How to Read MT5 Backtest Results Like a Pro: Profit Factor, Drawdown & More

You download an EA from MQL5 Market. The vendor shows a backtest with $50,000 profit on a $1,000 account. Incredible, right?

Not so fast. That screenshot might be worthless.

The MT5 Strategy Tester gives you dozens of metrics, and most traders don't know how to read them. Worse, many EA vendors exploit this ignorance — showing only the numbers that make their product look good while hiding the ones that matter.

This guide teaches you to read every important metric, understand what it really means, and spot the tricks vendors use to make bad strategies look amazing.

Setting Up a Proper Backtest

Before you read a single number, make sure the test itself is valid. A perfect-looking result on a flawed test setup is worthless.

Tick Model: Use "Every tick based on real ticks" — this is the highest quality modeling MT5 offers. "Every tick" is acceptable if real tick data isn't available. "Open prices only" is garbage for most strategies. It skips intra-bar price action entirely, meaning your stop losses and take profits won't trigger at realistic points.

Testing Period: Minimum 2 years. Ideally 5 or more. A strategy tested on 6 months of trending market proves nothing about how it handles ranging, volatile, or crash conditions. You need to see your strategy survive 2020 COVID, 2022 rate hikes, and everything in between.

Spread: Use "Current" or set a realistic fixed spread for the instrument. Some vendors test with 0 spread — their results are pure fiction. On XAUUSD, the difference between 0 spread and a realistic 20-point spread can turn a profitable strategy into a losing one.

Initial Deposit: Match your real trading capital. A strategy that works on $10,000 with 0.10 lot might not work on $500 because the minimum lot size (0.01) creates a completely different risk profile.

The Metrics That Actually Matter

Profit Factor

The simplest metric to understand: Gross Profit divided by Gross Loss. If the strategy made $10,000 and lost $5,000, the Profit Factor is 2.0.

Profit Factor	Rating	What It Means
1.0 – 1.3	Marginal	Barely profitable. Slippage and commissions might eat the edge.
1.3 – 1.5	Decent	Usable in live trading if other metrics are solid.
1.5 – 2.0	Good	Solid strategy with a real edge.
2.0 – 3.0	Very Good	Strong performance. Worth investigating further.
3.0+	Excellent	Great if validated. Verify it's not overfitted.
5.0+ (under 200 trades)	Suspicious	Almost certainly overfitted. The strategy was tuned to match historical data perfectly.

A Profit Factor of 1.8 across 500 trades over 5 years is far more trustworthy than a Profit Factor of 4.0 across 30 trades over 3 months.

Maximum Drawdown (Absolute & Relative)

The largest peak-to-trough decline in your account. This is arguably the single most important number because it tells you how much pain you'll experience before the strategy recovers.

Max Drawdown	Risk Level	Reality Check
Under 10%	Conservative	Suitable for most traders.
10% – 20%	Moderate	Experienced traders only.
20% – 30%	Aggressive	Not for the faint-hearted. Can you stomach watching $3,000 disappear from a $10,000 account?
50%+	Extreme	You need a 100% gain just to get back to breakeven. Avoid unless you fully understand the risk.

Always focus on relative drawdown (percentage) rather than absolute dollar amount. A $500 drawdown sounds small, but on a $1,000 account that's 50% — devastating.

Remember: backtested drawdown is the minimum you'll experience. Live drawdowns are almost always worse due to slippage, wider spreads during news, and execution delays.

Total Net Profit

How much money did the strategy make? Simple, but context is everything.

$5,000 profit on a $10,000 account (50% return) over 5 years is roughly 8.4% annually — you'd be better off in an index fund. $5,000 on a $1,000 account (500% return) looks incredible, but check the drawdown first. Was there a point where the account was down 80%? If so, most traders would have pulled the plug long before the recovery.

Always evaluate profit relative to the initial deposit and the maximum drawdown endured to achieve it.

Number of Trades

This is your statistical significance check. No amount of fancy metrics matters if the sample size is too small.

Number of Trades	Statistical Value
Under 50	Meaningless. Could be pure luck.
50 – 100	Borderline. Not enough to draw conclusions.
100 – 200	Acceptable. Starting to show a pattern.
200 – 500	Good. Reliable sample size.
500+	Strong. High statistical confidence.

Many overfitted strategies have very few trades. The developer kept adding filters and conditions until only "perfect" setups remained — setups that will never appear again in the same way in live markets.

Win Rate vs. Risk-Reward Ratio

Win rate alone is one of the most misleading metrics in trading. A 90% win rate means nothing if the 10% of losers are 20x larger than the average winner. Some strategies win often with small gains but have occasional large losses — the key is whether the math works out in your favor over hundreds of trades.

What actually matters is the combination:

Win Rate x Average Win vs. Loss Rate x Average Loss

Example: A strategy with 40% win rate and a 3:1 reward-to-risk ratio earns, on average: (0.40 x $300) - (0.60 x $100) = $120 - $60 = $60 per trade. Compare that to an 80% win rate with 0.3:1 ratio: (0.80 x $30) - (0.20 x $100) = $24 - $20 = $4 per trade. The "ugly" 40% win rate strategy makes 15x more per trade.

Expected Payoff

The average profit per trade, calculated across all trades. This number should be positive and — critically — large enough to cover your real trading costs.

If Expected Payoff is $2.50 per trade but your broker charges $7 round-trip in spread plus commission, the strategy is actually losing $4.50 per trade in real conditions. A backtest won't show this if the spread was set unrealistically low.

For XAUUSD on a typical ECN broker, expect $5-15 in round-trip costs per 0.10 lot. Your Expected Payoff needs to comfortably exceed this number.

Sharpe Ratio & Recovery Factor

Sharpe Ratio measures risk-adjusted return. It answers the question: "How much return am I getting per unit of risk?" A Sharpe above 1.0 is acceptable. Above 2.0 is very good. Below 0.5 means the strategy takes on too much volatility for the returns it delivers.

Recovery Factor is Net Profit divided by Maximum Drawdown. It tells you how efficiently the strategy recovers from losses. If a strategy made $10,000 with a max drawdown of $2,000, the Recovery Factor is 5.0 — healthy. Below 3.0 means recovery is slow. Below 1.0 means the strategy hasn't even recovered from its worst drawdown yet — a major red flag.

Reading the Equity Curve

Numbers lie. The graph doesn't — at least, not as easily.

A good equity curve shows a smooth upward slope with small, brief drawdowns and consistent growth across different time periods. You should see the strategy making money steadily in 2020, 2021, 2022, 2023 — not just one banner year carrying the entire result.

A bad equity curve tells specific stories:

• Staircase pattern — long flat periods with sudden jumps. The strategy only works in specific market conditions and sits idle (or slowly bleeds) the rest of the time.
• Hockey stick — flat for most of the period, then a sudden spike at the end. Classic sign of curve-fitting: the developer optimized to exploit a specific recent move.
• Deep V-shapes — massive drawdowns followed by sharp recoveries. Check if the strategy has proper risk management to handle prolonged adverse moves, not just favorable conditions.
• Diverging balance and equity lines — if the balance line (closed trades) looks great but the equity line (including open trades) shows deep dips, the strategy holds large floating losses. It's hiding the real risk.

Common Vendor Tricks

Knowing these tricks will save you money. Every single one of these is used regularly on MQL5 Market and other EA marketplaces.

1. Cherry-picked test periods. The vendor runs backtests across dozens of date ranges and publishes only the one where the strategy happened to perform best. A strategy that made 200% from Jan-Jun 2024 might have lost 80% from Jul-Dec 2024. You'd never know.

2. Unrealistic spreads. Testing with 0 spread or best-case fixed spreads. Gold with 0 spread doesn't exist in real trading. This alone can turn a losing strategy into a "winner" on paper.

3. Low-quality modeling. "Open prices only" runs fast and can look just as good in screenshots. But it completely ignores intra-bar price action. Stop losses, take profits, and trailing stops all trigger at wrong prices.

4. Balance line only. The vendor shows you the balance curve (closed trades only) instead of the equity curve. This hides floating drawdowns from open positions. Any strategy holding multiple positions can show a smooth balance line while carrying significant unrealized losses.

5. Overfitting through optimization. The developer runs MT5's optimizer with hundreds of parameter combinations, picks the one that looks best, and presents it as the "real" result. This strategy worked perfectly in the past but has zero predictive power for the future.

6. Missing trade count. A screenshot showing massive profits but cropping out the trade count. 20 trades over 5 years has zero statistical value. It's noise, not edge.

What a Legitimate Backtest Report Looks Like

A serious developer doesn't hand you a screenshot. They publish the full backtest report — every trade, every metric, across years of data. The entire trade history should be available for inspection, not just a summary.

See an example of what a transparent backtest report looks like — full trade log, equity curve, and all metrics exposed for scrutiny.

Beyond raw results, legitimate developers explain how they validate their strategy. Do they use walk-forward analysis? Out-of-sample testing? Cross-validation? If a vendor has never heard of these terms, their "profitable backtest" is almost certainly the product of overfitting.

Read about walk-forward validation in practice — the methodology behind building strategies that survive real markets.

Beyond Backtesting — What Else to Check

A great backtest is a starting point, not a finish line. Here's what separates real strategies from paper fantasies:

• Forward testing on demo accounts. At least 3 months of live demo results. This eliminates look-ahead bias and confirms the strategy works in real-time, not just in hindsight.
• Verified live signals. MQL5 Signals or MyFXBook with verified broker connections. If a vendor refuses to publish verified live results, ask yourself why.
• Multi-instrument testing. Does the strategy only work on XAUUSD H1? A robust strategy shows its logic works across related instruments or timeframes, even if it's optimized for one.
• Sensitivity analysis. Change the key parameters by 10-20%. Does the strategy still work? If a small parameter change destroys profitability, the strategy is overfitted to a specific historical pattern that won't repeat.

The best traders treat backtests as evidence to be scrutinized, not results to be celebrated. Now you have the tools to do exactly that.