Explanations > Settlement model

Settlement model

Explain how Quovra resolves markets and pays winners.

Settlement model

Quovra separates market resolution from payout.

Why separate settlement and claim?

The settle instruction verifies proof and records the winner. The claim instruction transfers payout to one winner.

This avoids a compute-heavy transaction that both verifies a proof and loops over all winners.

Invariant

text

No valid TxLINE proof -> no market resolution -> no winner payout

Payout formula

text

payout = stake + stake * losing_total / winning_total

Dust can remain in the vault because payouts use integer division.

State transitions

mermaid

stateDiagram-v2
  [*] --> Open
  Open --> Closed: close_ts reached
  Open --> Funded: deposit
  Funded --> Closed: close_ts reached
  Closed --> Resolved: settle(valid proof)
  Closed --> Closed: settle(invalid proof) reverts
  Resolved --> Claimed: claim(winner)

Tradeoffs

Tradeoff	Decision
Faster payout vs safer compute	Pull payout
Trusted resolver vs proof verifier	Proof verifier
Real money vs safe demo	Devnet test USDC
Permissioned operator vs permissionless keeper	Anyone can submit a valid proof

Search keywords

settlement model, pull payout, proof, escrow, claim

Last updated: [YYYY-MM-DD]

Settlement model