How long do you have to meet a bitcoin loan margin call?
Strike's page says 72 hours, Arch says typically 24, Ledn and SALT print LTV levels rather than hours. Each lender's cure window, read 4 October 2026.
Lend Ledger — Crypto lending, risk first
This week
Latest writing
At what LTV does a bitcoin-backed loan get liquidated?
SALT prints a ladder from 75% to 90.91% LTV; Lava sets a price, not a ratio. We read four lenders' own pages and turned SALT's triggers into BTC falls.
Which bitcoin lenders do not rehypothecate collateral?
Unchained, SALT, APX Lending and Ledn each publish a promise about reusing bitcoin collateral. We read all four word for word and graded who can check it.
Ledn vs Unchained: which bitcoin loan liquidates first?
Ledn's page prices an LTV-triggered loan; Unchained's commercial-loan page defines a CTP violation and prints no sale level. Both read 20 September 2026.
Does Ledn rehypothecate bitcoin collateral? Not for yield
Ledn says collateral is not lent for yield since May 2025, yet its own explainer allows re-posting to a funding partner. What its Open Book Report shows.
Is Strike's volatility-proof bitcoin loan liquidation-proof?
No. The price triggers are gone, but a missed payment still sells collateral after a 10-day grace. Both Strike ladders and Ledn's, tabulated and priced.
The $9.3m ankrFLOW headline was wrong by a factor of twenty
Blockaid's initial estimate put the MORE Markets drain at $9.3 million. The corrected figures reported four days later are roughly $410,000 at spot. What a lender should take from the gap.
Arch Lending's tokenised-gold loans, assessed on what a 90% liquidation threshold assumes
A $250,000 minimum, up to 75% initial advance, margin call at 85% and liquidation at 90%. Every threshold is published in the launch release. What the release does not publish is the price source that decides when they are crossed.
The MAMO post-mortem on Moonwell's forum puts residual debt at $9.13m
The $8.7m the trackers reported matches the stablecoins the post-mortem traces off Base. What a forensic report published on the protocol's forum a day later adds is a different number: $9.13m of borrower obligations still outstanding. The mechanism is not a vote and not a broken contract, but a donation to a market contract and a price feed that believed it.
Galaxy's Crypto Portfolio Line of Credit, assessed on a no-rehypothecation promise
A revolving line against BTC, ETH and staked SOL at 50% loan-to-value and an advertised 8.99% variable APR. The pitch rests on collateral the firm says it does not lend on — a claim the borrower cannot check, and the launch coverage never describes the margin call.
A governance vote, not a code bug, emptied Term Finance's vaults
An attacker bought majority control of a thinly held governance token, passed proposals against Term Finance's strategy vaults and moved roughly $8.5 million out. The contracts executed exactly as written.
Coinbase's crypto-backed borrowing, assessed as a custodial door onto an on-chain loan
The app is a familiar exchange account. The loan underneath it is a smart-contract position that a price feed can close without asking. Assessed on the gap between those two facts.
The same lending engine is now running behind three different logos
Coinbase, Uniswap and Crypto.com each ship a lending or borrowing button with their own branding on it. The documented integrations keep pointing at one shared backend, and that is a risk question.
Custodial crypto lending, assessed as a counterparty question
Hand assets to a firm, receive an advertised yield. Assessed on the only question that has ever mattered in this model: who is paying that yield, and what happens if they cannot.
'Insured' keeps appearing in crypto-lending copy without a named underwriter
The word does real persuasive work on a landing page. Read closely, a great deal of it describes something narrower than customers assume — and sometimes nothing at all.
Over-collateralised on-chain lending, assessed from liquidation risk inward
The model that outlasted the last cycle, assessed the way this desk assesses everything: what can go wrong first, and how much of it the published parameters actually tell you.
Over-collateralisation is the only lending model that survived the cycle
The crypto lending sector ran two models. One of them is largely gone, and the reason is structural rather than accidental.
