💹 GoldApex WF

Complete strategy & performance report — v4.3-CLEAN (recommended profile A12R25)

Symbol: XAUUSD (Gold) Timeframe: H1 Market: Forex CFD Test model: Every Tick (Model=4) Period: 2024-07 to 2026-08

77%Win rate

1.5Profit Factor

~6.5%Monthly return (simple)

10-15%Max drawdown

~640Trades in 2 years

Executive Summary

GoldApex WF is a fully mechanical trend-based system running on gold (XAUUSD) on the H1 timeframe. It uses two complementary sleeves: (1) Momentum — entries with trend strength, and (2) Pullback — entries during trend corrections. Exits use 3-step profit taking, a floating stop loss and strict capital protection rules.

In a 2-year backtest, the recommended profile produced +163% profit with a max drawdown of ~15% (~6.5% monthly, simple).
The result held up in Out-of-Sample validation: ~6.4% monthly in the second half of the data.
Multi-layer risk management (daily/weekly stops, pause after consecutive losses, weekend flattening) is the key to drawdown control.
Important: this is a narrow-edge strategy (PF~1.5); losing months are normal and results are meaningful from ~$2,500 of capital.

1) Strategy Logic

Why gold? Why H1?

Gold (XAUUSD) is one of the most volatile and most "trending" global markets. H1 balances signal frequency and trend quality: fast enough for 20–30 trades per month, slow enough to filter noise.

Two Complementary Sleeves

Momentum sleeve (M): SuperTrend aligned on H1 and H4 + EMA and ADX filters — enters only when the trend is strong and fresh.
Pullback sleeve (P): H4 trend (EMA50/200) + price pullback to EMA with RSI — enters corrections with half the momentum risk.

Where the Edge Comes From

(1) Two-timeframe alignment reduces false signals, (2) 3-step profit taking locks 30%+30% and lets the rest run, (3) ATR-based stops adapt to market volatility. Over 2 years: 77% winning trades and average net +$25.7 per trade.

2) System Architecture

Figure 1 — GoldApex WF architecture (two sleeves + 3-step exit + multi-layer protection)

3) Risk Management

Layer	Function	Default
Daily loss stop	Trading stops for the day after realized loss limit	4%
Max losses per day	After 3 losses, no trades until next day	3
Weekly risk budget	Total weekly risk cap	15%
Consecutive loss pause	After 2 SLs in a row, sleeve rests 8 bars	2 → 8 bars
Drawdown risk cut	If equity is 15% below peak, risk drops 30%	15% / ×0.7
Halve risk after a loss	Next trade risk ÷ number of losses today	On
ATR-based stop	SL distance adapts to volatility (2.5×ATR)	2.5 ATR
Weekend flattening	All positions closed before Friday close	Fri 21:00

4) Backtest Performance — Profile A12R25

+16,309$2-year net profit (10,000$ initial)

+163%Total return

1.49Profit Factor

14.9%Max drawdown

77.0%Win rate

635Trades (~1.2/day)

Metric	Value	Note
Total profit	+16,309$	From 10,000$ to 26,309$
Total / monthly return (simple)	+163% / 6.6%	Average monthly profit ÷ initial capital
Monthly return (compound)	~4.0%	Real compounded growth rate
Profit Factor	1.49	Gross profit ÷ gross loss
Max drawdown	14.9%	Deepest drop from peak
Profit/drawdown ratio	11.0	Main decision metric (edge)
Win rate	77%	Wins ~3.4× more frequent (losses larger but rare)
Avg net profit per trade	+25.7$	Total profit ÷ number of trades
Sharpe / recovery factor	4.0 / 7.8	Return per unit of volatility

Profile A12R25 settings: momentum risk 2.5% • pullback risk 1.25% • SL 2.5×ATR (momentum) and 2.0×ATR (pullback) • ATR period 12 • magic 204741/204742

5) Monthly Returns (24 months, 10,000$ deposit)

Figure 2 — Monthly return (%) — profile LIN4 at 1.5/0.75% risk — 19 of 24 months positive; worst month -4.8%

Profitable monthLosing month

6) Risk Ladder — Return vs Drawdown

Key reading: the profit-to-drawdown ratio (edge) is almost constant (~8 to 12); raising risk only increases "size", not quality. Profile A12R25 is the balance giving ~6.5% monthly with 15% drawdown — where risk is still "reasonable".

7) Out-of-Sample Validation — Real Edge or Overfit?

Validation result: in the second half of the data, all 6 configurations stayed profitable. Profile A12R25 delivered 6.4% monthly out-of-sample (vs 6.6% full period) — rare stability showing the edge is real, not overfit.

8) Warning: $100 Accounts and Minimum Capital

Capital	Target risk (2.5%)	Calculated lot	Actual lot	Result
100$	2.5$	~0.001	0.01 (locked)	Real risk 15-50% — -98% months
1,000$	25$	~0.008	0.01 (locked)	Still below the optimal level
~2,500$	~62$	~0.02	0.02 ✓	Risk management works correctly
10,000$	250$	~0.08	0.08 ✓	Results of this report

⚠️ Important: below ~$2,000–2,500 the percentages in this report are not achievable — the minimum lot multiplies risk 5–20× and can wipe out the account in a bad month.

9) Risks and Limitations (Honest)

Narrow edge: Profit Factor ~1.5; losing months (up to -5%) are a natural part of the system.
Simple ≠ compound: "6.5% monthly" is simple; the real compounded growth is ~4% per month.
Backtest limits: slippage, variable spread and trader psychology can reduce live results by 10–20%.
Market regime change: gold performs worse in a long-term range market.
Financial markets: no strategy is risk-free. This system works with 10–15% drawdown.

10) Installation Guide — Step by Step

Required Files

File	MT5 install path	Description
GoldApex_WF_v4.3_CLEAN.ex5	MQL5\Experts\	Main compiled EA
GoldApex_WF_v4.3_A12R25.set	MQL5\Presets\	"7% monthly" profile settings (recommended)
GoldApex_WF_v4.3_A12.set	MQL5\Presets\	"5% monthly" profile settings (conservative)

Installation Steps

Copy the .ex5 file into MQL5\Experts\ (File → Open Data Folder).
Copy the .set files into MQL5\Presets\.
Restart MetaTrader (or Refresh).
Open an XAUUSD chart on the H1 timeframe.
Drag GoldApex_WF_v4.3_CLEAN onto the chart.
Press Load and select GoldApex_WF_v4.3_A12R25.set.
Mandatory: enable Allow Algo Trading.
Test at least 1 month on a demo account before going live.

Important Notes

Minimum capital ~$2,500 — below this the minimum lot multiplies the risk.
Magics: 204741 (momentum) / 204742 (pullback). Any change = new magic.
Server time: session 8–20 is based on the broker server time.
Spread: brokers with spreads above 40 points may filter signals.