Bitcoin as collateral, and how a loan is liquidated

When bitcoin is pledged, it has to be held somewhere for the term of the loan, and who holds it determines what can happen to it. The arrangements sit on a spectrum. At one end is collaborative multisignature custody, where control is split across several keys and the borrower holds at least one, so no single party can move the collateral alone. In the middle is a qualified custodian — a regulated third party that holds the bitcoin segregated from the lender's own assets, so a lender failure does not automatically put the collateral into the lender's bankruptcy estate. At the far end is the lender holding the collateral in its own wallet, which concentrates every operational and counterparty risk in one place.

The second question is whether the collateral can be reused. Rehypothecation is the practice of a lender re-lending or otherwise deploying pledged collateral while the loan is open. It is a normal and regulated practice in parts of traditional finance; in the crypto lending market of the last cycle it was neither capped nor reliably disclosed. A lender that reuses collateral is exposed to its own other positions, and if those fail, the borrower's bitcoin can be caught in the failure even when the borrower has met every obligation under the loan. A written no-rehypothecation policy is the specific protection against that risk, and it is a different protection from custody: custody decides who can move the asset, and the loan contract decides whether the lender is permitted to deploy it.

The 2022 failures are the reason this section exists. Celsius, BlockFi, Voyager and Genesis froze withdrawals and entered bankruptcy proceedings between May and November of that year, and customer assets were caught in the wreckage. The cause was not primarily the price of bitcoin; it was opaque custody and the reuse of customer assets in other positions. The lesson the surviving market drew is that reused collateral is collateral that can be lost for reasons unrelated to the borrower's own loan-to-value, and that the questions to ask a lender are who holds the bitcoin, whether it is segregated, and whether it is reused — in writing, not in marketing language.

As the collateral value falls, the loan-to-value rises toward two thresholds. The first is the margin call, at which the lender notifies the borrower that the position needs attention. The borrower can respond by adding collateral or repaying part of the loan, and the agreement specifies a cure window in which to do so. The second is the liquidation threshold, at which the lender sells collateral to restore the required ratio. Some agreements provide for partial liquidation, selling only enough collateral to bring the loan back into compliance; others liquidate the whole position.

The mechanics are unforgiving in a specific way that is worth stating plainly. Bitcoin's price can move several per cent in a day, and a sharp move can cross both thresholds before a borrower has time to act. A cure window measured in hours is not much protection against a move that happens in minutes, and a liquidation executed into a falling market sells at the worst available price. The borrower's defence is not the cure window; it is borrowing at a low enough loan-to-value that the trigger price sits far below the current price, and keeping reserve collateral that can be posted quickly.

The regulatory treatment of bitcoin as collateral is still being settled, and it differs by market. In the United States, the SEC's Staff Accounting Bulletin 122, issued in 2025 to replace SAB 121, clarified that banks and broker-dealers may hold crypto for clients without recording it as a balance-sheet liability, which removed an obstacle to institutions accepting it. The CFTC has run a pilot programme permitting bitcoin, ether and USDC as eligible margin at registered futures commission merchants, with final rulemaking on custody standards and settlement finality targeted for 2026. Those developments concern derivatives margin rather than spot lending, but they establish the custody and segregation standards against which spot lenders are likely to be measured.

The practical summary is that a bitcoin-backed loan is a well-understood instrument when the collateral is independently custodied, not reused, and the borrowing is conservative, and a way to lose the bitcoin when any of those three conditions is missing. The liquidity page explains why a forced sale into a thin market realises less than the quoted price, and the corporate treasury page covers the related case of bitcoin held against issued debt rather than pledged against a single loan.