LAB #1834 HARMFUL: LEDGER BENCH: Apply filter: USDCHF LONG @ LONDON_OPEN with 5 conf — improves win rate from 31.9% to — E8 Intelligence Research
Year: 2026
Type: preprint
Abstract: IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: USDCHF LONG @ LONDON_OPEN with 5 conf — improves win rate from 31.9% to 80.0% (+48.1pp, n=5) SAME-WINDOW EFFECT: Over the current live window, this filter would have blocked all non-USDCHF trades — that's 10 of 10 trades removed. Since none of the live trades were USDCHF LONG at London open with 5 conf, the filter would have produced zero trades, eliminating both the 3 wins and 4 losses (plus 3 open/rejected). Net effect: PnL delta = 0 (no trades taken), but win rate becomes undefined (0 trades), not 80%. The claimed +48.1pp improvement is based on n=5 — statistically meaningless and not observable in this window. OVER-TIME PROJECTION: Over 3-6 months, this filter would drastically reduce trade frequency — likely to near-zero, since USDCHF LONG at London open with 5 conf is a rare confluence. With n=5, the 80% win rate has a 95% confidence interval roughly 28%-99% — the true rate could be worse than Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
Author: Caldin, Andrew Stewart
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7208837368
Cites:
Cited by:
Related to: