LAB #4440 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.5% to 32.3% ( — E8 Intelligence Research

@misc{caldin20264440,
  title = {{LAB #4440 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.5% to 32.3% ( — E8 Intelligence Research}},
  author = {Andrew Stewart Caldin},
  howpublished = {Zenodo},
  year = {2026},
  month = oct,
  doi = {10.5281/zenodo.23490938},
  url = {https://zenodo.org/records/23490938},
  note = {CC BY 4.0},
}

TY  - UNPB
TI  - LAB #4440 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.5% to 32.3% ( — E8 Intelligence Research
AU  - Andrew Stewart Caldin
PY  - 2026
DA  - 2026/10/20
PB  - Zenodo
DO  - 10.5281/zenodo.23490938
UR  - https://zenodo.org/records/23490938
AB  - IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.5% to 32.3% (+0.8pp, n=1535)  SAME-WINDOW EFFECT: Applying the MANIPULATION-phase filter to the live window would have excluded the three wkjudas trades (GBPUSD LONG, USDJPY SHORT, NZDUSD SHORT) — all closed as LOSSES — and likely the two rizzy7m AUDUSD trades (one LOSS, one open/REJECTED). That removes 3 confirmed losers and 1 probable loser from 10 trades, leaving 5 wins and 1 loss among the remaining rizzy5m/7m signals. Expected win rate jumps from ~50% (5/10) to ~83% (5/6) in this window, with a clear positive PnL delta of roughly +3R (assuming ~1R per win and −1R per excluded loss). However, the filter also excludes the two REJECTED/open trades, so the live effect is partially confounded by execution status.  OVER-TIME PROJECTION: Over 3–6 months, the +0.8pp win-rate gain (31.5%→32.3%) is statistically marginal (n=1535, p≈0.15) and likely  Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
ER  -