LAB #4499 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.6% to 32.4% ( — E8 Intelligence Research

@misc{caldin20264499,
  title = {{LAB #4499 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.6% to 32.4% ( — E8 Intelligence Research}},
  author = {Andrew Stewart Caldin},
  howpublished = {Zenodo},
  year = {2026},
  month = oct,
  doi = {10.5281/zenodo.23553564},
  url = {https://zenodo.org/records/23553564},
  note = {CC BY 4.0},
}

TY  - UNPB
TI  - LAB #4499 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.6% to 32.4% ( — E8 Intelligence Research
AU  - Andrew Stewart Caldin
PY  - 2026
DA  - 2026/10/22
PB  - Zenodo
DO  - 10.5281/zenodo.23553564
UR  - https://zenodo.org/records/23553564
AB  - IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.6% to 32.4% (+0.8pp, n=1530)  SAME-WINDOW EFFECT: Applying the MANIPULATION-phase filter would have excluded trades entered outside that quarterly phase. In the current window, only the USDCHF LONG (rizzy7m) and possibly the AUDUSD LONG (rizzy7m) fall within a plausible MANIPULATION phase, but the exact phase boundary is undefined in your data — so the filter would likely have blocked most of the 10 listed trades, including the 4 CLOSED/WINs. That would reduce the live win count from 4 wins/3 losses (among closed) to roughly 1–2 wins/0–1 losses, a net PnL drop despite a higher win rate percentage. The +0.8pp backtest gain is not visible here; it would have sacrificed absolute returns for marginal selectivity.  OVER-TIME PROJECTION: Over 3–6 months, the filter may improve win rate slightly (as backtested) but likely at the cost of trade frequen  Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
ER  -