Circle Chief Economist Gordon Liao has called for a sharp increase in USDC borrowing rates on Aave V3’s Ethereum Core market after the fallout from the KelpDAO exploit left the lending pool effectively jammed and depositors struggling to access liquidity. The proposal, submitted on April 22, 2026, came after a major dislocation triggered by the April 18 exploit of KelpDAO’s rsETH bridge infrastructure. While Aave’s own smart contracts were not directly compromised, the incident quickly spread stress across the protocol’s core markets and exposed the risks that can arise when unbacked collateral enters DeFi lending systems.
How the KelpDAO exploit spilled into Aave
According to the governance proposal, attackers exploited a vulnerability in KelpDAO’s cross-chain bridge, allowing them to mint or move unbacked rsETH and deposit it into Aave V3 as collateral. With that collateral in place, they were able to borrow an estimated $200 million to $300 million worth of assets, including WETH and major stablecoins. Aave responded by freezing rsETH markets on both V3 and V4 to limit further damage, but by then the shock had already spread throughout the broader lending environment.
The result was a rapid liquidity crunch. Aave’s total value locked reportedly fell by billions of dollars within days as users reacted to the incident and tried to withdraw funds. Core markets including ETH, USDT, and USDC were pushed to 100% utilization, leaving depositors unable to exit normally. Some users reportedly tried to obtain liquidity indirectly by borrowing against stablecoins such as GHO, DAI, and USDe rather than withdrawing from the affected pools outright.
USDC pool became the center of the crisis
The USDC pool on Aave V3 Ethereum Core became the clearest symbol of the problem. Data cited from Aavescan on April 22 showed total supply and total borrows both near $1.89 billion, with less than $3 million in available liquidity. The pool remained stuck at roughly 99.87% utilization for four consecutive days. Despite that extreme stress, variable borrow rates were capped near 13.82%, while supply rates were around 12.42%.
Liao argued that this rate structure was too flat to restore market balance. In his view, the post-kink portion of Aave’s interest rate curve for USDC did not create enough pressure on borrowers to deleverage or repay. Instead, borrowers who were relatively insensitive to financing costs treated the roughly 14% borrowing cost as an acceptable fee for maintaining access to scarce capital. That meant the market was not clearing, even while depositors faced delays and constraints in accessing their funds.
What Circle’s economist wants changed
The proposal outlines a two-step adjustment to Aave’s USDC interest rate model. First, Liao asked Aave’s Risk Steward to immediately raise Slope 2 from 10% to 40% and reduce optimal utilization from 92% to 87%. In a second phase, a full governance vote would seek to raise Slope 2 to 50% and lower optimal utilization further to 85%. Under the proposed target configuration, the maximum USDC borrowing rate at 100% utilization would climb to approximately 53.5%, a dramatic increase from the current ceiling of around 14%.
Other parameters would remain unchanged. Liao’s framework keeps Slope 1 at 3.5%, the base rate at 0%, and the reserve factor at 10%. If the Risk Steward action were approved, the initial adjustment could take effect the same day, while the complete target configuration would still require Aave’s normal governance process, which the report says would take roughly five to seven days.
The logic behind the rate hike
Liao said a much steeper post-kink curve would help restore price discovery in a market that had stopped functioning properly. By making borrowing materially more expensive at extreme utilization levels, the protocol could encourage rate-insensitive borrowers to exit, reduce excessive demand for stablecoin liquidity, and attract fresh USDC deposits from market participants seeking yield. He pointed to traditional short-term funding markets, including Treasury repo and fed funds, as examples of how steep rates can help resolve temporary liquidity dislocations.
Importantly, Liao stated that the proposal reflected his personal views and was not an official Circle policy position. Even so, the recommendation drew broad attention, especially after Circle CEO Jeremy Allaire shared it publicly on X. That helped elevate the issue within the Aave governance community, where participants such as Llamarisk and Aave Labs were expected to weigh in on the appropriate response.
Broader implications for Aave and DeFi risk management
The KelpDAO episode did not originate inside Aave, but it highlighted how bridge-related failures can generate protocol-level risk for lending markets. When unbacked collateral enters a borrowing system, it can create bad debt even if the lending protocol’s core contracts remain secure. In this case, estimates cited in the report suggest potentially unrecoverable losses across Aave markets could range from $124 million to $230 million, depending on how losses are ultimately recognized and distributed.
Aave has already published an rsETH incident report and is discussing potential recovery paths. One of the more sensitive governance questions is whether loss absorption mechanisms such as slashing stkAAVE should be considered. That debate goes beyond the immediate rate adjustment proposal and speaks to the broader question of how decentralized lending protocols should allocate the costs of cross-protocol failures.
The market reaction underscored the seriousness of the event. AAVE’s token price reportedly fell around 20% to 26% in the days following the exploit as users pulled capital and reassessed protocol risk. Against that backdrop, some observers interpreted Liao’s proposal as a constructive signal that governance participants were moving toward a practical stabilization response rather than leaving the market stuck in a near-frozen state.
Whether the proposal is adopted in full or modified through governance, the episode has become a high-profile test of DeFi resilience. It illustrates how quickly liquidity can vanish when collateral quality is compromised, and how interest rate models may need to react aggressively when utilization remains pinned at extreme levels. For Aave, the immediate issue is restoring normal USDC market function. For the wider sector, the bigger lesson may be that bridge exploits can reverberate far beyond the protocol where the original vulnerability appears.