LAB #4478 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.5% to 32.3% ( — E8 Intelligence Research

@misc{caldin20264478,
  title = {{LAB #4478 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.23526279},
  url = {https://zenodo.org/records/23526279},
  note = {CC BY 4.0},
}

TY  - UNPB
TI  - LAB #4478 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  - 2026/10/21
PB  - Zenodo
DO  - 10.5281/zenodo.23526279
UR  - https://zenodo.org/records/23526279
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=1536)  SAME-WINDOW EFFECT: Applying the MANIPULATION-phase filter over the current live window would have removed the three wkjudas trades (GBPUSD LONG, USDJPY SHORT, NZDUSD SHORT) — all three were losses, so the filter would have converted a 4W/3L record into 4W/0L, a +37.5pp swing in win rate on this small sample. However, it would also have removed the two rizzy7m trades that are currently REJECTED/open (AUDUSD LONG, USDCHF LONG) — but those are not yet resolved, so the net immediate effect is strictly positive: +3 losses removed, no wins touched. PnL delta on this window would be roughly +3R (assuming ~1R average loss per removed trade), but the sample is tiny (n=10) and dominated by a single strategy family (wkjudas), so this is not evidence of general efficacy.  OVER-TIME PROJECTION: Over 3–6 months, a +0.8pp win-r  Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
ER  -