analysis: run without DD/filter limits to measure trade ceiling and diagnose losses

[cfournel]

Remove/relax post-MSS filters (SWEEP, TOP_IMP, DD stop if any) to see how many trades the EA could make on 2022-2026 data. Goal: understand which filters are responsible for trade losses and whether they are each justified.

[cfournel]

Analysis complete — DD limit NOT the bottleneck.

Ran 2022-2026 backtest with inpMaxTesterDD=100.0 (removed 40% DD limit). Result is identical:

• 10 trades, PF 2.75, balance $21,041 — exactly the same as with DD=40%

The DD cap never triggered during this backtest. Max observed DD was ~10.53%, well within 40%.

Full OB funnel (2022-2026):

6520 total OBs detected
3695 fc_ob (passed basic quality filter)
  13 fc_mss (MSS + SWEEP passed in checkForMSSBefore)  ← 0.35% pass rate
   1 fc_imbalance_filled
   8 fc_timp (topImpValid flag — not a clean funnel step)
  15 traded (all isAllGood() checks passed, orders placed)
  10 closed trades (reported as total_trades)

Key insight: The overwhelming bottleneck is MSS swing structure detection — only ~16 of 3695 valid OBs ever pass the swing H-L-H or L-H-L pattern + SWEEP check inside checkForMSSBefore(). Removing the DD limit changes nothing.