Research note · Collateral

Collateral, expressed as state.

NotCentralised·7 min read

A repo is a promise to repurchase. Express that promise as a single stateful instrument and three things follow: collateral that can move, rehypothecation you can trace, and small-ticket repo that finally pays for itself. This note walks the mechanism end to end.

What a repo is — and why it is expensive

A repurchase agreement is a sale paired with a binding agreement to buy the asset back: financing dressed as a trade. The economics are simple; the operations are not.

Custody of the collateral, margining, enforcement on default, and tracking who re-pledged what to whom — that machinery is a fixed cost. It is why repo concentrates in large, standardised tickets and stays out of reach for smaller or bespoke exposures.

The repo as a composed contract

In YieldFabric the lender does not hold a stack of documents and a custody relationship. The lender holds a single stateful instrument — a composed contract bundling the pledged collateral, held in escrow, and the right to the repurchase payment.

It encodes a defined fork — repurchase by expiry and the collateral releases, or the collateral transfers to the lender. The outcome is rule-based, not discretionary: no notice period, no enforcement action, no intermediary holding economic control. The legal right is unchanged; it is the operation of that right that becomes deterministic.

Instrument
Composed contract

The single stateful instrument the lender holds — collateral plus the right to the repurchase payment, bundled.

Leg
Pledged collateral

The tokenised asset locked in escrow for the term of the repo.

Leg
Repurchase right

The lender’s claim on the repurchase payment — itself a clearly-defined token.

Rule
The defined fork

Repurchase by expiry and the collateral releases; or it transfers to the lender. No discretion.

Capability
Re-pledge

The repurchase right can be pledged as collateral for the next repo, without moving the underlying.

The lifecycle, end to end

A repo is bilateral — an obligor and a lender — and runs as a short, deterministic sequence. Because the outcome is encoded up front, settlement is the contract resolving its own fork, not a negotiation.

  1. 01
    Pledge

    The obligor locks tokenised collateral in escrow for the term of the repo.

  2. 02
    Fund

    The lender delivers the upfront payment via an atomic swap and receives a single composed contract — collateral in escrow plus the right to the repurchase payment.

  3. 03
    The forkRule-based

    The contract encodes two outcomes: repurchase by expiry and the collateral releases, or it transfers to the lender. Rule-based, not discretionary.

  4. 04
    Re-pledge

    Optional. Because the repurchase right is itself a token, the lender can re-pledge it as collateral for the next repo — collateral mobility without moving the underlying.

  5. 05
    Settle

    At expiry the fork resolves: repay and release, or transfer. The structure settles to zero; every layer still traces back to the originating obligor.

Rehypothecation you can trace

Because the repurchase right is itself a clearly-defined token, it can be re-pledged as collateral for the next repo — and the next. Each link is its own bilateral repo, yet every layer traces back to the originating obligor, who stays the root anchor of credit risk the whole way down.

Strictly, this is the tidier cousin of classic rehypothecation: each lender re-pledges its own position — the repurchase right it holds — rather than directly re-using the borrower’s collateral. Economically equivalent, legally cleaner — a matched-book, back-to-back chain.

Rehypothecation is precisely where collateral chains went dark in 2008: the same assets re-pledged through opaque chains, the true depth and encumbrance unknown, so when one link failed the downstream exposure was a black box and everyone scrambled to discover who was owed what. Here it is the opposite by construction — every pledge and re-pledge is an explicit, time-stamped on-chain event.

Default does not reintroduce the black box. When a link fails to repurchase, the fork at that layer resolves by rule — the contingent claim passes to the layer above — while every other layer still references the collateral it actually holds. A default propagates as a set of defined outcomes, not as a discovery process.

Inside a multi-layer chain

A worked chain shows what “traceable” buys you. One originating pledge funds several layers; none of the collateral moves physically; each layer is a defined fork over the layer beneath it, and all of them reference the same root obligor.

  1. 01
    Originating obligorRoot of risk

    Pledges the collateral and owes the first repurchase. Every layer above ultimately prices this party’s performance.

  2. 02
    Lender A

    Funds the first repo and holds the composed contract — collateral in escrow plus the right to the repurchase payment.

  3. 03
    Re-pledge → Lender B

    Lender A pledges its repurchase right. Lender B holds a claim contingent on Lender A’s repurchase, which is contingent on the obligor’s.

  4. 04
    Re-pledge → Lender C

    Another funding layer, still anchored to the same originating collateral — no asset has changed hands.

  5. 05
    Resolution

    At each expiry the layer’s fork resolves; a default at any layer resolves by rule. The obligor’s performance is what the whole chain ultimately settles against.

What you can trace

Because every pledge and re-pledge is an explicit, linked event, the questions that were unanswerable in a paper chain become answerable — some read off directly, some by walking the chain.

Provenance

Any layer back to the originating obligor — one hop per link, no reconstruction.

Depth

How many times the collateral has been re-pledged, and where each layer sits — read off the linked pledge graph.

Encumbrance

What is locked and what is free, per layer — no pledging the same collateral beyond what exists.

Concentration

Total exposure to the originating obligor — derivable by walking the chain, since every link is explicit.

History

When each pledge, re-pledge, and repurchase happened — every state append-only and versioned, replayable as of any recorded moment.

Small-ticket repo becomes viable

Strip out the discretionary enforcement and the intermediary that holds economic control, and the fixed cost of a repo collapses.

Bilateral repo and loan-level funding gain the safety of structured finance — defined seniority, limited recourse, collateral mobility — at scales where that protection was previously uneconomic. Spreads can anchor to asset risk rather than process or intermediary risk.

A defined fork
Repurchase by expiry, or the collateral transfers. Rule-based, not discretionary.
Collateral mobility
Re-pledge the contingent right across funding layers — no physical re-transfer.
Traceable risk
Every layer traces back to the originating obligor — the root anchor of credit risk.
Viable at small scale
No intermediary holding economic control, so the fixed cost collapses.
Proven in Project Acacia

In Project Acacia, the repo leg of a five-party collateralised-lending structure settled in wholesale central-bank digital currency on a public blockchain: tokenised collateral was locked in escrow, the lender delivered the upfront payment, and the repurchase resolved on-chain — rule-based, with no intermediary holding economic control. References are descriptive; no endorsement is implied.

From research to product.

Project Acacia was the pilot. YieldFabric is the platform — the same rails run private-credit facilities, securitisations, and bilateral deals today.