A loss vector most guides skip, partly because it does not fit the usual villain shape. Nobody steals your keys, nothing is hacked, and the money leaves anyway. Here is the attack, step by step; the scam industry that has grown on top of it; and what actually reduces your exposure. Part of Know the Scam by SimpleSwap.

Start with an honest admission, because this entry sits slightly awkwardly in a series about fraud.

A sandwich attack is not a crypto scam in the way phishing is. Nobody deceives you, nobody obtains your seed phrase, and no contract is exploited. A bot notices what you are about to do and arranges to be standing in front of you when you do it. Everything it does is a valid transaction that anyone is permitted to make.

It is closer to a tax than a theft. But it is a tax you pay without being told; it is paid to a private party, and a genuine scam industry is built directly on top of it. So it belongs here.

The mechanic, step by step

On a public blockchain, transactions are not private the moment you sign them. They sit in the mempool, a waiting room of pending transactions visible to anyone running a node, until a block producer includes them. Bots read that room continuously.

Say you swap $50,000 of ETH for USDC on a DEX. Here is what happens in the seconds that follow.

Step 1. The bot sees your order before it executes. Your transaction is public while pending, and it contains everything a bot needs: the pool you are trading against, your size, and your slippage tolerance.

Step 2. It calculates how your trade will affect the price. In an automated market maker, price is a function of the pool’s reserves, so a large buy predictably increases the price. The bot knows exactly how much, because the formula is public.

Step 3. It front-runs you. The bot submits its own buy of the same asset, paying a higher fee so it lands before yours in the block. This pushes the price up before your trade executes.

Step 4. Your trade was executed at the worst possible price. You still receive tokens, and your wallet reports success, so nothing looks wrong. You got less than you would have had thirty seconds earlier.

Step 5. It backruns you. Immediately after, in the same block, the bot sells what it just bought at the price your own trade created. The difference is its profit, and it came out of your output.

Your transaction sat in the middle. Hence the name.

Two details make this hard to notice. The whole thing resolves inside one block, so there is no window in which anything appears wrong. And you do not have a counterfactual: you cannot see the price you would have received, only the one you got.

The damage also scales with your size, because the bigger your trade, the more it moves the pool and the more there is to extract. On a $200 swap, it is invisible. On a $200,000 swap, it is real money.

The numbers

Cointelegraph Research, using an EigenPhi dataset, analyzed more than 95,000 sandwich attacks on Ethereum between November 2024 and October 2025 and estimated annual trader losses at around $60 million. Cumulative extraction on Ethereum has been estimated at roughly $410 million.

Four findings are worth pausing on, and the first one runs counter to the alarmist version of this story.

The trend is downward. Monthly extraction fell from nearly $10 million in late 2024 to about $2.5 million by October 2025, even as monthly DEX volume climbed past $100 billion. Researchers attribute part of that decline to traders adopting MEV protection, a rare case of a defensive tool visibly working.

Frequency did not fall with it. Attacks held steady at 60,000 to 90,000 a month. What dropped was the take per attack, which now averages about $3. This is a volume business, not a series of spectacular hits.

Almost 40% of attacks targeted low-volatility pools, meaning stablecoins, wrapped assets, and liquid staking tokens, with about 12% hitting stable swaps specifically. The intuition that “I am only swapping USDC; there is no price risk” does not protect you, because the bot is not exploiting volatility. It is explained by the slippage tolerance you set casually because the pair felt safe.

It is remarkably concentrated. Roughly 70% of all attacks during that period trace back to a single operator.

Solana runs the same play at different scales, though the two headline numbers below are separate snapshots, not a single total. Over a 30-day stretch (Dec 7–Jan 5), a single bot, DeezNode’s Vpe program, executed 1.55 million sandwich transactions for a profit of 65,880 SOL, about $13.43 million, per Helius/Jito data — with an average profit per attack of about $8.67. That is one operator, not the network. A separate 30-day snapshot found 77,188 sandwich attacks against 49,247 victims, perpetrated by 203 individual attackers, with each attacker having more than 10,752.

The scam that grew on top

Here is where this becomes a proper Know the Scam entry.

Because “MEV bot” sounds technical and profitable, an entire category of fraud now uses the term as bait. The pitch arrives via DM or YouTube tutorial: deploy this MEV bot contract, fund it, and watch it earn passive income from front-running trades. You are walked through the pain of going to Remix and deploying it yourself, which makes you feel in control.

The contract does one thing. It forwards your funds to the author. One documented operation collected roughly 156 ETH, around $411,000, from retail victims.

The tell is simple and worth memorizing: real MEV infrastructure is boring, open source, and fiercely competitive. It is a business of milliseconds and infrastructure costs, run by people who do not need your capital and would not share the edge if they did. Anything flashy and confident that invites you to fund a contract is the scam version.

What actually reduces exposure

Ranked roughly by how much they help.

Use a private transaction relay for DEX trades. Services such as Flashbots Protect route your transaction to block builders without it passing through the public mempool, so searchers cannot see it in flight. Flashbots Protect has served 2.1 million Ethereum accounts and shielded around $43 billion in DEX volume as of late 2024, and Polygon launched its own private mempool in April 2026.

The honest caveat: private is not airtight. Academic work has documented sandwich attacks executed through private channels, and benchmarks have found that execution quality varies meaningfully between providers. Treat “MEV protected” as a claim to check rather than a guarantee.

Prefer intent-based or batch-auction venues for same-chain trades. CoW Swap and similar designs settle orders in batches at a uniform price rather than in a sequenced queue, which removes the ordering advantage the attack depends on.

Set slippage tolerance deliberately. Your slippage setting is the ceiling on how much a sandwich can take. A 5 on a large trade is an invitation. Tighten it, accept that some transactions will fail, and treat those failures as evidence that the protection is working.

Split large orders. Sandwich profit scales with how much your trade moves the pool. Several smaller trades across venues move each pool less, so there is less to extract.

Trade deeper pools. The same order in a deeper pool results in less price movement by default and yields less profit for a bot.

Where this meets your swap

Worth being precise here rather than claiming immunity.

Sandwiching requires two things: a trade whose intent is visible before it executes, and a price that your own order moves. A standard SimpleSwap exchange gives an attacker neither in the usual form. You are not broadcasting a signed DEX trade with your slippage tolerance attached. You send an ordinary transfer from your wallet to a deposit address e to exploit, and execution occurs on the provider’s side afterward.

Beyond that, routing itself changes the arithmetic. Where liquidity is sourced from centralized venues, there is no public mempool involved. Where a route splits a large order across several sources, each portion moves any individual pool less than the whole order would have, which is the same mechanism that keeps price impact contained and happens afterward, what a bot could extract.

The limits, plainly: this is a structural property of the flow rather than a security guarantee; DEX liquidity remains part of the wider market, and none of it protects trades you make directly on a DEX. Those need the tools in the section above.

What comes next

Each entry above will get its own breakdown in “Know the Scam by SimpleSwap” over the coming months. Follow #KnowTheScam to catch every new one as it drops — phishing kits, fake support DMs, drainer contracts, the whole map: new scam, same hashtag, one place to check.

FAQ

What is a sandwich attack in crypto? A bot spots your pending DEX trade in the public mempool, buys the same asset just before you to push the price up, lets your trade execute at that worse price, then sells immediately after. Its profit is calculated from what you received.

Is MEV illegal or a scam? Sandwiching is not fraud in the legal sense. Every transaction involved is valid and permitted. It is extraction rather than theft, which is why it persists and why the defense is technical rather than legal. What is a scam is the “MEV bot” industry, which sells contracts promising passive front-running income and takes your deposit.

Is sandwiching getting worse? By the measured numbers, no. Extraction on Ethereum fell from about $10 million a month in late 2024 to roughly $2.5 million by October 2025, even as DEX volumes grew, and researchers credit part of that to the adoption of protection tools. Attack frequency stayed high, though, so the risk to any individual trade has not disappeared.

How do I know if I have been sandwiched? Usually, you do not, which is the core problem. Your transaction succeeded normally. Block explorers such as EigenPhi let you inspect a transaction and see whether it was bracketed by trades from the same address in the same block.

Does slippage tolerance protect me? It caps your maximum loss rather than preventing the attack. A sandwich cannot take more than your slippage setting allows, so a wide default is what makes a large trade worth attacking.

Are stablecoin swaps safe from this? No. Nearly 40% of measured attacks targeted low-volatility pools. The attack exploits transaction ordering, not price volatility.

Is SimpleSwap safe from sandwich attacks? A standard SimpleSwap exchange does not broadcast a signed DEX trade with your slippage tolerance attached, which is the thing bots hunt for in the public mempool. You send an ordinary transfer to a deposit address, and routing occurs on the provider’s side afterward, partly through centralized liquidity, where no public mempool exists. That is a structural property of the flow, not a security guarantee, and it does nothing for trades you make directly on a DEX. Our only official domain is simpleswap.io.

Is SimpleSwap a scam? No. SimpleSwap has operated since 2018 as a self-custodial swap aggregator, moving funds between wallets you control without holding a balance in between. Two unrelated things push that phrase into search results.

The first is brand impersonation: cloned sites and fake support accounts that borrow a recognizable name in order to phish people. Our only official domain is simpleswap.io; we never message you first, and nobody at SimpleSwap will ever ask for your seed phrase.

This article is for educational purposes only and is not financial or security advice. SimpleSwap’s only official domain is simpleswap.io.