Aave Liquidity Crunch Worsens, $300M Borrowing Surge Traps Depositors

The Aave liquidity crunch triggered by Saturday’s KelpDAO exploit has produced one of DeFi’s most quietly brutal mechanics: depositors who cannot withdraw borrowing against their own trapped funds at a deliberate loss, just to get any money out at all. According to data from Chaos Labs, roughly$300 millionin new USDT-collateralized borrowing hit Aave in the single day following theApril 18attack, not because anyone wanted use, but because the exit doors were welded shut.

How the Aave Liquidity Crunch Began: The KelpDAO Bridge Exploit

The KelpDAO attack didn’t target Aave directly. That’s the detail that makes it so damaging. OnApril 18an attacker found a way to manipulate KelpDAO’s cross-chain bridge infrastructure into releasing116,500 rsETHa liquid re-staking token representing staked ETH positions, without any real backing behind them. That’s roughly18%of rsETH’s entire circulating supply, worth about$292 millionat the time of the attack.

Those counterfeit tokens were immediately deposited into lending protocols, primarilyAaveand used as collateral to borrow real assets: ETH, wETH, and stablecoins. Fake tokens in. Real money out. The attacker didn’t need to crack Aave’s code; they just needed the system to believe the collateral was legitimate, and it did.

Aave founder Stani Kulechov confirmed that Aave’s own contracts were never compromised. The protocol frozersETHmarkets on both V3 and V4 within hours of the exploit going public. That freeze stopped the active drain. But it also triggered a second crisis that nobody saw coming.

Whales Left First, and the Pool Ran Dry

When exploit news broke, large holders moved fast. Justin Sun, MEXC exchange, and other institutional-scale wallets began pulling funds from Aave almost immediately. Within hours, over$6 billionin assets had left the protocol, a withdrawal pace that no lending pool is built to absorb gracefully.

Aave’s liquidity pools work on a simple mechanic: depositors earn yield, borrowers pay interest, and the gap between total deposits and total loans is the “available liquidity”, the amount depositors can actually withdraw at any moment. When outflows eat through that buffer entirely, utilization hits100%and the exit is gone.

ETH markets hit100% utilizationfirst. Then the contagion spread to USDT and USDC pools. “When the rsETH exploit happened and AAVE incurred bad debt, whales like Justin Sun, MEXC exchange, and others immediately withdrew billions from AAVE,” crypto analyst Duo Nine said in a post-incident breakdown. “These markets are now also stuck with money locked.”

That [borrowed] WETH is gone. The rsETH holding its place in the vaults is worth whatever an unbacked claim is worth, approaching zero on the L2 side, where 20+ chains held bridged rsETH backed by a now-empty mainnet lockbox.

What Does the $300 Million Borrowing Surge Actually Mean?

Here’s the perverse logic that produced the$300 millionborrowing spike. Depositors stuck in maxed-out pools still have money in Aave, they just can’t withdraw it. Aave allows borrowing up to75% loan-to-valueagainst deposited collateral, so a user with$100,000in locked USDT can borrow$75,000in a moveable stablecoin.

That’s not a clever trade. That’s paying a10-25% tollto exit your own position. You’re accepting 75 cents on every dollar, leaving behind a debt that still earns interest, and walking away lighter than you walked in. The fact that$300 millionworth of people chose this option tells you exactly how desperate the situation became.

monetsupply.eth, pseudonymous head of strategy at Spark, a competing DeFi lending protocol, flagged the cascading effect this creates. Borrowing against locked USDT to access USDC doesn’t free the system; it just shifts the pressure. “With a 75% max LTV, users with stuck USDT deposits can take out up to 3/4 of the value of their Aave position. But this ends up reducing liquidity in other markets, with USDC and USDe markets now at 100% utilization as well,” they observed.

We’re now seeing some negative secondary effects of illiquidity in Aave stablecoin markets. Because users can’t withdraw due to 100% utilization, there has been a ~$300 million increase in borrowing with USDT collateral in just the past day since the rsETH exploit.

The Bank Run Analogy, And Why It Breaks Down

The comparison to a traditional bank run is almost accurate. A fractional reserve bank holds far less cash than it owes depositors, in a panic, the first people in line get paid and the last ones don’t. But banks have emergency mechanisms: central bank lending facilities, FDIC insurance, overnight repo markets, a phone call to a finance minister.

DeFi has code. Specifically, it has interest rate curves that are supposed to make borrowing expensive when utilization climbs high, which should attract fresh capital into the pool and naturally restore liquidity. The assumption baked into that mechanism is that utilization spikes are temporary noise. A coordinated$6 billionwithdrawal triggered by an external exploit is not noise, it’s a scenario the interest rate model doesn’t have a good answer for.

TheKelpDAO exploitdidn’t just steal money. It demonstrated that whale coordination during a crisis can overwhelm a DeFi protocol’s self-correction mechanisms entirely. Smart contracts can freeze bad collateral. They cannot freeze the panic response of the people holding good collateral.

Who Pays the Price When DeFi’s Exit Doors Close?

Not the whales. Justin Sun and MEXC got their money out in the first hours. The people paying the 25% haircut to escape are retail depositors, smaller funds, and anyone who wasn’t watching a Discord alert feed on a Saturday afternoon.

That asymmetry, between who gets out clean and who pays to leave, is the part of this story that deserves more scrutiny than it’s getting. The standard DeFi narrative says that smart contracts remove gatekeepers and give everyone equal access. What this week showed is that block-by-block on-chain execution speed still creates a tiered system: those with the fastest bots and the biggest positions move first, and everyone else absorbs the consequences.

Duo Nine put it plainly: “Some users decided to borrow against USDT/USDC and exit via other markets at a 10-25% loss. Basically you borrow GHO/DAI/USDe against your locked USDT/C.” He wasn’t describing a strategy. He was describing a distress exit. And$300 millionin distress exits is a number that should be uncomfortable for anyone who believes permissionless finance is categorically fairer than the alternative.

As whales took out their money, USDT and USDC also hit 100% utilization. These markets are now also stuck with money locked.

This article is for informational purposes only and does not constitute investment advice. Every investment and trading decision involves risk. Readers should conduct their own research before making any financial decisions.