White paperCollateral

Bitcoin as institutional collateral

A framework for treating BTC as pledgeable collateral: custody, rehypothecation, haircuts and insolvency treatment.

By Dylan CarterFebruary 2026Executive summary

The framework

The paper sets out a framework for a lender, a prime broker or a credit committee deciding whether and how to accept Bitcoin as collateral. It organises the decision into four questions, control, reuse, discount and failure, and argues that the first and the last are the ones that determine eligibility, while the middle two determine price.

Custody: who controls the asset

A lender’s first question is whether it controls the collateral well enough to sell it on default. The paper compares the arrangements in use: a qualified custodian holding the asset for the borrower with a control agreement in the lender’s favour; a tri-party structure in which an agent holds the asset for both; and multi-signature arrangements in which the custodian, the lender and the borrower each hold a key and a transfer requires an agreed quorum. It argues that the multi-signature arrangement is the one that adds something the securities world does not have, because the asset itself enforces part of the agreement, and that this is likely to be the basis on which the most favourable terms are eventually offered.

Rehypothecation: whether the asset can be reused

The paper treats the reuse of collateral as the question on which Bitcoin differs most from securities in practice. Because the asset can be moved without an intermediary, a lender who is permitted to reuse it can do so immediately and across borders, and the borrower’s exposure to the lender’s own solvency rises accordingly. The paper sets out the terms a borrower should insist on, segregation, a prohibition on reuse without consent, and a right to verify on-chain that the asset remains where the agreement says it is, and argues that the last of these is a protection no securities borrower has ever had.

Haircuts: how the discount is set

The paper decomposes a haircut into the three risks it compensates for, price movement before the position can be closed, the cost of selling in a hurry, and uncertainty of control, and argues that the three should be set separately and disclosed separately, so that improvements in custody and liquidity are reflected in the terms rather than lost in a single number. It sets out no figures, because haircuts vary by lender, jurisdiction and borrower, and it argues that a published haircut is less informative than a published method.

Insolvency: what happens when a party fails

The paper’s longest section asks what each party holds when the borrower, the lender or the custodian becomes insolvent. It works through the perfection of a security interest in a digital asset, the treatment of a custodian’s holdings as client property or as the custodian’s own, and the enforceability of an on-chain transfer made in the hours after a failure. It argues that the law in this area is being settled case by case and jurisdiction by jurisdiction, that the documents can do a great deal to narrow the uncertainty in the meantime, and that a lender who has not read the custodian’s terms of business has not assessed the collateral.

Conclusion

The paper concludes that Bitcoin is already pledgeable collateral for lenders who have done the structural work, that the work is mostly legal and operational rather than technical, and that the terms on offer will converge toward those for volatile securities as the custody market matures and the case law accumulates. It expects the collateral use to precede, and to lead to, the use of the asset for settlement by the same institutions.

The full paper

The full paper is available on request.

Request the paper

More papers

More papers