highlights the urgent need for PBS (Proposer-Builder Separation) and SUAVE.
~500ms
Submission Latency
>90%
Of Ethereum Blocks
04
The Endgame: Encrypted Mempools & SUAVE
The final cure requires cryptographically enforced privacy at the protocol level. Encrypted mempools and shared sequencing layers like SUAVE aim to neutralize information asymmetry by design.
- Encrypts transactions until block inclusion, preventing frontrunning.
- Decentralizes block building by creating a transparent marketplace for execution.
- Turns MEV from a rent extracted from users into a commodity competed over by builders.
Protocol Level
Solution
Commoditized MEV
End State
thesis-statement
THE MEMPOOL PROBLEM
The Core Disease: Naive Transaction Exposure
Sandwich attacks are merely the most visible symptom of a foundational flaw in how blockchains broadcast and order transactions.
The root cause is public mempools. Every transaction is broadcast in plaintext before inclusion, creating a free-for-all for MEV extraction. This is a protocol-level design flaw inherited from Bitcoin and Ethereum's early architecture.
Sandwich bots are arbitrageurs. They are a symptom of predictable, naive execution. The disease is the lack of execution privacy, which turns every user's intent into a public signal for extractive bots.
Compare intent-based systems. Protocols like UniswapX and CowSwap solve this by hiding orders and batching settlements off-chain, removing the mempool exposure vector entirely.
Evidence: Over $1.2B in MEV was extracted from Ethereum in 2023, with sandwich attacks comprising a dominant share. This is a direct tax levied by the public transaction broadcast model.
SYMPTOM VS. DISEASE
The Symptom's Toll: Quantifying Sandwich MEV
Comparing the direct costs of sandwich attacks against the underlying systemic failures that enable them.
| Metric / Feature | The Symptom (Sandwich Attack) | The Disease (Systemic Failure) | The Ideal State (No MEV) |
|---|---|---|---|
Extractable Value per Attack (Avg.) | $500 - $5,000+ | N/A (Enabler) | $0 |
Victim Slippage Incurred | 2x - 10x expected rate | N/A (Enabler) | 1x expected rate |
Primary Enabler | Public Mempool Visibility | ✅ Transparent Block Building | ❌ Encrypted/Private Order Flow |
Network-Level Annual Drain (Est.) | $1B+ (Ethereum) | Pervasive across L1s/L2s | $0 |
User Experience Impact | Failed trades, worse prices | Constant latency optimization tax | Predictable execution |
Solver/Builder Profit Source | Arbitraging user mispricing | Order flow auction (OFA) payments | Protocol fees only |
Mitigation Example | Frontrunning protection (e.g., 1inch) | SUAVE, MEV-Share, MEV-Boost | Fully encrypted mempools |
deep-dive
THE ROOT CAUSE
Architectural Cures: Beyond Band-Aid Solutions
Sandwich attacks are a symptom of the public mempool's fundamental design flaw.
Public mempools are the vulnerability. The core issue is not the bots, but the transparent transaction ordering that allows them to operate. Every pending trade on Uniswap or Curve is visible before execution, creating a predictable profit opportunity.
Private order flow is the cure. Solutions like Flashbots Protect, CoW Swap's batch auctions, and UniswapX's off-chain intent settlement eliminate the public signal. They separate transaction submission from execution, removing the front-running surface.
The future is intent-based architecture. Protocols like Across and 1inch Fusion process user intents off-chain, matching them directly with solvers. This shifts the paradigm from broadcasting a vulnerable transaction to privately expressing a desired outcome.
Evidence: After Flashbots' adoption, Ethereum's proposer-builder separation (PBS) made over 90% of MEV extraction permissioned and competitive, proving that architectural change, not just detection, mitigates the problem.
protocol-spotlight
DISMANTLING MEV EXTRACTION
Protocol Spotlight: Building the Antidote
Sandwich attacks are just the most visible symptom of a deeper disease: opaque, extractive transaction ordering. The real cure is redesigning the system.
01
The Problem: Opaque, Centralized Order Flow
Traders blindly send transactions to public mempools, creating a predictable feast for searchers and validators. This is a structural subsidy paid by users to block producers.
- ~$1B+ extracted annually via MEV.
- Creates network congestion and unpredictable fees.
- Centralizes power with a few block builders.
$1B+
Annual Extract
>80%
Builder Market Share
02
The Solution: Commit-Reveal & Encrypted Mempools
Hide transaction intent until it's too late to front-run. Protocols like Shutter Network and EigenLayer's MEV Blocker use threshold encryption.
- Zero information leakage before block inclusion.
- Renders naive sandwich bots ineffective.
- Preserves composability and decentralization.
~0s
Info Leakage
TEE/MPC
Tech Stack
03
The Solution: Intent-Based Architectures
Don't submit a transaction, declare a goal. Let specialized solvers (like in UniswapX, CowSwap, 1inch Fusion) compete to fulfill it optimally.
- User gets best execution, not first execution.
- Solvers internalize MEV for user benefit.
- Naturally aggregates liquidity across Uniswap, Curve, Balancer.
10-100x
More Liquidity
Better Price
Guarantee
04
The Solution: Proposer-Builder Separation (PBS)
Formally separate who builds blocks from who proposes them. Ethereum's PBS roadmap and Solana's Jito create a competitive builder market.
- Transparent auction for block space.
- MEV revenue is democratized and can be redistributed.
- Reduces validator centralization incentives.
>95%
Efficiency
Redistributed
MEV Revenue
05
The Solution: SUAVE - A Universal MEV Marketplace
A dedicated chain for decentralized block building and order flow. Flashbots' SUAVE aims to be the preference layer for all chains.
- Decentralized sequencer network.
- Cross-chain MEV opportunities unlocked.
- Users retain control over their order flow.
Cross-Chain
Scope
Decentralized
Execution
06
The Outcome: MEV as a Feature, Not a Bug
When these systems mature, extracted value is minimized and recaptured. The endpoint is fair ordering and optimal execution as a public good.
- Predictable finality and costs.
- Level playing field for all participants.
- Enables new DeFi primitives reliant on fairness.
~0
Sandwich Loss
Max Value
To User
counter-argument
THE CORE PROBLEM
Steelman: Is Privacy Just a New Cartel?
Sandwich attacks are a symptom of a deeper disease: public mempools create a zero-sum, extractive market for transaction ordering.
Public mempools are the disease. They broadcast pending transactions, creating a predictable, monetizable attack surface. This is a market structure failure, not a security bug. Protocols like Flashbots Protect and CoW Swap exist solely to bypass this toxic environment.
Privacy enables new cartels. Private transaction channels (e.g., Flashbots SUAVE, EigenLayer) shift power from public searchers to private order-flow auctioneers. This doesn't eliminate rent-seeking; it centralizes it into a few opaque, permissioned systems.
The solution is credible neutrality. The goal is not privacy for its own sake, but fair ordering. Networks like Espresso Systems and Astria are building decentralized sequencers to make ordering a public good, not a private revenue stream.
Evidence: Over 90% of Ethereum blocks are built by centralized builders (e.g., beaverbuild, rsync), proving that private order flow already dominates. The fight is over who controls and profits from it.
FREQUENTLY ASKED QUESTIONS
FAQ: For Architects and Builders
Common questions about why sandwich attacks are a symptom of deeper market structure flaws, not the core disease.
Sandwich attacks are a symptom of predictable, on-chain transaction ordering in a public mempool. The root disease is the lack of a private, fair, and efficient execution layer. Protocols like Flashbots SUAVE and CoW Protocol aim to solve this by separating transaction ordering from block production and enabling batch auctions.
takeaways
ARCHITECTURAL IMPERATIVES
Takeaways: Prescription for the Future
Sandwich attacks are a market structure failure; fixing them requires redesigning the core trading stack.
01
The Problem: Public Mempools Are Toxic
Broadcasting transactions publicly is an invitation for frontrunning. This is the root cause of sandwich attacks and MEV extraction, costing users ~$1B+ annually in lost value.
- Real-time auction for your trade's slippage.
- Creates a negative-sum game for end users.
- Forces protocols to build complex, reactive defenses.
$1B+
Annual Loss
100%
Visibility
02
The Solution: Encrypted Mempools & SUAVE
Hide transaction intent until execution. Flashbots' SUAVE and similar architectures (e.g., Shutter Network) encrypt orders, breaking the MEV supply chain.
- Removes the profitable information asymmetry.
- Enables fair ordering through threshold decryption.
- Turns MEV from an extractive tax into a public good via redistribution.
0ms
Frontrun Window
Sealed
Bid Auction
03
The Problem: AMMs Are Predictable
Constant function market makers like Uniswap V2/V3 have deterministic price impact. Bots can perfectly calculate profitable sandwich parameters before you trade.
- On-chain liquidity is a public price oracle for attackers.
- Passive LPs subsidize the attack via arbitrage profits.
- Creates a structural vulnerability in DeFi's core primitive.
100%
Predictable
LP Loss
Subsidy
04
The Solution: Proactive Liquidity & CowSwap
Move liquidity off the predictable AMM curve. CowSwap (Coincidence of Wants) and UniswapX use batch auctions with off-chain solvers.
- No on-chain pre-execution liquidity to frontrun.
- Solvers compete to provide best execution, internalizing MEV.
- Results in better prices for users and no sandwich risk.
~$2B+
Traded Volume
Net Positive
User Price
05
The Problem: User UX is Hostile
Expecting users to manage RPC endpoints, custom gas, and slippage is absurd. The default experience is optimized for bots, not humans.
- Slippage tolerance is a blunt, user-set parameter for attack profitability.
- Wallet defaults are often insecure or extractive.
- Shifts security burden to the least sophisticated participant.
User Error
Primary Risk
Bot-Optimized
Default State
06
The Solution: Intent-Based Architectures
Users declare what they want (e.g., "swap X for Y at best price"), not how to do it. Protocols like UniswapX, Across, and layerzero's OFT standard abstract execution.
- Removes complex transaction parameters.
- Solver networks compete on fulfillment, baking in protection.
- Represents the endgame for user-facing DeFi.
Declarative
User Action
Solver Competition
Execution
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall