Counterparties and settlement

Who bears the risk between trade and performance — and what takes the clearing house's place when there is none.

This lesson has had no expert review. It was written for this platform and against the evidence it cites; nobody has gone through it independently.

Learning objectives

  • You can explain which risk arises between trade and performance.
  • You can name what atomic settlement solves — and what it removes with nothing in its place.

Check your prior knowledge

Answer these for yourself before reading on. Wherever you hesitate is where this lesson pays off.

Core concept

Fact

The window between trade and performance

Between the moment two parties agree and the moment both have delivered lies a period. Inside that window each side bears the risk that the other fails after it has performed itself. That is no edge case: it is the reason clearing houses, margin and the delivery-versus-payment rule exist at all.

Fact

A clearing house does several things at once

It steps between both sides as counterparty, so neither is exposed to the other. It nets opposing positions, so only the balance has to move. It requires collateral, which it realizes on a default. And it allocates a residual loss to the remaining members by fixed rules. Four functions, one bearer — and they are routinely thought of as one.

Interpretation

Atomic settlement removes the window, not the risk

On-chain, a swap happens inside one transaction: either both sides change hands, or neither. The risk from the window genuinely disappears — a real improvement over multi-day settlement. What also disappears are the other three functions: nothing is netted, there is no shared collateral management, and nobody allocates a residual loss. Seeing only the first point mistakes a subset for a replacement.

Definitions

Counterparty risk
The risk that the other side fails to perform its obligation.
Delivery versus payment
A procedure in which delivery and payment take effect only together.
Atomic swap
A swap that happens completely within one transaction, or not at all.

Model

  1. Become the counterparty → unnecessary, because the window is gone

  2. Netting of positions → no counterpart

  3. Collateral management → per protocol, not shared

  4. Allocation of a residual loss → set per protocol, not across them

Four clearing-house functions and their on-chain counterpart — Only the first line is solved; the other three are displaced or open.

Worked example

The same trade, two settlements

Traditional
T+2, clearing house as counterparty, margin posted
On-chain
atomic, no third party involved
Other side defaults on day 1
traditional: the clearing house performs; on-chain: the case does not arise

For this single trade the on-chain variant is superior: no window, no default, no margin required. Once a hundred trades a day run between the same parties, though, netting is missing — instead of one balance, a hundred gross amounts move, each with its own transaction cost.

Better for the single case, more expensive in operation.

Reading: The useful question is not “which is better” but which of the four functions the organization actually needs.

Retrieval

A swap happens atomically. Which risk does that remove?
Which clearing-house function has no on-chain counterpart?

Exercise on real data

Look up how a transaction proceeds on a chain, and locate where the settlement window would sit in the traditional system.

Read up on the basics →

Application

An organization settles many small trades daily with the same counterparty. What follows from the absence of netting?

Related case studies

Institutional reading

Bank
Which of the four functions would the institution itself perform under on-chain settlement?
Advisory
Does the client know that if an error occurs, nobody unwinds the trade?

Key takeaways