EXP-000001
Equity-basis vs balance-basis drawdown across the catalogue
Hipotesis
Balance-basis drawdown (closed trades only) materially understates peak-to-trough risk for systems that hold losing positions, and must not be ranked against equity-basis figures silently.
Perubahan terkontrol
Importer reads equity drawdown wherever the report states it and records the basis; every page carrying a drawdown states which basis it shows; balance-basis pages carry an explicit caveat.
Dataset & kondisi
All 14 tester reports (5 MT5 with both bases, 9 MT4 balance-only).
Sebelum / Sesudah
Sebelum
- SUPREMACY drawdown as first imported (balance basis)
- 12.44%
Sesudah
- SUPREMACY drawdown on equity basis, same report
- 48.82%
Validasi
Hasil OOS, walk-forward, Monte Carlo dan forward hanya ditampilkan bila datanya ada. Entri ini tidak memilikinya.
Keputusan
The gap on the worst case was a factor of ~4 (12.44% vs 48.82%). Publishing the balance figure as "the" drawdown would have materially understated risk.
Keterlibatan AI
- Model
- Claude (Anthropic)
- Yang dikerjakan AI
- Read all 14 tester reports, surfaced the equity/balance mismatch and drafted the importer change. The publication rule (always label the basis) was decided by a human.
- Tinjauan manusia
- ✓ Ditinjau dan disetujui oleh manusia
Entri berbantuan AI hanya dipublikasikan setelah tinjauan manusia; entri buatan AI tanpa tinjauan itu ditahan oleh pipeline build. AI tidak memprediksi pasar, dan tidak ada entri yang mengklaim demikian.
Catatan
Basis labelling now appears on detail pages, the compare matrix and the risk/return map.
Perbaikan backtest tidak berarti perbaikan masa depan. Makin banyak eksperimen, makin besar peluang sebagian berhasil kebetulan — log ini ada sebagian agar risiko multiple-testing tetap terlihat.