Loading
Operators
Strategies run
20
6 live · 14 closed
Live TVL
$544.4k
Agent-managed
—
with evidence
Best result
129.6%
their headline number
Cohort average
3.0%
equal-weighted, 8 of 20 with data
Management mode
discretionary
inferred
Realised P&L
+$7.8k
across 4 of 20 strategies
Strategies with a trade record
4 of 20
16 with no execution record
Losing strategies
1 of 4
of those we can price
Median lifetime
672d
6 alive · 14 closed
The P&L above covers 4 of 20 strategies. The other 16 carry no execution record, almost always because they closed before we began observing, and a dead vault’s fills cannot be fetched back. Those are disproportionately the failures, so read this figure as a floor on the losses and a ceiling on the gains.
What this firm is actually holding
4 books on hyperliquid.clearinghouseState · read 2026-10-08 06:42 UTC$443.7k
gross, what can be forcibly closed
$443.7k
net, the view actually held
—
offset across margin accounts, which is not a hedge
100% of the gross is directional. Netting is only real where both sides share margin. Where they do not, the exchange liquidates each account on its own: a long in one vault does not rescue a short in another, so the net understates what can be closed against this firm and the gross is the number that matters.
| Instrument | Gross | Net | Cancels | Across |
|---|---|---|---|---|
| ETH | $95.4k | −$95.4k | — | 4 margin accounts |
| NEAR | $56.4k | +$56.4k | — | 4 margin accounts |
| PUMP | $56.3k | +$56.3k | — | 4 margin accounts |
| SAND | $30.5k | +$30.5k | — |
Financial record
6 strategies · 1 venue · 1 chainCapital under management$544.4k
Return, capital-weighted-25.68%over $530.2k, 97% of what could carry a return
No return measurable$14.2ktoo little history $14.2k
Not established yet
Earned outside these strategies—service revenue, jobs
Cost of operating—inference, data, compute
Borrowed—credit and leverage
Blank because nobody has read them, not because they are zero. Each will carry the reading it came from when it arrives, a figure assembled by assigning payments to categories is a guess, and a guess cannot be checked the way 0x3b29…c7e7’s capital can.
Evidence record
20 strategies across the firmHow each record can be checkedNote 9
A · attestedpre-trade committed, sequence complete, anchored on chain0
A− · pre-registereddeclared to us before it ran, so the claim precedes the result0
B · venue-verifiableobservable on chain or at the venue, cohort unproven20$544.4k
C · reportedfrom an aggregator, not read directly0
D · uncheckednothing we can check, usually no contract address0
Why we say software decidesNote 3
None of this firm’s strategies are judged machine-managed. That is a finding, not a gap, most capital here is run by people, and the comparison between the two is the point.
Survivorship
This operator has run 20 strategies that we can see. Their best shows 129.6%; the equal-weighted average across everything they ran is 3.0%. A leaderboard would show you only the first number. 12 of the 20 have no usable history, so the average is computed over 8 and is likely still flattering.
Every strategy, best to worstclosed ones included, ranked last
| # | Strategy | Platform | Network | Status | Autonomy | TVL | Return | Max DD | Sortino | History | Metric conf. |
|---|