A Trader Swapped $50 Million on a DEX and Got Back $36,000. Here's What DEX Slippage Actually Does
6 min read
Loading...
6 min read
On 12 March 2026, a trader tried to swap $50.4 million worth of stablecoin tokens into AAVE on a decentralized exchange. The trade went through. No hack, no stolen wallet. The trader walked away with about $36,000 worth of AAVE, roughly 0.07% of what they put in. This is DEX slippage taken to its most extreme: the price you're quoted on screen and the price your trade actually fills at can be two very different numbers, and the gap between them comes down to simple math, not bad luck.
A decentralized exchange (DEX) usually doesn't match your order against another person's order the way a stock exchange does. Instead, most DEXs use an automated market maker (AMM): a pool of two tokens, sitting in a smart contract, that prices trades using a formula instead of an order book.
The most common version, used by Uniswap and copied by most others, keeps the product of the two token amounts in the pool constant as trades happen:
Here's what that means in practice. When you buy a token from the pool, you add the token you're paying with and remove the token you're buying. To keep x × y equal to the same k, the price has to rise as more of the token you want gets removed. A small trade barely moves it. A large trade, relative to how much money is sitting in that pool, moves it a lot. That price movement, between the price quoted when you started the trade and the price you actually get once it settles, is called price impact, and slippage is what you experience as a result.
The trade above is real, and it's the clearest illustration of price impact you'll find. According to reporting from The Block, The Defiant, CCN and several other outlets, the trader was moving $50.4 million out of an Aave deposit position and into AAVE tokens, using a swap tool built into Aave's own interface. The trade routed through several hops: first into USDT, then into ETH through a large, liquid pool, then finally into AAVE through a much smaller pool on a different exchange that held only around $73,000 in total liquidity.
That last pool is where things fell apart. Pushing tens of millions of dollars worth of value through a pool holding $73,000 doesn't just move the price a little, it collapses it almost entirely. The trader ended up with roughly 324 AAVE tokens, worth about $36,000 at the time, a loss of nearly the entire $50 million.
According to Aave founder Stani Kulechov, the interface had already shown a warning about extraordinary slippage before the trade was confirmed, and the quoted return was already far below a fair value. The trade went through anyway. Separately, on-chain analysis attributed to Arkham Intelligence suggested a block-building operation also extracted tens of millions of dollars from the same transaction using techniques related to the kind covered in NetWort's explainer on MEV and sandwich attacks. Both problems can stack on the same trade: a thin pool creates the price impact, and whoever is watching the transaction can extract even more value once the price has already moved.
Price impact isn't fixed. It scales with how big your trade is compared to how much liquidity sits in the pool you're trading against.
Uniswap's own documentation gives a simple comparison: a $100,000 swap between two stablecoins, going through a pool holding more than $10 million in liquidity, causes such a small price move that a trader could reasonably set a slippage tolerance as tight as 0.05% and still expect the trade to complete normally. That's a deep pool. The trade barely dents it.
Now compare that to the AAVE case above, where roughly $50 million in value tried to move through a pool holding about $73,000. The ratio between trade size and pool depth is what decides whether you get a fair fill or lose almost everything. The formula doesn't care whether the pool is deep or shallow. It reacts to both exactly the same way.
Every DEX asks you to set a slippage tolerance before you confirm a swap. It's the maximum gap you're willing to accept between the price you were quoted and the price your trade actually fills at. If the gap would be wider than your tolerance, the trade fails instead of going through at a price you didn't agree to.
Uniswap's default slippage tolerance is 0.5%, with an "auto" setting that adjusts anywhere from 0.5% to 5% depending on the pair and network conditions, according to Uniswap's own support documentation. PancakeSwap uses the same 0.5% default with a comparable auto range. Both platforms let you set a custom number instead.
If you hold a token like AAVE or other assets that trade through liquidity pools like these, checking an asset's price and volume history on NetWort's Explore screener gives you a better sense of how liquid a market actually is before you size a trade, well before an AMM's own math decides the price for you.