NoteInfrastructure
Layer two, read as plumbing
What Bitcoin scaling networks actually change for settlement time, cost and finality, without the token narratives.
Scaling networks built on Bitcoin are usually discussed as products, each with a name, a community and sometimes a token. This note reads them as plumbing instead. The question it asks of each is narrow: what does it change about the time, cost and finality of settlement, and what does it leave unchanged?
Three properties, separately
Settlement on the Bitcoin base layer has three properties that a scaling network can alter. Time: how long before a transfer is confirmed and then considered final. Cost: what a transfer pays to be included. Finality: how certain the transfer is, and who or what that certainty rests on. The base layer trades the first two for the third. It is slow and, at busy times, expensive, and in exchange its finality rests on nothing but the chain itself.
Every scaling network moves along that trade. The firm’s rule for reading them is to ask, for each, what finality now rests on. If a transfer is final because the base chain has confirmed it, the network has inherited Bitcoin’s finality. If it is final because a set of operators has signed it, or because a challenge period has passed without objection, the network has introduced a new assumption, and the assumption should be named.
Payment channels
A payment channel locks funds on the base layer and lets two parties update the balance between them off-chain, settling to the chain only when the channel closes. Time and cost fall to nearly nothing for each update. Finality rests on the base layer plus one new assumption: that each party is watching the chain, or has delegated the watching, so that an attempt to close on an old balance can be contested. For payments between parties who already transact repeatedly, that assumption is light. For settlement between institutions that do not, it is a design constraint rather than a flaw, and it tells you what the network is for.
Sidechains and federated networks
A sidechain holds Bitcoin in a structure controlled by a federation of operators and issues a representation of it on a separate chain with its own rules. Transfers are fast and cheap, and the chain can support instruments the base layer cannot. Finality rests on the federation: the transfer is final when enough operators say it is, and the Bitcoin backing it can be returned when enough of them agree. That is a trust assumption of a familiar kind, closer to a clearing house than to a public chain, and it should be assessed as one. Who are the operators, how many must agree, under what law, and what happens if they disagree?
Rollups and validity proofs
A newer class of network posts data or proofs to the base layer and relies on it to order and anchor them. Where the base layer can verify a proof that a batch of transactions is valid, finality can rest on the chain again, with no operators to trust. Where it cannot, and Bitcoin’s scripting rules limit what can be verified, the network falls back to a challenge period or to a committee. The firm watches this area closely because it is where the inherited-finality case is strongest in principle and least settled in practice. The question to ask of any design is the same: what, exactly, is the base layer verifying?
What the plumbing reading leaves out
It leaves out the token. Most scaling networks do not need one to function, and where one exists its purpose should be stated and tested against the plumbing: does it secure the network, pay for its operation, or merely fund its promoters? The firm’s interest is in networks that extend Bitcoin’s settlement properties to more transactions and more instruments. A token that does not serve that end is a distraction from the engineering, and the engineering is the thing being bought.
Read this way, the scaling networks are not competitors to the base layer or to each other. They are different answers to the same trade, suited to different uses: channels for repeated payments, federations for instruments the base layer cannot carry, proof-based designs for settlement that inherits finality. The firm’s view is that institutional settlement will use all three, and that the one that matters most is the one whose finality an institution can explain to its own risk committee.