Backtest evaluation shows that filtering short gold positions increases algorithmic win rate from 32.0% to 55.2%, highlighting potential profitability alongside sample overfitting risks.
IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: XAUUSD SHORT — improves win rate from 32.0% to 55.2% (+23.1pp, n=29) SAME-WINDOW EFFECT: Applying the XAUUSD SHORT filter would have blocked the three XAUUSD SHORT trades in the window (entries at 4048.74, 4051.79, 4066.35), all of which closed as LOSSES. Removing those three losses from the 10-trade window would shift the record from 5W/5L to 5W/2L (71.4% win rate), a +21.4pp improvement. PnL delta is positive but modest — the three blocked trades were likely small losses, so the net gain is roughly +1.5R to +2.5R, not a game-changer. OVER-TIME PROJECTION: Over 3–6 months, the filter's value depends on whether the 55.2% win rate (n=29) is stable or a lucky artifact of a specific gold regime (e.g., strong USD or risk-off). The risk is overfitting to a short sample — XAUUSD SHORT may have been favorable only during a narrow volatility cluster, and the filter could block profitable long-side gold trad Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
No takes yet. Share an insight, caveat, or question.
Andrew Stewart Caldin (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: