Layer 5 · a worked example

A Worked Example

How coupled credit could actually run — shown in full, including the hard parts.

Why this page exists A layer this important cannot rest on a promise that it works. So here is one concrete way it could actually be built — deliberately specific, and deliberately honest about the parts that are hard. This is not the only design, and not a final one. It is a worked example, offered so the mechanism can be inspected rather than trusted. Where it reaches outside the protocol, or leans on something softer than the base, that is named plainly, not hidden.

The binding principleEvery instrument must expose its path to the ground

Before any mechanism, the rule that makes the mechanism safe. Layer 5 permits real creativity — loans, collateral, interest, netting, even pooling and securitization higher up. It permits all of it under a single, unbreakable constraint: every credit instrument must expose its settlement path downward. No matter how elaborate an obligation becomes, anyone must be able to open it and trace it, step by step, to a determinate settlement on the base.

This is the exact inverse of the failure that took down the tower in 2008, where claims were stacked on claims until no one could see what any of it finally rested on. Here, that opacity is forbidden by construction. Elaboration is allowed; detachment from the ground is not.

The division of laborBitVM is the court, not the economy

The most important thing to understand about this design is what BitVM does not do. It does not process every payment, recompute every interest accrual, or update every collateral ratio. Asking it to would defeat the whole architecture. BitVM is not the economy. It is the court.

Think of how ordinary contracts work. Two businesses sign an agreement and simply perform it; the vast majority never see a judge. The court matters not because it is used constantly, but because both parties know it is there if one of them defects. BitVM provides exactly that, cryptographically: execute the agreement however you like, off-chain, at the speed of commerce — but if you assert a state that violates the agreed program, someone can prove you wrong, and the false assertion is punished. Layer 5 is where people transact. BitVM is where disputed truth gets adjudicated. As the seam page puts it: the base does not execute, it adjudicates.

The exampleAlice borrows against Bitcoin

Take a single, concrete case, and follow every step.

Alice posts 1 BTC as collateral and borrows 50,000 JU worth of credit. The contract encodes deterministic rules, agreed in advance:

  • the collateral ratio must stay above 150%;
  • the obligation is denominated in JU, valued from the published Layer 4 energy index;
  • the BTC-to-JU price is read from specified market data;
  • interest accrues by a fixed formula;
  • if the collateral ratio falls below 125%, liquidation becomes permissible;
  • repayment releases the BTC; default transfers just enough collateral to make the lender whole.

Almost none of this belongs on Bitcoin. An ordinary L5 credit engine computes it continuously, off-chain. The base is never asked to do the work — only, if needed, to judge it.

Here is how the judging arises. The operator running the contract asserts a state transition — for instance: "As of epoch N, Alice's collateral ratio fell to 121%; therefore, under contract C, liquidation is permitted." If everyone agrees, almost nothing happens on-chain; the assertion simply stands. But suppose a party disputes it — "No: you used the wrong JU index value," or "The threshold was never actually crossed." Now the computation can be challenged, and this is precisely where BitVM enters. The operator has committed, in advance, to every step of the calculation; the challenger need only point to the one step that is wrong; and Bitcoin's minimal base can check that single disputed step and punish a false claim. The court convenes only for the contested point, and only when there is a contest.

The honest partTwo oracles — and they are not equally hard

Here is where a lesser account would wave its hands, so here is where this one slows down. The contract consumes two observations from outside the protocol, and they are different in kind. A reader deserves to know which is solid and which is soft.

The JU energy index. This is the Layer 4 measure — published each epoch alongside a commitment to its source data, so that anyone can recompute it from the public basket and catch a discrepancy. It is the harder of the two, because it is anchored to the physical, adversarial, continuously metered reality of energy prices. It is difficult to fake for the same reason the lights staying on is difficult to fake.

The BTC-to-JU market price. Computing Alice's collateral ratio requires knowing what her 1 BTC is worth — and Bitcoin's worth is a market price, not an energy cost. This is a second oracle, and it is frankly the softer one: BTC's price is reflexive, a creature of the belief-driven top of the stack, exactly the kind of number the architecture is otherwise wary of. An honest implementation must treat it with the same discipline as the JU oracle — many independent sources, published, challengeable — while admitting it can never be made as hard as the energy anchor, because a reflexive price never can be. This is the softest link in this particular instrument, and pretending otherwise would be the very magic trick this page exists to refuse.

What BitVM does with these observations is the crucial boundary, and it is worth stating exactly: BitVM does not decide what reality is. It verifies that the agreed rules were correctly applied to the agreed observations of reality. It cannot know, by itself, that today's true JU reference is any particular number — that value originates outside Bitcoin, at Layer 4. What it can do is guarantee that once the observation is fixed, no one can lie about what the contract does with it.

The instrumentA credit object, not another floating token

Notice what the loan does not create: a new, freely circulating token spawned to be speculated on. Instead, imagine a credit object — an inspectable record that carries its whole life inside it:

  • Principal: 50,000 JU
  • Collateral: 1 BTC
  • Maturity: epoch 8,240
  • Interest rule, margin rule, settlement condition: formulas I, M, S
  • Contract hash: H

This object can be transferred, pooled, cleared, or securitized by the layers above — L5 does not forbid sophistication. But the settlement path travels with it and cannot be stripped off. However far up the stack it goes, anyone can open it and trace the same chain: claim → JU denomination → settlement rule → BTC collateral → Bitcoin. That is value grounds downward, obligations settle home, made literally true of a single instrument.

Closing the loopThe credit is extinguished, not retained

And when the obligation is met, the credit object does not linger as one more claim in circulation. It is extinguished. Repayment releases the collateral and destroys the credit; default resolves it through liquidation and likewise ends it. Either way, the instrument passes out of existence when its work is done. This is the discipline of the whole layer, made concrete: the credit is a wave, not a reservoir. It carries value across the field and then it is gone, having carried the thing it was made to carry — leaving no residue to pool into a hoard.

The relationships, drawn correctlyWhat "coupled" actually means

It is easy to blur the two relationships this instrument has to the layers below it, so here they are kept sharp, because the difference is the whole point:

  • The credit is denominated in the JU — a measurement relationship. The JU is the coordinate the obligation is expressed in; it is not the collateral, and the credit is not "backed by" it.
  • The credit settles through BitVM, down to Settlement Finality, the bearer asset, and Proof of Work — a settlement relationship. This is the downward path the obligation travels to come to rest.

So the word coupled is precise, and it is not the same as "backed." The credit is not backed by the JU, which only measures it. It is not even, strictly, backed by BTC in the strongest sense the stack knows — because BTC collateral terminates in a market price, and a price is a belief, not a synthetic invariant. What makes this credit coupled is something more robust than any backing: it remains bound, at every step, to a deterministic path toward settlement that anyone can inspect and no one can erase. That is a real and honest claim — and it is worth being clear that it is a different, and slightly humbler, claim than terminating in pure physical cost. Collateral in a reflexive asset buys a determinate path, not a synthetic floor. The stack does not pretend those are the same.

Why it mattersFreedom above, without detachment below

Set against how modern credit actually works, the difference is stark. A bank balance today can sit at the end of a long chain — deposit, to bank liability, to loan, to security, to repo, to derivative, to collateral transformation, to another liability — each link free to spawn another representation of value, until what anything finally rests on has been lost from view. This design permits the same creativity and forbids the same disappearance. Credit may expand upward — pooled, traded, layered — without ever becoming ontologically detached from the settlement beneath it. The upper layers keep enormous freedom. What they cannot do is erase the obligation to settle.

That is precisely the problem the whole stack exists to solve, and it is this layer's move in the infinite game: the web of credit is allowed to grow, to reach more players and carry more of the world — but it can never float free of the ground, because every strand of it, followed down, still ends on spent work no one can rewrite.