How we audit

Official data, as published

Every figure on Linceo comes from each platform's official data. An account's profit curve is exactly the one the platform itself publishes, shown as is and never rebuilt from the trade tape. It comes from the allTime history on Hyperliquid and the user-pnl API on Polymarket. On Polymarket we show the figure on the trader's profile (user-pnl). Polymarket's own leaderboard (lb-api) is a different official figure and may not match, sometimes by a lot.

We do this on purpose: rebuilding the PnL trade by trade produces errors (a merge read as a liquidation, a history cut short), so we only rebuild it when the official figure is missing, and the page says so. We only read public data that the platform itself publishes.

The metrics

We measure with each account's official daily data and the drops it actually went through:

• Sharpe / Sortino on the daily returns of the wallet's official PnL series (resampled to the last value of each day, since Hyperliquid's series is irregular), annualised with √365; with a short record the ratio is pulled toward 0, so a handful of lucky periods can't fake a high one.
• Biggest drop (drawdown), measured on the official profit series, peak to trough: the worst run from a high that a follower would have sat through.
• A check on where the profit comes from (skeleton check): it separates real skill from liquidity rewards, from one-hit luck and from dormant accounts.

Copyability (★1–5)

A high return is useless if you can't actually mirror it. The copyability score rates timing and structure: whether a wallet's entries and exits can be mirrored in time with real signal-to-execution latency — how long it holds, how often it trades, how many positions it keeps open at once, and mechanics nobody can copy. A very profitable trader who moves faster than you can follow scores low on purpose.

Execution, capped and without surprises

When you copy, your money never leaves your account. Your wallet signs a permission to trade; orders are signed by that permission's key, which Linceo keeps encrypted and which can never withdraw. Every copied order passes a per-order cap and safety checks before it is placed. The portfolio backtest warns you how many scaled entries would exceed that cap: those would be skipped, so the real copy can be smaller than the curve shows.

What this can't tell you

Every metric looks backward: it describes what an account did, not what it will do. Data refreshes periodically, so the leaderboard can lag the latest moves; the account page recomputes the metrics from the official curve when you open it. “≥” marks a last-trade date that could only be estimated (at least that many days ago); on Polymarket, “≈” marks a last trade taken from a calculation more than 24 h old. None of this is a recommendation to copy anyone.

Linceo is analytics and signals software: it never holds your funds, your money stays in your account and none of this is financial advice. See our Terms of Service.