High returns are easy to find. Copyable behavior isn't.
We put 100 top-scoring leaderboard candidates through FiatBear's deep audit. 12 passed. The report classified 87 as not suitable for FiatBear's polling-based copying constraints and 1 as inconclusive. That is not a claim that any of them are bad traders.
Returns backed by behavior we can actually follow
0x29998e...65de
audit #5
Passed FiatBear's copyability checks.
0xe41218...f674
audit #18
Passed FiatBear's copyability checks.
0xaefbbf...0114
audit #22
Passed FiatBear's copyability checks.
0x841eb4...ec1b
audit #28
Passed FiatBear's copyability checks.
0x8215c4...07d8
audit #29
Passed FiatBear's copyability checks.
0x47ded9...81e5
audit #37
Passed FiatBear's copyability checks.
0x27e2f8...bf03
audit #42
Passed FiatBear's copyability checks.
0x1e0825...1584
audit #50
Passed FiatBear's copyability checks.
0xf19b3c...2695
audit #61
Passed FiatBear's copyability checks.
0xb14d15...8560
audit #70
Passed FiatBear's copyability checks.
0xcf218b...8a36
audit #91
Passed FiatBear's copyability checks.
0x95d250...c3c1
audit #95
Passed FiatBear's copyability checks.
The leaderboard number wasn't the whole story
0x8fbb98...d714
audit #7
Drawdown above 30%: month drawdown 51%.
0xf73b15...a5b9
audit #33
Mostly spot or DEX assets: mostly spot/dex assets (96% of fills).
0x06ecb2...0b19
audit #2
Less than 90 days of history: account too young (19d).
0x9cd0a6...e5fe
audit #39
No recent activity: inactive (15d since last fill).
Most rejections came from four measurable constraints
Copying adds constraints the original trader never had
A large return can hide a path most followers would not tolerate or survive.
A trader making hundreds of fills per day can move before a polling-based copier sees the change.
Spot and DEX-only activity cannot be reproduced by FiatBear's Hyperliquid perpetuals copier.
A short winning run gives less evidence than behavior that held up over several market regimes.
Reproducible, dated, and deliberately narrow
Evaluated 2026-08-21 15:16 UTC using rules hl-scout-v1.
The scout first filtered public Hyperliquid leaderboard data for account range, profitability, turnover, and edge. It then checked public portfolio and fill history for age, drawdown, recency, pace, holding behavior, and asset coverage.