LAB #4611 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 32.2% to 33.1% ( — E8 Intelligence Research

@misc{caldin20264611,
  title = {{LAB #4611 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 32.2% to 33.1% ( — E8 Intelligence Research}},
  author = {Andrew Stewart Caldin},
  howpublished = {Zenodo},
  year = {2026},
  month = oct,
  doi = {10.5281/zenodo.23587280},
  url = {https://zenodo.org/records/23587280},
  note = {CC BY 4.0},
}

TY  - UNPB
TI  - LAB #4611 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 32.2% to 33.1% ( — E8 Intelligence Research
AU  - Andrew Stewart Caldin
PY  - 2026
DA  - 2026/10/23
PB  - Zenodo
DO  - 10.5281/zenodo.23587280
UR  - https://zenodo.org/records/23587280
AB  - IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 32.2% to 33.1% (+0.9pp, n=1498)  SAME-WINDOW EFFECT: Applying the MANIPULATION-phase filter to the live window above would have blocked the three REJECTED/open trades (USDCHF LONG 0.81152, EURJPY SHORT, AUDUSD LONG 0.71755) — all of which are currently unresolved and likely entered during that phase. Among the seven closed trades, none show a clear phase label, so the filter would not have altered any of the four wins or three losses. Net effect on this small sample: no change to realized PnL, but it would have removed three open positions that carry unknown forward risk, potentially reducing drawdown if those trades fail. Expected win-rate delta here is ~0% on closed trades, but variance reduction is plausible.  OVER-TIME PROJECTION: Over 3–6 months, a +0.9pp win-rate improvement (32.2%→33.1%) on ~1,500 backtested trades is marginal and within n  Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
ER  -