Since the KelpDAO incident on April 18, 2026, Aave founder Stani Kulechov has consistently made it clear that the protocol was not hacked or compromised in any way during the attack. Still, that hasn’t stopped Aave from losing nearly $10 billion over the next four days, and Fluid is among those picking up the slack.
As Aave grappled with its pools being completely drained, withdrawals frozen, and lenders stuck with collateral they didn’t want, Fluid launched a WETH Redemption Protocol without warning.
The concept of the protocol was surprisingly good: Fluid’s light vault contained wstETH as collateral $ETH debts on Aave, creating inverse positions that directly matched users stuck on Aave.
This way they could hand over their property $ETH collateral and get wstETH or wETH in return without Aave’s frozen pools having to do anything.
It’s time to shine
According to Castle LabsFluid processed 166,722 aETH (approximately $400 million) through the redemption protocol in just two days. Dune-on-chain data had confirmed that cumulative aETH collateral swap volume soared to over 84,000 between April 20 and 21 $ETH before the Castle Labs update the full information in their X newsletter.
Fluid then expanded the protocol to Arbitrum and Base, opening a queue-based system that allowed aggressive traders to safely exit their risky positions by matching them with lenders trying to acquire different assets.
The mechanism changed slightly for Layer 2 (as most new networks require more manual debt handling), but the main principle remained the same.
When asked whether the protocol was a one-time solution or the start of something bigger, the Fluid team compared it to traditional finance. Products like credit default swaps exist for these types of situations, so Fluid expects DeFi to have its own equivalent.
Although the crisis was unplanned, it was still the most effective demonstration of that theory.
How much damage did Aave take?
The figures after operations tell it all. Aave’s total delivered dropped from $45.8 billion to $35.7 billion. The TVL also fell from $26.3 billion to approx $16.4 billion (a loss of $9.94 billion).
As Aave’s pools were completely empty, interest rates for stablecoins also rose. This caused even more damage to Aave. With liquidity depleted and withdrawals frozen, lenders began lending against their own frozen assets just to cut their losses, which in turn pushed interest rates even higher.
Total DeFi TVL also dropped from $99.5 billion to approx $86.7 billionthe lowest level in more than a year and a decline of 37% from the $119 billion recorded in early 2026. While the KelpDAO incident didn’t cause any of this, Castle Labs data confirms that it was the largest contributor, with Aave being the biggest victim.
However, not all the capital that left Aave disappeared. According to on-chain data maintained by Look at chainSpark’s TVL rose to $4.552 billion (up from $825 million), while Aave bled. DefiLlama also confirmed the change, indicating a sharp TVL increase since April 19, unlike the chart of every other project in the same time frame.
Will things work out for Aave in the end?
Launched on March 30 on the Ethereum mainnet, Aave’s V4 features an innovative liquidity architecture and collateral framework that requires bridged tokens to prove a DVN minimum of 3 out of 5.
That upgrade was supposed to take eleven days when the KelpDAO attack occurred. Combined with the existing governance crisis (the departure of BGD Labs and the dissolution of ACI), Aave has clearly had a difficult quarter.
However, it still retains the highest TVL base despite a $10 billion loss. It currently earns $560 million in annualized fees, and still has the institutional backing from Grayscale and the Bank of Canada that came in weeks before the exploit.
Aave is still the market leader and stakeholders support the protocol to be in order.