Two halves of secured credit, on one record.
Traditional finance and decentralised finance each have half of what secured credit needs.
Traditional finance has the legal half. The law says how a pledge, charge or security interest is created, how third parties learn of it, how it ranks, when the creditor may enforce, how the asset is sold, and who receives what from the proceeds. What it lacks is a shared record: each lender, custodian, servicer and trustee keeps its own books, reconciled at cost after the fact.
Decentralised finance has the other half: one open record that anyone can read and settle against. What most of its lending lacks is any resemblance to the law of secured credit: a position is liquidated when a price feed crosses a line, and nothing in the mechanism ranks a second lender or returns a surplus to the borrower as a matter of course.
| attachment | a lien binds one vault lock to one set of obligations |
|---|---|
| publicity | the lock, the lien and the obligations are public on chain |
| priority | a locked asset cannot secure a second set, and a second lender takes a charge over the borrower's position, ranked behind the first |
| default, then enforcement | breach only after the due date and a grace period fixed at creation, enforcement by a rule anyone can trigger |
| disposition | under sale, at venues named when the lien was created, within a window fixed then |
| application of proceeds | one waterfall, every creditor paid pro rata |
| surplus | what is left goes to the holder of the borrower's position |
The first deployment's lien takes the collateral in kind. Sale, the waterfall and a surplus from a sale are features of a variant, the full lien; the second charge, of the nested lien.
The claim is not that "code is law". It is narrower: that the mechanical core of secured credit — attachment, publicity, rank, default, sale and distribution — can be carried out by code that does what the law of secured credit already sets out in outline, so that the outcome on chain and the outcome a court would recognise can coincide.
Where the law depends on judgement — a stay on enforcement in insolvency, whether a sale was commercially reasonable, the protections owed to consumers, sanctions and identity requirements, a court's view of what a transaction really was — the protocol does not pretend to supply it. That belongs to the legal agreement that refers to the record and to the orchestrators built above the core.
Parties need not know each other. A bond issuer does not know who holds its bonds; the instrument and the record stand between them. Terms are set by auction; the creditor's right is a transferable claim; the borrower's equity is a transferable position; and the collateral, held under a rule, stands where a central counterparty's balance sheet would otherwise stand.
Compliance lives at the edges, not in the core. Institutions that must know their counterparties can do so where they meet the record: tokens that carry their own transfer allowlists, and orchestrators that admit only verified participants to a given market. These write to the same record as fully open markets.
Why neutrality matters here. A shared record is only useful if each side can trust it without trusting the other side's vendor. A protocol that charges no fees, has no owner with discretion over it and cannot be changed under its users is the kind of infrastructure two rival institutions can both adopt.
The limits. A shared record is valuable in proportion to its use; until both sides use it, it is a record of one side. Other shared ledgers for secured credit are being built, most of them permissioned and owned by a consortium or a single firm. And the on-chain record governs the token: whether the token carries the asset behind it is a matter for the issuer, the custodian and the legal agreement, which the protocol cannot replace.
Status: Specified.