Loading
Exit register
Vault dashboards lead with size and yield. This measures what appears withdrawable at the current state: per market, the smaller of what the vault supplied and what that market has left unlent, summed, read from Euler, Morpho and Lista.
50 days of the register, recorded daily, from 18 August 2026. That is short, and it is short because we started recording it then rather than because nothing happened before. This is a reading we took and stored rather than an event a chain kept, so the series itself begins when we began; the underlying activity may well be reconstructable from an archive node or the venue’s own history, and we have not done it.
Cannot be withdrawn
$1.05B
86 of 618 vaults in the register
Supplied to lending
$8.56B
gross, across the vaults we read
Withdrawable, each vault alone
$1.80B
21% of what was supplied; not all at once
Register
51 days
recorded daily since 2026-08-18
If everybody asked at once
An hour is not a scenario: it is what the register says is free now, with no model behind it. A day and a week add how fast each vault\u2019s own free liquidity has actually come back, taken from its own history in this register and never from a market-wide rate.
Within an hour
$1.76B
free now, measured
Within a day
$1.77B
plus one day at its own rate
Within a week
$1.79B
plus seven days at its own rate
Waiting buys less than it sounds. The median vault\u2019s free liquidity grew by <0.01% a day over the register, and for half of all vault-days nothing came back at all. Over 481 vaults holding $8.33B, each with at least 14 days of its own history. A further 148 vaults holding $39.81M free are NOT in these figures: they have too little history here for a rate of their own, and the market median is not their history.
What is not here. Of 3,649 open vaults holding $11.34B, 2,809 holding $728.25M have no exit reading at all, neither what can be withdrawn nor what sits unlent. That is 6% of the capital on this site, and the figures above say nothing about it. The venues behind most of it are Concrete, Hyperliquid, ether.fi Liquid, Lagoon and Veda, where withdrawal is asynchronous or on a timer and the reading we have elsewhere does not apply.
All at once is less. Across the $7.51B we read market by market (Morpho and Lista), each vault on its own could take out $2.72B; if every one of them left together, the 232 markets they share would give up at most $965.54M, because $1.76B of free cash is counted once per vault that could claim it. Queue order, reallocation and withdrawal caps are not modelled: this is the free-cash ceiling, not a forecast.
Why this is not in the TVL figure
A lending vault’s size and its liquidity are different facts. Venues show both for their own vaults; this page puts the liquidity side by side across every lending vault we read, which no single venue does. Utilisation here is computed from the two halves of one reading, lent out over lent out plus left, never against a capital figure written by another job at another time, which is a wrong number built from two right ones.
A vault absent from this list is one whose liquidity we do not read, never one with nothing locked. Where the reading is missing the site says so rather than showing a zero.
Can the money behind the best rates leave?
The venue’s own advertised rate, beside what its own markets say could be withdrawn today. Both numbers are published elsewhere on this site; together they answer a question neither answers alone.
The highest advertised rates are on vaults nothing can leave
7 of the 40 highest-yielding vaults we can read both figures for have nothing withdrawable at the last reading, holding $4.59M between them. An advertised rate is a claim about what is being earned; it says nothing about whether it can be taken, and an accrued claim that cannot be redeemed is not a return anybody has had.
| Vault | Advertised rate | Size | Could leave today | Of the vault |
|---|---|---|---|---|
| Vaultik USDCMorpho | 297,995.8% | $255.8k | $0.00 | 0.0% |
| Duplicated KeyMorpho | 297,995.8% | $2.11M | $0.00 | 0.0% |