The headline is an arithmetic sum of per-signal OHLCV price outcomes before sizing or costs; it is not a portfolio return. The equity card is a separate sequential simulation that risks 1% of modeled current equity on every eligible sized signal and deducts per-asset fee/slippage assumptions. It does not model broker fills, overlapping exposure, margin, leverage, latency, funding, or subscriber selection. Pending, zero-value expiry placeholders, and gate-blocked rows are surfaced separately instead of being folded into win-rate statistics.
Risk notice: TradeClaw provides educational signal analytics and historical research. Signals are not financial advice, investment recommendations, or guaranteed outcomes. Trading involves substantial risk, including loss of capital. Past performance does not predict future results. Always test with paper trading and independent risk controls before using real funds.
Eligible rows for every tracked symbol in the current archive.
Research Alerts
Entry-like alerts remain paused whenever the published cost-adjusted evidence gate is not ready.
Inspect Strategy Research
Review the available modeled backtests and trade logs. Fixture metrics are research diagnostics, not live or broker performance.
Signal Study — Sequential SimulationEligible stream
Eligible signal stream. Outcomes require approved observed-OHLCV provenance; no broker fills are used.
Per-Symbol Outcomes
current archive| Pair | Recorded rows | 4h Hit | 24h Hit | Avg P&L | Price-move sum | Trend |
|---|---|---|---|---|---|---|
Recorded Signal Rows
| Bar Open | Pair | Dir | Entry | 4h | 24h | P&L |
|---|---|---|---|---|---|---|
Population and Limitations
Stored signals carry their generation timestamp. New counted outcomes require an approved observed-OHLCV source and are not broker fills. Legacy outcome values without approved source provenance remain visible as unverified audit rows and do not count. Counted outcomes include source-backed TP, SL, and nonzero 24-hour closes; they exclude simulated, gate-blocked, pending, and zero/missing force-expiry placeholders. The unsized price-move sum uses counted rows, while the sequential simulation further requires a recorded stop-loss. The broadcast view includes only gate approvals recorded since 2026-06-10. These populations are signal studies, not customer trades or actual portfolio returns.