Open any MT4/MT5 backtest report and you'll find dozens of metrics: net profit, gross profit, gross loss, profit factor, expected payoff, absolute drawdown, maximal drawdown, relative drawdown, total trades, short trades, long trades... the list goes on.
The problem isn't too few metrics — it's too many, without any guidance on which ones to trust. Some metrics predict future performance. Others simply describe what already happened. And a few are actively misleading when used in isolation.
A Tiered Metrics Reference
| Metric | Tier | Reliability |
|---|---|---|
Profit Factor Gross profit ÷ gross loss — the clearest edge indicator |
Core | ★★★★★ |
Expectancy per Trade Average expected gain per trade — confirms the math works |
Core | ★★★★★ |
Max Drawdown Worst peak-to-trough decline — survival check |
Core | ★★★★☆ |
Recovery Factor Net profit ÷ max drawdown — efficiency of risk taken |
Core | ★★★★☆ |
Sharpe Ratio Risk-adjusted return — useful for comparing strategies |
Useful | ★★★☆☆ |
Max Consecutive Losses Longest losing streak — tests psychological tolerance |
Useful | ★★★☆☆ |
Trade Count Statistical sufficiency check — fewer than 100 is unreliable |
Context | ★★☆☆☆ |
Win Rate Wins ÷ total trades — only meaningful with R/R context |
Use with care | ★☆☆☆☆ |
Net Profit (alone) Total return without drawdown context — incomplete picture |
Use with care | ★☆☆☆☆ |
Why Profit Factor Beats Net Profit
Net profit is the outcome. Profit factor is the mechanism. Two strategies can produce identical net profits through completely different quality of trading, and profit factor exposes that difference.
Consider: Strategy A makes 100 trades, $200 average win, $150 average loss, 60% win rate. Strategy B makes 12 trades, all concentrated in one favorable period. Both show $5,000 net profit. Strategy A has a profit factor of 2.0, consistent across the period. Strategy B's profit factor over its trade history is inflated by the cluster — and nearly meaningless as a forward predictor.
Metrics that measure the process of how money was made — profit factor, expectancy, consistency — predict future performance better than metrics that measure the result — net profit, total return.
The Problem with Win Rate (Again)
Win rate appears prominently in most backtest reports and is the first thing many traders look at. As we've covered in a previous article, it's essentially meaningless without the context of average win size vs average loss size.
But there's a deeper problem: high win rate strategies actively encourage bad behavior. They generate constant positive feedback, making it psychologically easier to keep running a strategy that might be fundamentally unprofitable. The losses are infrequent but large — and by the time you notice the pattern, significant damage may already be done.
Metrics That Work Together
- Profit Factor + Max Drawdown (edge vs risk)
- Expectancy + Trade Count (reliable edge signal)
- Recovery Factor + Drawdown Duration (resilience)
- Rolling PF + Consistency (temporal stability)
- Win Rate alone (no R/R context)
- Net Profit without drawdown (result without cost)
- High PF on <50 trades (statistically irrelevant)
- In-sample metrics without OOS validation
One Metric That's Underused: Max Consecutive Losses
Most traders check max drawdown but ignore max consecutive losses. This matters for a different reason: it determines your psychological endurance, not just your financial survival.
A strategy with a 10% max drawdown that achieves it through 12 consecutive losing trades is a very different experience than one that reaches 10% through alternating wins and losses. If you know your strategy can produce 12 consecutive losses, you can psychologically prepare for it and set a rule in advance about when to intervene. If you don't know, you'll be making that decision under maximum stress.
Before going live, ask: what is the maximum consecutive loss run in my backtest? Now double it. Is that a losing streak you could sit through without breaking your rules? If not, reduce position size until the answer is yes.
The Minimum Viable Metric Set
If you only ever check five numbers on a backtest report, make them these: Profit Factor (is there edge?), Expectancy (is the math positive?), Max Drawdown (can I survive it × 2.5?), Recovery Factor (is the risk worth the reward?), Trade Count (is there enough data to trust any of the above?).
Everything else in the report is context. These five are the foundation.
See All Your Key Metrics Instantly
EA Analyzer Pro extracts and highlights the metrics that matter from your MT4/MT5 backtest report — free, no installation required.
Open EA Analyzer Pro →