For an institution placing bitcoin with a custodian, rehypothecation is the risk that the assets it pledged are quietly reused to back someone else's obligations. The term has a precise meaning in traditional finance, where reuse is permitted within limits, and a sharper edge in digital-asset custody, where weaker controls let reuse run unseen. This article is educational and is not legal or investment advice; custody and collateral decisions should be made with qualified counsel.
Hypothecation vs. rehypothecation
The two terms describe two links in the same chain, and confusing them is the most common error in this topic.
Hypothecation happens when an owner pledges an asset as collateral while keeping ownership of it. A margin trader who pledges securities to back a loan has hypothecated those securities. The lender now holds a claim against the collateral if the borrower defaults, but the borrower still owns it.
Rehypothecation happens one step downstream. The party that received the pledged collateral, a broker or custodian, takes that same collateral and pledges it again, this time for its own funding or trading needs. The client's asset is now backing two obligations at once: the client's original loan, and the custodian's separate borrowing. The client often never sees this happen and has not specifically approved each reuse.
| Property | Hypothecation | Rehypothecation |
|---|---|---|
| Who acts | The asset's owner | The party holding the pledged collateral |
| What happens | Owner pledges an asset as loan collateral | Holder reuses that collateral for its own borrowing |
| Ownership | Stays with the original owner | Stays with the original owner, who now bears added risk |
| Visibility to the owner | Direct, the owner initiates it | Often opaque, buried in account agreements |
| Order in the chain | First pledge | Second-order reuse of the same collateral |
The risk that matters is the second-order reuse. When a custodian rehypothecates client collateral, the client's asset is exposed to the custodian's solvency, not just the client's own credit. If the custodian fails while the client's asset is pledged somewhere else, the client joins a queue of creditors trying to claw back property that has already moved. The same reuse mechanics shape how lending against bitcoin is structured, covered in Bitcoin-backed loans.
Where rehypothecation is legal, and where it is limited
Rehypothecation is a normal part of prime brokerage and securities lending. It supplies liquidity to markets and lets brokers finance their operations against collateral they already hold. The question regulators have wrestled with is not whether to allow it, but how much reuse to permit and how to protect clients when a firm collapses.
The US model: a hard cap and a lock-up
In the United States, the SEC's Customer Protection Rule (Rule 15c3-3) draws a line between two kinds of client securities. A broker must take fully-paid securities and excess-margin securities and lock them up in its physical possession or in a designated control location, segregated from the firm's own assets (17 CFR 240.15c3-3). Those securities cannot be reused at all.
The broker is permitted to rehypothecate a client's margin securities, but only up to a ceiling. Federal Reserve Regulation T and Rule 15c3-3 together cap the value a broker may re-pledge at 140 percent of the client's debit (margin loan) balance. A client who borrows 100 dollars on margin can have at most 140 dollars of their securities reused (17 CFR 240.15c3-3). The cap exists so that the amount of client property put at risk stays tethered to the size of the loan the client actually took.
The UK model: historically no statutory cap
The contrast with the United Kingdom is instructive. Historically, UK law set no statutory limit on the amount of client assets a prime broker could rehypothecate. Reuse was governed by contract rather than by a fixed regulatory ceiling, which meant a broker could, by agreement, re-pledge the full value of a client's pledged assets or more.
The consequences became concrete when Lehman Brothers International (Europe) entered insolvency in 2008. Client assets that had been rehypothecated under UK arrangements were caught in the failed entity's estate, and the clients who had pledged them spent years as unsecured creditors trying to recover property that had been reused (IMF working paper, Singh and Aitken). The episode is the standard cautionary case for why disclosure, limits, and segregation matter, and it pushed both regulators and clients toward tighter contractual controls on reuse.
The lesson that carries into digital assets is simple: the protection a client actually has depends on what the custodian is permitted to do with the asset, and on whether the client can verify what the custodian is doing.
Why rehypothecation is the hidden risk in digital-asset custody
Securities markets have decades of rules, reserve formulas, and audit infrastructure built around reuse. Much of the digital-asset custody market grew up faster than that scaffolding, and several custodians during the 2022 market stress operated with weaker controls than a regulated broker would.
The recurring pattern in those failures was commingling. Client deposits and the platform's own funds flowed into the same pool, and that pool was lent out, used as trading collateral, or moved to affiliated entities to chase yield. Clients believed their funds were held one-for-one and available on demand. In reality the funds had been reused, and when withdrawals spiked, the platform could not produce what it had promised to hold.
What made the reuse invisible was the absence of independent verification. A client of a traditional custodian relies on audits, regulatory examinations, and the firm's own statements; for how those examinations work, see What SOC 2 and audits mean for Bitcoin custody. A client of a digital-asset platform often had only the platform's word. When the only evidence that funds are segregated is the custodian's own attestation, rehypothecation can run for a long time before anyone outside the firm can see it.
"Trust us" attestations vs. cryptographic proof
A traditional attestation is a point-in-time statement, frequently from the custodian itself or a third party the custodian hired, that the books balance. It can be honest and still miss reuse that happens between attestation dates, and it asks the client to trust the attester. The client learns what someone reported on a given day rather than what the chain shows continuously.
A public ledger changes the evidence available. When a custodian holds client funds in segregated on-chain accounts, the holdings in those accounts are recorded on a ledger anyone can read. The custodian cannot quietly move funds out to back its own borrowing without the movement appearing as a transaction. Instead of a contract clause that a client has to take on faith, reuse becomes visible to anyone who checks the addresses.
How segregation plus proof of reserves makes reuse provably absent
Two design choices, used together, turn "we don't reuse your funds" from a promise into something a client can check.
The first is segregation. Each client's or each purpose's funds sit in their own on-chain account rather than in a shared pool. Segregated accounts mean there is no commingled balance for a custodian to quietly draw against, and they make it possible to attribute every satoshi to the account it belongs to.
The second is proof of reserves. Because the segregated accounts are on a public ledger, the custodian can demonstrate control of specific addresses and the balances those addresses hold, and an outside party can confirm that the on-chain reserves match what the custodian claims to be holding for clients. If funds had been rehypothecated out of a segregated account, the on-chain balance would no longer match the claim, and the discrepancy would be visible to anyone running the check.
Proof of reserves on its own proves what a custodian holds, not what it owes; a complete picture also needs proof of liabilities. Both are covered in What is proof of reserves?. For the rehypothecation question specifically, the combination of segregated on-chain accounts and verifiable reserves is what lets a client move from trusting an attestation to checking the ledger. Whether a custodian holds qualified custodian status, examined in What is a qualified custodian?, is a separate question from whether its reserves can be verified on-chain.
Blockstream Enterprise applies this model in its Custody Engine. Client funds are held in segregated accounts, every transaction is traceable to a user and role through a cryptographic audit trail, and reserves are verifiable on-chain rather than merely attested on paper. Because the reserves are recorded on the ledger, reuse of segregated funds cannot happen silently: it would change the on-chain balance and break the match between the reserves and the custodian's stated holdings. The governance layer reinforces this by separating who can draft, approve, and authorize a transaction, so no single operator can move client funds unchecked. Details are at blockstream.com/enterprise.
The Liquid Network adds a privacy dimension that institutions often need alongside verifiability. Confidential Transactions hide transaction amounts and asset types from outside observers while still allowing selective disclosure to auditors and regulators, so a custodian can keep client positions private without giving up the ability to prove reserves to the parties entitled to verify them.
Frequently Asked Questions
What is rehypothecation in simple terms?
Rehypothecation is when a broker or custodian takes collateral a client pledged to it and pledges that same collateral again for its own borrowing or trading. The client still owns the asset, but it is now backing two obligations at once, which means the client is exposed to the custodian's solvency on top of their own.
Is rehypothecation illegal?
No. Rehypothecation is legal and common in traditional finance, particularly in prime brokerage and securities lending. In the United States it is limited: fully-paid securities must be locked up and cannot be reused, and reuse of margin securities is capped at 140 percent of the client's margin loan balance under SEC Rule 15c3-3 and Regulation T. The risk arises when reuse happens without disclosure, without limits, or without a way for the client to verify it.
What is the difference between hypothecation and rehypothecation?
Hypothecation is the owner pledging an asset as collateral while keeping ownership of it. Rehypothecation is the next step, where the party that received that pledged collateral reuses it by pledging it onward. Hypothecation is initiated by the owner; rehypothecation is done by the holder, often without the owner's direct involvement.
How did rehypothecation contribute to the 2022 digital-asset custody failures?
Several platforms commingled client deposits with their own funds and reused the pooled balance for lending and trading, frequently without clear disclosure. Clients assumed their funds were held one-for-one and available on demand, but the funds had been reused, so when withdrawals surged the platforms could not return what they claimed to hold. The shared cause was a lack of independent verification: clients had only the platform's word that funds were segregated.
Can rehypothecation be prevented?
Rehypothecation can be made provably absent rather than merely promised against. When a custodian holds client funds in segregated on-chain accounts and publishes proof of reserves, reuse of those funds would change the on-chain balance and break the match between the reserves and the custodian's stated holdings. Any outside party can run the check, so reuse becomes visible instead of hidden.
What is the difference between an attestation and proof of reserves?
An attestation is a point-in-time statement, often from the custodian or a firm it hired, that the books balance, and it asks the client to trust the attester. Proof of reserves on a public ledger lets the custodian demonstrate control of specific addresses and their balances, which an outside party can verify directly against what the custodian claims to hold. The first reports what someone said; the second can be checked on-chain.