The Strategy Tester at a beginner level

12 minutesLesson 8 of 10
Femi, the HustlCraft strategy mascot

MT5 has a machine inside it that replays the past and tells you how a set of rules would have done. That is a useful thing to own. It is also the tool behind almost every screenshot that has ever separated a beginner from his savings, so we will be exact about what it knows and what it is only guessing.

Where it is and what it tests

Go to View, Strategy Tester, or press Ctrl+R. A panel opens at the bottom of the terminal with an Overview tab and a Settings tab. The tester runs a program, an Expert Advisor from your Navigator, over past price data and records every trade it would have taken. It does not test you, your hesitation or your Friday afternoons. It tests a written rule a program can follow, which is the first honest thing it tells you: a plan that cannot be written that clearly cannot be tested this way at all.

  1. In Settings, pick the program in the first drop-down, then the Symbol and the Timeframe your plan uses.
  2. Set the Date range. Custom period, and at least a year, so the rule meets more than one kind of market.
  3. Set Modelling. Every tick based on real ticks is the closest to what happened; 1 minute OHLC and Open prices only are quicker and cruder.
  4. Set Deposit and Leverage to match your real account. Leave Optimization at Disabled for now.
  5. Click Start, at the right-hand end of the tester's tab row. The Backtest tab fills with figures, and a Graph tab draws the balance over time.
The Strategy Tester Settings tab: Expert Examples\Moving Average\Moving Average.ex5, Symbol EURUSD, timeframe H1, Date Custom period from 2025.09.09 to 2026.09.10, outlined, Forward No, Delays Zero latency, ideal execution, Modelling Every tick based on real ticks, outlined, Deposit 10000 USD, leverage 1:100, Optimization Disabled, and the tab row with Settings outlined.
The tester set up for MetaQuotes' own example program, Moving Average, on a year of EURUSD hourly data. The empty rows below are cut out.

Trading carries risk and you can lose money.

What it can tell you

Read four numbers on the Backtest tab and ignore the rest at first: Total trades, Profit factor, Balance drawdown maximal and Total net profit. The graph matters more than any of them, because a smooth line that rises and a jagged line that ends at the same place are two different lives to sit through.

One run, one change

  1. Run a plain moving average crossover on EURUSD H1 over the last twelve months. Say it reports 184 trades, a profit factor of 1.18 and a worst drawdown of 11 percent.
  2. Now open Settings again, change nothing except the spread from Current to a fixed 30 points, and run it again.
  3. Say it now reports the same 184 trades, a profit factor of 0.96 and a drawdown of 16 percent.

The rule did not change. The cost did, and the edge went from thin to gone. That is the tester at its most useful: it shows you how fragile a rule is when one honest cost is added, exactly the lesson from a thin edge over 200 trades.

Trading carries risk and you can lose money.

The much longer list of what it cannot

It cannot knowWhy
What happens nextIt replays the past. A year that has happened tells you what the rule did, never what it will do.
Your broker's real spread at newsUnless you set it, the tester uses today's spread, or a fixed one, for every candle. Real spreads widen exactly when your stop is close.
Slippage, requotes and disconnectionsEvery fill is at the price on the data, at once, every time. Live execution is none of those things.
Swap and commission unless you enter themA rule that holds overnight can be profitable in the tester and a loser once the nightly charge is added.
The ticks inside a candleOn 1 minute OHLC or Open prices only, the tester invents the path price took between four numbers. A stop and a target inside the same candle is a coin toss it decides for you.
That you tuned the settings to fitTry twenty values, keep the best, and the report describes a rule that fitted noise. This is over-fitting, and the tester will never warn you.
What you would have doneYou did not sit through the eight losing trades in a row. The program did, without feeling anything.

This is the same warning as Part One's why backtests and demos mislead, now with the exact buttons in front of you. A backtest is evidence, weak evidence, that a rule is not obviously broken. It is never evidence that the rule works.

Femi, the HustlCraft strategy mascot
Femi's warning

A tester report is a story about a road that has already been driven, on a day with no traffic. When someone shows you a rising line from the tester and calls it proof, they are selling you the empty road.

How a beginner should use it

The picture to keep

The tester is the empty road behind the driving school: it proves the car starts and the brakes work, and it says nothing about Lagos traffic on a wet Friday.

PreviousNext: Reading the History tab as a journal