0x3cc61bd23cc6…3cc61bd5
Blockaid flags ongoing $3.43 million sBOLD vault exploit
Blockaid flagged an ongoing $3.43 million exploit against K3 Capital's sBOLD vault, where discounted liquidation shares drained deposits and raised broader DeFi integration risks.

Outputs
Blockaid flagged an ongoing $3.43 million exploit against K3 Capital's sBOLD vault on November 18.
The attacker reportedly used a flash loan and self-liquidation to mint discounted sBOLD shares.
K3 Capital launched sBOLD around June 2025 after a May 2025 audit by ChainSecurity.
Reental reported a separate approximately $454,000 exploit on its Polygon platform on November 18.
The two unrelated incidents brought reported losses for the day to approximately $3.9 million.
By Marcus Bennett
Blockaid flagged an ongoing exploit on November 18 that has caused about $3.43 million in losses at K3 Capital’s sBOLD vault, according to its research. The report said the total could still rise while the attack remained active, making the current figure a snapshot rather than a final loss estimate.
How the exploit worked
sBOLD is an ERC-4626 tokenized vault on Ethereum. The standard lets other protocols integrate a vault without custom wiring. K3 Capital pools users’ $BOLD deposits into Liquity V2’s stability pools, where they earn yield sourced from borrower interest and liquidation premiums.
Those premiums make liquidation activity central to the vault’s accounting, not a peripheral event. Stability pools earn when troves get liquidated, and sBOLD deposits sit directly inside that mechanism.
The attacker reportedly used a flash loan and their own trove to target that mechanism. A flash loan is a short-term DeFi loan that must be borrowed and repaid within a single transaction. The report describes wstETH as wrapped staked Ether, a liquid token representing ETH staked through Lido.
On Liquity V2, a trove is a borrowing position: a user locks collateral and mints $BOLD, the protocol’s stablecoin. By liquidating their own trove into the stability pools, the attacker could mint sBOLD shares at a discount.
The attacker then swapped collateral and redeemed deposits belonging to other users. That sequence turned an accounting advantage around liquidations into a direct loss for the vault.
The exploit matters because liquidation premiums, share pricing and deposited $BOLD all meet in the same accounting system. Its effects therefore extend beyond a single transaction and reach the assumptions supporting the vault’s composability.
The audit did not remove every risk
K3 Capital launched sBOLD around mid-2025, with Blockaid’s research placing the launch around June. The firm described the product as delta-neutral and said composability with other DeFi protocols could improve capital efficiency.
ChainSecurity audited the vault in May 2025, shortly before launch. Its findings identified weaknesses and highlighted areas for improvement. That distinction narrows what an audit can promise: it can identify design weaknesses and suggest improvements, but it cannot guarantee that every vulnerability has been removed.
On the same November 18, Reental, a tokenized real-estate platform running on Polygon, reported a separate exploit costing approximately $454,000. Blockaid’s research treats the incidents as unrelated. Together, the two cases pushed reported losses for the day to approximately $3.9 million.
What does this change for sBOLD depositors?
The most immediate concern is whether the reported $3.43 million loss grows. Because Blockaid described the attack as ongoing, sBOLD holders need updates from K3 Capital and security firms before treating the current figure as final.
Composability creates a second channel of exposure. Any protocol that accepts sBOLD as collateral or integrates it into a strategy may need to assess its own risk, even if it did not participate in the exploit. For depositors, the key issue is containment; for builders, it is exposure across linked positions.
Integrators built around sBOLD must treat the vault as a risk-bearing dependency, not just another source of yield. That is the operational cost of composability when a shared accounting failure reaches multiple products.
The next material checkpoint will come when K3 Capital or investigators confirm that the exploit has ended and publish final loss and exposure figures; until then, sBOLD remains an active security dependency for depositors and integrators.
via cryptonews.net (Source)
More from Marcus Bennett
Show full bio
Senior reporter covering money, funding and market shifts around crypto and web3 industry at Mempool Brief.
1,456 articles