Quantitative Analysis

Backtest Results

Historical simulation on real MT5 OHLC data. Includes commission ($7/lot) and slippage. London + NY killzone filter applied.

Equity Progression

50k M15 bars · 1% risk · XAUUSDm

$15,433

from $5,000

$5k$8k$10k$13k$15kStartMo 3Mo 6Mo 9Mo 12Mo 15Mo 18Mo 21Mo 24End

Key Parameter Findings

H4 Trend Confirm

Requiring H4 EMA-20 direction to agree with H1 EMA eliminates counter-trend noise universally across all tested pairs.

Impact

PF ↑ all pairs

Setting

always true

Signal Cascade

Three-path OTE → FVG → OB cascade increased trade frequency from ~5/month to ~6/month while maintaining high PF.

Signals/mo

5 → 6+

PF retained

3.90

Trailing 2R = false

Moving SL to 1R when 2R is hit consistently reduced PF across all 32 sweep combinations. Partial TP is superior.

vs trailing on

PF higher

Setting

always false

All Runs

SymbolPeriodTFRiskTradesWin RateProfit FactorReturnMax DD
XAUUSDm50k bars (~25 mo)M151%14854.1%3.90+209.0%-9.5%Optimal config
GBPUSDm50k barsM151%7256.9%3.42+143.0%Parameter sweep
NZDUSDm50k barsM151%5453.7%2.91+87.0%Parameter sweep
EURUSDm50k barsM151%4852.1%2.67+62.0%Parameter sweep
BTCUSDm50k barsM151%3158.1%2.44+48.0%Parameter sweep

Parameters

Timeframe
M15
Symbol
XAUUSDm
Killzone
08:00–13:00, 15:00–18:00 UTC
EMA H1 Period
20
EMA H4 Confirm
true (period 20)
DOW Filter
Skip Mon + Fri
Min RR
2.5×
Swing Period
2
Partial TP 1R
50% at 1R
Partial TP 2R
25% at 2R
Trailing 2R
false (hurts PF)
Friday Close Hour
21:00 UTC

How It Works

  1. 1

    Detect Structure

    BOS or CHoCH confirms directional bias on the rolling candle window.

  2. 2

    Trend Alignment

    H1 EMA-20 and H4 EMA-20 must both agree with the BOS direction.

  3. 3

    Killzone + DOW

    Entry only during 08–13 UTC and 15–18 UTC. Skip Monday and Friday.

  4. 4

    Signal Cascade

    Try OTE (Fib 61.8–78.6%) first. Fallback to FVG, then OB if no OTE.

  5. 5

    SL / TP / Partials

    SL at PD array boundary. TP at swing extreme. Close 50% at 1R, 25% at 2R.

Disclaimer — Backtests simulate on historical OHLC data and do not account for requotes, broker restrictions, or changing market regimes. Past results do not guarantee future performance. For informational purposes only.