04a7fd75-8

ConfirmedSecurity646 words3 min decode

Ethereum block builders received $1.78M from sBOLD vault exploit

Bitquery traced 97 transactions that drained $3.4 million from K3 Capital’s sBOLD vault in about eight minutes; Ethereum builders received $1.78 million in payments.

Ethereum block builders received the most in a $3.4 million vault hack, data shows
WitnessEthereum block builders received the most in a $3.4 million vault hack, data showssteve_a_johnson / pixabay

Outputs

  1. Bitquery traced 97 transactions that drained $3.4 million from K3 Capital’s sBOLD vault in about eight minutes on Wednesday.

  2. Ethereum block builders received about $1.78 million, more than half of the stolen proceeds.

  3. A copied trading bot extracted $1.91 million, or 56%, and paid most of it to builders.

  4. One Ethereum block-inclusion payment reached 44 ETH.

  5. ChainSecurity audited sBOLD in May 2025 and reported a later-code mitigation for a self-liquidation extraction path.

By Tom Whitfield

A $3.4 million exploit drained K3 Capital’s sBOLD Ethereum savings vault in about eight minutes, according to on-chain tracing from blockchain analytics firm Bitquery. Ethereum block builders received approximately $1.78 million in inclusion payments from bots competing to exploit it.

The trace covered the 97 transactions that emptied the vault on Wednesday. More than half of the stolen proceeds went to the companies responsible for assembling Ethereum blocks. A separate CoinDesk transaction review found that the original attacker and a wallet it funded ended with about $520,000.

How did attackers manipulate sBOLD shares?

The attackers manipulated how the vault calculated the value of its shares. They repeatedly deposited $BOLD, introduced wrapped ether that the vault initially excluded from its accounting, and converted that ether into $BOLD before redeeming their shares at an inflated value.

The sequence could repeat several times inside one transaction. That let the attackers extract funds belonging to other depositors rather than only their own balances, turning a valuation error into a direct loss for the shared pool.

K3 Capital develops sBOLD, a crypto savings product whose users deposit $BOLD, a token designed to remain worth $1, into a shared pool. Its software spreads those deposits across pools on crypto lending platform Liquity to earn interest.

That structure places share pricing at the center of redemption risk. The attackers repeatedly changed the effective asset mix before the vault calculated the value of the shares they redeemed.

Why did block builders receive more than half the loss?

One automated trading bot copied the attack and extracted $1.91 million, or 56% of the stolen funds. It paid away most of those proceeds to block builders to ensure its transactions were processed ahead of rival bots.

Those payments formed part of maximal extractable value, or MEV, under which traders and bots profit from transaction ordering and pay to get ahead of competitors. Competition became intense enough that inclusion in one Ethereum block drew a payment of 44 ETH. The payment measured competition for ordering, not a builder’s final profit.

The copied bot’s $1.91 million extraction was more than three times the roughly $520,000 that the original attacker and a wallet it funded ended with. The bot’s figure was gross extraction because it paid most of the proceeds to builders.

Builders also pay validators, the computers that check transactions and add blocks to Ethereum, to publish their blocks. Bitquery’s $1.78 million figure therefore measures incoming payments, not confirmed profits, and the analysis does not establish how much any builder retained after those costs.

Could the May 2025 audit findings have covered the exploit?

ChainSecurity audited sBOLD in May 2025 and identified a way to extract value through self-liquidations. It said the issue had been mitigated in a later version of the code.

The audit result does not establish which code version Ethereum used on Wednesday. The traced transactions likewise do not show whether the version containing that mitigation processed the attack.

It remains unclear whether Wednesday’s attack exploited a related mechanism, whether the reviewed fixes were deployed, or whether another accounting weakness was involved. The missing deployment history prevents a definitive link between the May 2025 finding and the $3.4 million loss.

The immediate operational question is which code version processed Wednesday’s withdrawals and whether its accounting logic excluded wrapped ether at the point used by the attackers. A K3 Capital disclosure on the deployed version and remediation would establish whether Ethereum used the version with the reported self-liquidation mitigation and which share-calculation changes remain necessary.

For Ethereum’s MEV market, the next checkpoint for risk assessment is a K3 Capital disclosure of Wednesday’s code version. Until that information is available, block-inclusion competition could redirect any future sBOLD accounting loss into builder payments.

via cryptonews.net (Source)

More from Tom Whitfield

Tom Whitfield

Show full bio

News editor covering technology and product news in crypto and web3 industry at Mempool Brief.

1,440 articles