Loading
Depositors
Every return figure elsewhere on this site belongs to a vault. These belong to the people in it, and they are different numbers whenever people arrived at different times, which is always.
Depositor books are read for 693 of 2,621 open vaults we can size, holding $7.82B of $11.35B (69% of the capital). A venue that does not publish who is inside its vaults is absent from everything below, which is an absence rather than a finding about its depositors.
Growth
How much of it was people arriving?
Growth
Money that arrived against money that was earned
- Money earned
- $1.94B51%
- People arriving
- $1.84B49%
Of $3.78B of growth across every vault that grew. A venue reports capital and capital growth; neither says which of the two happened, and on a rising market they look the same.
How this figure was arrived at
- Metric
- growth-arrivals
- Population
- vaults_with_flow_historyVaults whose deposits and withdrawals we hold over the window, so growth can be split from performance.
- Denominator
- 3,779,680,699.093 ratio · 100.0% of it
- Source
- Computed here from figures this platform already holds.
- Grade
- B
- Read at
- 2026-10-15 19:41 UTC
- Methodology
- aspern-0.3
- Snapshot
- cmv9xzi8713s8le262ro6c41f
Three routes to the split, ranked. The vault’s own share price is best and needs no cooperation from the venue, because a deposit cannot move it, because a deposit mints shares at that price. The venue’s own profit figure is second and inherits its fee treatment, which is one of the things this is supposed to separate. Flows are the residual and the weakest: a deposit we failed to observe lands in the performance term, which is the flattering direction.
167 of 519 vaults where two routes could both be computed disagree by more than a tenth of the vault. That is published rather than resolved. Two routes agreeing is evidence; two disagreeing is a finding about the venue’s books, and the second is the more interesting of the two.
A vault whose share price cannot move is refused rather than reported at zero — that would be a valuation convention arriving as a result. The flow route is the weakest and errs in the flattering direction: a deposit we failed to observe lands in the performance term. Where more than one route is available they are compared and the disagreement is published rather than resolved.
Growth, decomposed
And what the earned half was made of
The ring above separates money that arrived from money that was earned. This takes the second half apart: closed trades, the fees paid on them, the funding paid or received for holding a position, and the change in what was still open. Every term is read from the venue, and what they do not explain is published rather than folded into the nearest one.
| Money in and outdeposits less withdrawals, and arriving is not earning | −$2,337,518 |
| Closed tradesrealised on the fills the venue reports | $2,177,768 |
| Feespaid to the venue on those fills | −$323,168 |
| Fundingpaid or received for holding the positions, from the venue's own ledger | −$952,008 |
| Open positions movedchange in unrealised profit on what was still open | −$2,477,594 |
| Not explainedwhat the terms above do not account for, 0.7% of the $61,554,702 that moved | $404,233 |
| Change in value |
Addresses seen
30,107
- In one vault only
- 24,125
In two or more vaults
5,982
- Timing measurable
- 557
- Not yet measurable
- 5,425
With measurable timing
557
Behind holding from the start
71
of 557 measured
Beat holding from the start
Read this before using the ranking. The venue lists only current depositors, so anyone who entered, lost money and left is invisible, every address below survived, which flatters all of them. Only the largest 100 holders per vault are published, so smaller and earlier positions are missing. And our value history begins when we started observing, not when the vault did. This is a lead worth investigating, not a verdict.
Chance alone would put the “beat holding” figure near 50%. It currently sits far below that, which at this stage says more about the shortness of our history than about anyone’s judgement: most entries we can measure landed after a vault’s early growth was already in its average. Treat the ranking as usable only once that figure approaches half.
Chance alone would put the “beat holding” figure near 50%. It currently sits far below that, which at this stage says more about the shortness of our history than about anyone’s judgement: most entries we can measure landed after a vault’s early growth was already in its average. Treat the ranking as usable only once that figure approaches half.
Best timing
Ranked by timing rather than by return, because picking a vault that was rising anyway is not skill. 489 addresses are held out for size , because their whole stake is under $1,000. A return on a leftover balance is arithmetic rather than a decision, and ranking it alongside a stake somebody chose would sell you the wrong list. They are counted here rather than dropped, because a shorter list that looks like the whole one is the thing this page exists not to publish. 4,899 addresses are not ranked here: a vault they hold annualises past the 1000% ceiling, and an edge measured from a ceiling is the ceiling minus the vault’s own history, which would rank the worst vaults highest.
| # | Address | Vaults | Equity | Return since entry | Timing edge | Hit rate | Confidence |
|---|---|---|---|---|---|---|---|
| 1 | 2 | $1.1k | 873.2% | 888.0% | 50% | 60% | |
| 2 | 2 | $1.4k | 341.4% | 299.1% | 50% | 60% | |
| 3 | 2 | $44.2k | 295.3% |
Analyst tier
3 of 60 shown in full. 12 more are listed with the wallet withheld. The other 45 are not on this page at all. A plan opens every depositor, concentration, and the timing of each wallet’s entries and exits. The count is free because the count is the finding; the detail is what Analyst carries.