Backtesting an MT5 Expert Advisor means running the EA against historical XAUUSD tick data inside MetaTrader 5's Strategy Tester to measure how it would have performed before risking real capital. Below is the exact data sourcing, modelling settings, and report-reading workflow to run a credible backtest of NextTrade Gold EA on your own broker. We deliberately do not publish a "headline number" — your live results depend on your broker, your settings, and the market regime. Always run your own test.
Want the input list? See all features documented. Want the decision flow behind each trade? Read how it works.
Almost every EA vendor advertises a fixed win-rate or monthly-return figure. Those numbers are easy to manufacture from a single cherry-picked window and they mislead more than they inform.
Garbage in, garbage out. Tick data quality is the single biggest reason backtests don't match live results.
Use the same broker you intend to go live with. XAUUSD spread, swap, and commission vary widely between brokers — testing with a different broker's history is testing a different instrument.
If your broker's history is thin, import third-party tick data. Be aware spreads will be modelled — set them to a realistic XAUUSD value (typically 20-35 points on ECN, 30-60 on standard).
The settings below are non-negotiable for a credible NextTrade Gold EA backtest.
XAUUSD on M15 (the EA's default trigger timeframe, sw_custom_tf). If you change sw_custom_tf, match the tester timeframe accordingly.
This is the only mode that approximates live execution for an intrabar EA. "1-minute OHLC" is fine for a quick sanity check; never publish a result from it.
Minimum 2 years. Include at least one full quarter of post-NFP volatility and a high-volatility stress window if available.
Match what you'll trade live. Tune fixlot to your equity — at 0.01 lot the per-trade exposure is small enough for sub-$1k testing.
Use your broker's actual cap (1:100, 1:200, 1:500). Backtesting at 1:1000 when you'll trade at 1:100 will mislead you about margin pressure.
Load the included .set file from the EA inputs panel. Don't tune inputs blindly — start from the preset and change one variable at a time.
A high net profit means nothing on its own. These are the numbers that actually matter.
The worst peak-to-trough loss the strategy ever endured. If it exceeds the pain you can stomach in real money, the equity curve is a lie — you'd have closed it down before it recovered.
Gross profit ÷ gross loss. Above 1.3 is real, above 1.6 is excellent for an XAUUSD H1 system. Suspiciously high (3+) usually means curve-fit.
Recovery factor = net profit ÷ max drawdown. Sharpe accounts for volatility. Both have to look reasonable across the full window — not just the last six months.
Look for hundreds of trades, not dozens. Look at monthly P&L — three good months and a flat year is curve-fit. Steady-with-drawdowns is real.
Want to run sensitivity tests? These are the EA inputs that have the biggest effect on backtest results. Change one at a time, not all at once.
fixlot — fixed lot per trade (default 0.01)sl_pips — fixed SL in pips (0 = no SL)tp_pips — fixed TP in pips (default 300)use_trail + trail_start / trail_distsw_custom_tf — trigger timeframe (M15 default)sw_strength — bars-each-side, higher = rarer signalssw_signal_mode — Classic vs Reverseuse_pattern_filter — require candle confirmationdaily_profit / daily_loss daily capsmax_dd drawdown halt %max_daily_orders trade-count capAlways split the data: optimise on the first 70%, validate on the last 30%. Anything else is curve-fitting.
For an EA that places SL/TP and uses M15 patterns, this mode skips the bar interior entirely. Results are meaningless.
XAUUSD is a wide-spread instrument. A backtest with 5-point spread will look great and trade like garbage live.
Get the EA, the presets, and the setup guide. Backtest it on your broker's data before you risk a dollar.
See Pricing