This address still holds
$229K
perps + spot + vaults + staking; perps account $0, no open positions
0x647d3a57ff93a5f46057e232e983b45f9e4ea435View on Hyperliquid ↗ ·
Not offered yetHigh frequency Mark-to-market cumulative profit $271K, ranked #1767 site-wide.
Those three numbers describe where he stands now. The curve below is a different thing — how much he made in the past. An address showing $271K of profit can be sitting on an empty account, and the P&L figure alone will never tell you.
Three definitions on one axis: green = our mark-to-market ledger · red = the exchange's public "closed P&L" · grey dashed = exchange all-in. The shaded gap is what closed-P&L misses ($5,924). The hatched band on the left is before our data starts (20251117); the exchange had him at $-76K by then — that is why our number differs from other sites.
The green line above is how much he made trading (rebuilt from on-chain fills; deposits and withdrawals are invisible there). This amber line is how much money is in his account (from the exchange API, deposits included). Wherever the two diverge, he put money in or took money out. So "made a lot but little left" and "made little but holds a lot" are both possible and neither is strange.
⚠️ Equity is sampled weekly for anything older than a month (near-daily only recently); this line connects those points rather than measuring every day. ⚠️ It covers perps + spot + vaults + staking, not just the perps account.
Funding is 19.8% of gross. Fees are negative, so they are listed under the ring as a deduction rather than forced into the pie.
Within each bar: green = directional, amber = funding. Right of the line is profit, left is loss.
Axes are percentiles among all addresses — further out means more pronounced, not better.
All 6 axes are facts, not a score. The right chart is cumulative profit across all 7,895 profitable addresses (log scale); the line is him — everyone to the right made more.
His positions rarely go flat. In 133 days he fully closed out only 6 times; the rest of the time he was adjusting an open position — 4,113 fills in that span.
This is not missing data — there are no rounds to count. A win rate needs discrete trades you can score one by one, and his position never closes. So we show how much profit each dollar of volume produced instead. (Leverage inflates this number.)
Why copy trading is not open yet · High frequency
294 fills per day. Mirroring this account needs order placement at his speed and fees low enough not to eat the thin per-trade margin. That is an execution-layer problem, and we have not wired it into copy trading yet.
His positions rarely go fully flat, so there is no "every round" table above — a round-trip win rate would not mean anything for him. But 1 of his positions did open and close cleanly, and those can be unfolded fill by fill:
| Cumulative P&L (mark-to-market) | $271K |
| Max drawdown (mark-to-market) | $144K |
| Return (÷ true intraday peak exposure) | 25.5% |
| Profit vs his peak account equity | 49.3% |
| His peak account equity | $550K |
| Resilience (P&L ÷ drawdown) | 1.88 |
| Largest exposure held | $1.06M |
| Open-to-flat rounds | 6 |
| Spread across | 4 coins |
| Fills inside those rounds | 4,113 |
| Typical round length | 30.5 days |
| Same-day in-and-out share | 33.3% |
| Taker share | 80.1% |
| Fills per day | 293.8 |
| Days active | 133 |
0x5b5d51203a0f9079f8aeb098a6523a13f298c060 · 0x2ea18c23f72a4b6172c55b411823cdc5335923f4 · 0x35d1151ef1aab579cbb3109e69fa82f94ff5acb1 · 0xa5b0edf6b55128e0ddae8e51ac538c3188401d41 · 0x418aa6bf98a2b2bc93779f810330d88cde488888 · 0x8e096995c3e4a3f0bc5b3ea1cba94de2aa4d70c9
Where these numbers come from. Rebuilt from Hyperliquid's raw fill stream: each coin's position is cut into rounds from first entry to fully flat, then marked to the market price of every single day, plus funding, minus fees. Summing all 1.24M rounds reconciles to the daily mark-to-market P&L to 2.2e-16.
What this page is not. It describes how he made money in the past. It is not a claim that following him will make you money. We have never tested whether addresses near the top of this board do better next month — no evidence, no claim.