The two figures the federation put out divide into an implied $80,080 a coin [16], which is the least interesting arithmetic available here. The pair that matters sits either side of the peg: 3,996.01834922 LBTC burned on the Liquid side, per Blockstream's own explorer [6], against roughly 3,996 BTC paid out on Bitcoin [5], a difference of about 0.018 BTC [17], which is fee-shaped. The transfer settled the way a valid redemption settles.
That is the difficulty. Liquid's peg burns LBTC to release locked BTC [10], and peg-outs are gated by a Peg-out Authorization Key documented so that "even if a set of functionaries were compromised, they couldn't redirect user funds to attacker-controlled addresses" [8]. With the named key reported uncompromised [7], two readings survive, or rather two that carry very different bills: the destination was already whitelisted and the LBTC that got burned came from somewhere it should not have, which makes this one integrator's accounting failure; or an authorization reached the watchmen without the key, which makes it the peg's failure, since releasing federation-held Bitcoin requires a greater-than-two-thirds watchman threshold and nothing more [9]. The published evidence does not pick one, and that choice is the whole exposure question.
Nothing in the account points to a failure in Bitcoin itself [15], and moving 3,996 coins on-chain is not the same as selling them, since there is no cited evidence the funds have reached an exchange [11]. What has been marked down is the claim rather than the coin, because LBTC is worth locked BTC only to the extent a quorum returns that BTC to its owner, and the white-hat framing stays an assertion until the coins come back [20].
TRM Labs' first-half 2026 numbers show why this category dominates the ledger even when it is uncommon: infrastructure and operations vulnerabilities were around 15% of incidents and 76% of losses in dollar terms [12], so the average incident of that kind carries about 5.1 times the loss share of the average incident overall [18].
The demand side rhymes. Heritage Falodun and Samson Ojo count roughly 0.8% of circulating BTC in DeFi against up to 30% of Ethereum, about 37.5 times more supply participation [13][19], and put the shortfall down to trust and institutional infrastructure rather than yield. Blockstream's May roadmap points the same way, emphasising reduced reliance on trust-based schemes and a BitVM 1-of-n bridge design [14].
What would break this read: the coins return within days and Blockstream publishes a control-path account locating the failure inside one integrator's key custody, in which case the concentrated-custody framing overstates it and this is SideSwap's incident wearing Liquid's name. Scale is unsettled too, since no figure for total federation holdings has been published [21], so 3,996 BTC is a large number without a denominator.
