#53 — H4 Donchian + FIFO stopped (live check)

Hypothesis — what it tests

Whether the H4 Donchian breakout and the FIFO-Trends P&F strategies — both of which passed backtest validation — survive real live execution at broker spreads and fills.

Description

This entry is the live-execution reckoning for two backtest winners. The H4 Donchian breakout (Donchian(10) on H4 bars + Wilder ATR(14) trailing stop, backtest OOS 16–25 p/d/pair at 84% WR) ran on account 010 and produced 3 trades, −61.6 pips, 0% win rate in 8 days — every exit a trail-stop, clearly worse than the +80 p/d OOS estimate — and was stopped with positions closed at market. FIFO-Trends (P&F trend-following with a 1-box trail on GBP_JPY, account 013) totalled 15 trades, −168.5 pips, 7% WR across all live attempts; the root cause was that a 1-box (5-pip) trail on a wide-spread pair is consumed by spread + slippage on every loser, and the S5 exit-monitoring fix improved fill accuracy but could not fix the structural problem. Both were stopped on 2026-05-26, with a rule not to redeploy FIFO until a 2-box-trail variant validated in paper (it later failed too).

Key result

Indicators

Algorithms

Code