Crime•

Block Builders Collected $1.78 Million of the $3.4 Million Drained From K3 Capital's sBOLD Vault, More Than Triple the Original Attacker's Take

On-chain data shows a copycat bot seized 56% of the funds drained from K3 Capital's sBOLD vault and passed most of it to Ethereum block builders for transaction priority.

Written By :MINRK
Block Builders Collected $1.78 Million of the $3.4 Million Drained From K3 Capital's sBOLD Vault, More Than Triple the Original Attacker's Take

Photo by Jonathan on Unsplash

Mentioned in This Article

ETH
ETH
Ethereum
$1,664.81
2.38%
B
BOLD
Price unavailable
W
WETH
Price unavailable

1. Where the sBOLD hack money went

More than half of the $3.4 million drained from K3 Capital's sBOLD vault on Wednesday, November 18, 2026, was paid to Ethereum block builders rather than kept by the people who attacked it, on-chain data shows. Blockchain analytics firm Bitquery traced 97 transactions that emptied the vault in about eight minutes, with builders receiving roughly $1.78 million from competing bots.

The numbers turn the usual picture of the sBOLD hack on its head. In most thefts from decentralized finance protocols, the question is how much the attacker walked away with. Here, the wallet that started the attack ended up with a small fraction of the total, an automated copycat captured the largest portion, and that copycat then handed most of what it took to the firms that assemble Ethereum's blocks in exchange for having its transactions placed ahead of everyone else's.

The result is a case study in how value moves once a flaw becomes visible on a public blockchain. The weakness sat in the vault's accounting, yet the largest single flow of money went to infrastructure operators that had no part in finding or exploiting it.

Three figures summarize the distribution. The copycat bot extracted $1.91 million. Block builders were paid about $1.78 million by bots bidding for position. The original attacker, together with a wallet it funded, finished with about $520,000. The sections below explain how the vault was drained, why a second party was able to capture most of the proceeds, and why the amount paid to builders should not be read as their profit.

2. What is sBOLD and who built it?

sBOLD is a crypto savings product developed by K3 Capital. Users deposit BOLD, a token designed to hold a value of $1, into a shared vault, and the vault's software distributes those deposits across Liquity lending pools so that they earn interest. Depositors receive shares that represent their claim on the pooled funds, and the value of each share is supposed to rise gradually as interest accumulates.

A structure of this kind depends entirely on its bookkeeping. Because every depositor's money sits in one pool, the vault must calculate at any moment what the whole pool is worth and how many shares exist against it. That ratio determines how many shares a new depositor receives and how much BOLD a departing depositor can withdraw. If the calculation can be pushed out of line, even briefly, someone redeeming shares at the wrong moment can leave with more than they put in, and the difference comes out of the balances of everyone else in the pool.

That is the category of weakness the attackers found. The vault held funds belonging to many users, and the theft consisted of withdrawing value that the accounting wrongly attributed to the attackers' shares.

3. How did the sBOLD exploit work?

The attackers manipulated the way the vault valued its shares, then redeemed those shares at an inflated price. The sequence began with repeated deposits of BOLD. The attackers then added wrapped ether to the vault, an asset that the vault's accounting initially ignored. Next, that ether was converted into BOLD, at which point it counted toward the vault's holdings, and the attackers redeemed their shares at a value higher than what they had paid for them.

The gap between the two moments is what made the attack profitable. Value was introduced in a form the vault did not count, then turned into a form it did count, which lifted the worth of each share after the attackers already held them. Redeeming at that point returned more BOLD than the deposits had cost.

The steps were also repeatable. The whole sequence could be run several times inside a single transaction, which allowed the attackers to keep pulling out money that belonged to other depositors until little remained. Across 97 transactions and roughly eight minutes, the vault was emptied.

4. A copycat bot took the largest share

The wallet that first exploited the flaw did not end up with most of the money. One automated bot copied the attack and extracted $1.91 million, equal to 56% of the stolen funds. It then paid most of that sum to block builders so that its transactions would be ordered ahead of rival bots chasing the same opportunity.

This pattern follows from how public blockchains operate. Pending transactions on Ethereum are generally visible before they are confirmed, and so is every transaction that has already been confirmed. Once an attack succeeds on-chain, the method is effectively published. Automated trading systems that monitor the network can reproduce a profitable transaction with their own addresses substituted in, and they can do so within seconds.

When several bots attempt the same extraction at once, only the first to be included in a block succeeds, since the funds can be taken only once. That turns the contest into an auction. Each bot offers a larger payment to the party that decides transaction order, and the winner is the one prepared to give up the most of its prospective gain. In this case the bidding became intense enough that the payment for inclusion in a single block reached 44 ETH.

5. What is MEV and why do block builders get paid?

Maximal extractable value, or MEV, is the profit that traders and bots can capture by controlling where transactions sit within a block, and they pay to be placed ahead of competitors. The term covers a wide range of activity, from arbitrage between exchanges to liquidations of undercollateralized loans. Copying an exploit and racing to execute it first is a version of the same thing: the profit exists only for whoever is ordered first.

Block builders are the entities that assemble Ethereum blocks. They collect transactions, arrange them in the order that produces the most revenue, and submit the finished block for publication. Because they decide the sequence, they are the natural recipients of payments from anyone who needs a specific position. A bot that wants its transaction processed before a rival's attaches a payment to the builder, and the builder includes the highest-paying version.

A review of the blocks involved found that Titan built most of the blocks containing the bot's transactions. BuilderNet, Ultrasound and Quasar also processed some of them. Together, block builders received about $1.78 million from the bots competing over the sBOLD flaw, which is more than half of everything taken from the vault.

6. Payments received are not the same as profit

The $1.78 million figure carries an important qualification: it measures incoming payments, not confirmed profits. Builders do not keep everything they collect. They in turn pay validators, the computers that check transactions and add blocks to Ethereum, to publish the blocks they have assembled. Builders compete with one another for that privilege, so a large part of what a bot pays a builder can be passed further along the chain.

The analysis does not show how much the builders retained after those onward payments. It is therefore accurate to say that builders received the largest flow of funds from the attack, but not that they earned $1.78 million from it. Some portion of the money taken from sBOLD depositors continued on to validators, and the data traced so far does not establish the split.

The distinction matters for anyone trying to follow the funds. Value taken from the vault did not stay in a single attacker's wallet. It was distributed across the bot that won the race, the builders that ordered its transactions, and the validators that published those blocks, each through the ordinary payment mechanisms of the network.

7. The original attacker ended with about $520,000

Against those sums, the originator of the attack fared poorly. The original attacker and a wallet it funded finished with about $520,000, according to a review of the transactions. That is well below both the $1.91 million extracted by the copycat bot and the roughly $1.78 million that flowed to block builders.

In other words, the party that identified the weakness and carried out the first successful transaction captured a minority of the $3.4 million. Most of the value was taken by a faster automated system that reproduced the technique, and most of that system's proceeds were then spent on winning priority. For the depositors whose BOLD was in the vault, the distribution makes no difference to the loss. It does, however, complicate any effort to trace or recover the funds, because the money has passed through several kinds of participants.

8. An earlier audit and the questions still open

The sBOLD code had been reviewed before the attack. ChainSecurity audited the vault in May 2025 and identified a way to extract value through self-liquidations. The firm said a later version of the code mitigated that issue.

Several points remain unresolved. It is not yet clear whether Wednesday's attack relied on the same mechanism the auditors flagged, whether the fixes they reviewed were actually deployed to the live vault, or whether an entirely different accounting weakness was responsible. Each possibility carries a different implication: a known issue that returned, a fix that never reached production, or a new flaw that the review did not cover.

An audit is a review of a specific version of code at a specific time. Its conclusions hold only if the version that was examined is the one that runs in production and if nothing material changes afterward. That is why the question of deployment matters as much as the question of which mechanism was used.

Until those questions are answered, the on-chain record offers the clearest account of what happened. It shows a vault emptied in about eight minutes, a copycat bot that outpaced the original attacker, and a bidding contest that sent the largest part of the proceeds to the builders of Ethereum's blocks.

Related Articles

NEWSLETTERS

Don't miss another story.

Subscribe to the MINRK Newsletter today.

Tick the newsletters you want. No selection? You'll get MINRK Headlines.

By signing up, you will receive emails about MINRK products and you agree to our terms of use and privacy policy.

MINRK Daybook

Market analysis for crypto traders and investors.

EVERY WEEKDAY

MINRK for Advisors

Defining crypto, digital assets and the future of finance for financial advisors.

EVERY THURSDAY

MINRK Protocol

Exploring the tech behind crypto one block at a time.

WEEKLY

MINRK Long & Short

A must read for institutions. Insights, news and analysis delivered weekly.

EVERY WEDNESDAY

MINRK Headlines

The biggest crypto news and ideas of the day.

EVERY WEEKDAY

MINRK Policy

Examining the intersection of cryptocurrency and government.

WEEKLY

MINRK Research

Join thousands of readers who rely on MINRK for data-driven insights on the latest digital asset trends.

MONTHLY