0x08a6b7ab08a6…08a6b7a8

ConfirmedSecurity710 vB93 sat/vB4 min decode

Kava Lend Attacker Extracted 55.2M KAVA via 127-Loan Liquidation Cycle

On-chain records show an attacker borrowed 55.2 million KAVA from Kava Lend through 127 loans, liquidating each with a second account to bypass the protocol's borrowing limit.

Kava Lend Attacker Used 127 Loans to Bypass Borrowing Limit
WitnessKava Lend Attacker Used 127 Loans to Bypass Borrowing LimitEivindPedersen / pixabay

Outputs

  1. Attacker borrowed approximately 55.2 million KAVA from Kava Lend

  2. The exploit used 127 loans, each liquidated by a second account to bypass the borrowing limit

  3. A nearly idle BUSD market's deposit value rose about 243,000 times in roughly three minutes

  4. The attacker repeatedly tested the market beginning Nov. 8 before the main sequence

An attacker borrowed approximately 55.2 million KAVA from Kava Lend by executing 127 separate loans and using a second account to liquidate each position, according to on-chain records reviewed in connection with the exploit. The mechanism allowed the attacker to circumvent the lending protocol's borrowing limit rather than break its code outright.

Kava Lend is a lending protocol built on the Kava blockchain. The exploit, previously covered by TokenPost, targeted the protocol's collateral accounting rather than a single smart-contract flaw.

How did the attacker inflate the collateral?

The operation began with a nearly idle BUSD market on Kava Lend. The attacker first cut that market's net supply to a very small level. Within roughly three minutes, the market's deposit value rose about 243,000 times — a figure that, translated plainly, means a marginal market suddenly carried collateral worth hundreds of thousands of times its prior scale.

That inflation raised the collateral value the attacker could use to borrow KAVA. The protocol accepted the inflated deposits as legitimate backing.

The scale deserves emphasis. A 243,000-fold increase in three minutes is not organic market activity under any reading. It is the arithmetic signature of a market with almost no liquidity being manipulated to a state the protocol's risk parameters did not anticipate.

What role did the 127 loans play?

The attacker repeated a two-step cycle. The first account borrowed KAVA against the inflated collateral. A second account then liquidated each loan as it was opened.

Each liquidation reset the position. The cycle let borrowing continue past Kava Lend's borrowing limit, because no single loan remained outstanding long enough to register against the cap. The on-chain record shows the sequence ran to 127 loans.

The second account is the load-bearing detail. Without it, the borrowing limit would have halted the attacker after far fewer transactions. With it, the limit became a threshold to be walked around, one liquidation at a time.

In plain words: the protocol's cap counted standing loans, and the attacker ensured almost none were standing.

Was the attack spontaneous?

The on-chain evidence suggests it was not. The attacker repeatedly tested the BUSD market beginning Nov. 8, probing its behavior before executing the main sequence.

That testing window implies reconnaissance. An attacker who spends days probing a market's response to supply manipulation is an attacker who understands which parameter will fail — and, notably, one who found no obstacle during the testing phase itself.

The absence of any intervention during those preliminary tests is not addressed in the available record. Whether monitoring existed and missed the probes, or did not exist, remains unstated.

What are the operational consequences?

The exploit exposes a structural weakness in how Kava Lend measures collateral in thin markets. A market described as nearly idle should not be able to absorb a 243,000-fold deposit inflation without triggering a circuit breaker. It did.

The borrowing limit, meanwhile, functioned as designed at the individual-loan level and failed at the aggregate level. The 127-loan count is itself a measure of that failure: the constraint was not defeated once, but 127 times in sequence.

For the protocol's operators, the incident raises two immediate questions. First, whether net-supply thresholds in low-liquidity markets carry adequate floors. Second, whether liquidation patterns — specifically, a single counterparty systematically liquidating one borrower's fresh positions — can and should be flagged in real time.

Neither question has a published answer yet. No statement on remediation, fund recovery, or parameter changes appears in the material reviewed.

What happens next?

The on-chain trail is complete and public: account addresses, loan sequence, liquidation events and the Nov. 8 testing activity are all recorded on the Kava blockchain. That record gives investigators and the protocol's team a full map of the exploit, though no recovery figure has been disclosed.

The market-structure lesson is narrow but consequential. Protocols that list low-activity markets as collateral sources now face a demonstrated attack template: shrink net supply, inflate deposit value, cycle loans through a liquidator. Expect Kava Lend, and lending protocols with similar architectures, to face pressure to cap or delist idle markets before the next reconnaissance campaign begins.

via tokenpost.com (Source)

More from Elena Vasquez

Elena Vasquez

Show full bio

Staff writer covering regulation, policy and their impact on crypto and web3 industry at Mempool Brief.

1,203 articles