Sources and context
Edited from our 2025 research “Profit By Waiting: AMMs for Redeemable Derivatives”, the design work behind the liquid staking marketplace.
This edition: .
Some assets can be redeemed for an underlying asset, but only after a delay. A holder who wants to exit immediately may accept a discount. A counterparty willing to wait can buy the claim and follow the redemption process.
This is a market in time preference as well as price. The engineering question is how to organize capital around the delay without hiding the associated liquidity and settlement risks.
Waiting has a cost
Capital committed to a redemption queue cannot be freely deployed elsewhere. The appropriate comparison is not simply the nominal yield of another asset. It is the return available over a comparable holding period, with comparable risk and access to liquidity.
A useful model separates the purchase discount, transaction costs, expected settlement time, and the possibility that settlement takes longer than expected. It also identifies exposure to the underlying asset, the redemption mechanism, and any lending venue used alongside the strategy.
Pool architecture changes the problem
An individual market maker may need to wait for a redemption to mature before reusing its capital. A pool can also receive deposits and trading flows from participants moving in the opposite direction. In some conditions, that incoming demand can replenish its liquid inventory earlier.
The possibility of faster rebalancing is valuable, but it should not be assumed during stress. A design that works well with balanced inflows and outflows still needs a plan for one-sided demand.
Our research examined how an AMM can integrate a redeemable asset’s withdrawal process directly. Instead of relying only on secondary-market trades to rebalance, the pool can use the underlying redemption route. That changes both the pricing problem and the accounting required to represent an LP’s position.
Liquid and pending are different states
A pool may hold immediately available reserves alongside claims awaiting settlement. Reporting one undifferentiated balance conceals that distinction. Deposits, exits, and valuation need rules for allocating the liquid and pending portions fairly.
An instant exit mechanism transfers a waiting position to another willing holder. It does not make the underlying waiting period disappear.
The terms of that transfer matter: pricing, available liquidity, and the treatment of remaining participants. If there is insufficient liquid capacity, a system must represent the remaining claim accurately rather than imply an unconditional immediate withdrawal.
A broader design pattern
The same questions arise in redeemable vaults, tokenized claims, and other systems with delayed settlement. Each asset has different legal and technical constraints, so a shared architecture does not imply interchangeable risk.
The recurring engineering task is to make time explicit: model settlement, separate available inventory from pending claims, and test what happens when the expected flow of new capital slows down.