XMBot Backtest Results: What 1,083 Trades Taught Us
Deep dive into our 6-month backtest on PAXG/USDT (tokenized gold) M5 data — win rate, profit factor, drawdown, and what walk-forward analysis reveals.
## Methodology
We backtested the RSI Sniper + ADX Filter strategy on PAXG/USDT (tokenized gold, traded on Binance) M5 data from July–December 2025 — not live XAUUSD forex/CFD data. Spreads, slippage, and execution on a real XAUUSD broker account will differ from this backtest. The test used: - $10,000 initial capital - 2% max risk per trade - 0.05 lot default size - RSI oversold/overbought filters - ADX trend strength confirmation
Key Metrics
| Metric | Value |
|---|---|
| Total Trades | 1,083 |
| Win Rate | 64% |
| Total Return | +84.3% |
| Max Drawdown | 4.3% |
| Profit Factor | 2.1x |
| Sharpe Ratio | 1.8 |
Walk-Forward Analysis
To validate the strategy isn't overfitted, we ran walk-forward analysis on unseen data. The result: **+19% return** on the out-of-sample period, confirming the strategy generalizes.
Risk Management Impact
Without the 2% risk rule, the same strategy would have seen 12%+ drawdown. The risk manager is what makes this sustainable.
Disclaimer
This backtest ran on PAXG/USDT, a tokenized gold proxy — not live XAUUSD execution. Past performance does not guarantee future results, and backtested results in particular tend to overstate real-world performance. Trading involves significant risk of loss.