Forex Tester Backtesting: Manual Backtesting on Desktop and Online

How Forex Tester backtesting works on desktop

Historical data

Basic Data comes with the license and covers the main currency pairs; Super Data is a paid subscription that adds tick data, floating spread and many more symbols. Own data can be imported as .txt, .csv or .hst files, and the FAQ notes that it is recommended to import 1-minute bars for accurate testing, but the program will work well with any timeframe. The terms add: All downloaded data is stored on your computer and remains valid for use after the subscription expires. The free data table describes its series as Price: Bid · Time: GMT (no Daylight Saving Time) · source Basic (Forexite free data).

The simulated account

A test starts with 10,000 USD of virtual deposit, which can be changed, and the features page lists the settings that bring the simulation closer to a given broker: spread, leverage, swaps, initial balance, minimum distance to price, digits after profits, margin, hedged margin, lot size, simulated deposits and withdrawals, and broker commission. Costs set too low make a manual result look better than it would have been.

Testing modes

• Ordinary mode: step back by one bar and change the trading decision. Useful for practice, but stepping back is a form of hindsight.
• “Forward testing only”: no stepping back. Every decision stays as it was made, closer to real-time trading.

Trading on the replayed chart

The chart area shows any number and combination of charts of various pairs and time frames, synchronized, as Japanese candlesticks, bars, range bars, lines and Heikin Ashi, with Renko charts, custom timeframes and built-in indicators. For sizing and exits there are risk as a % of equity per trade with automatic lot size, a Reward-to-Risk tool on the chart, and trailing stops. News can be part of the test: a historical news service shown on the chart; only USD-related high-priority news with Basic data, 9 currencies with high, medium and low priorities with Super data.

Statistics

The result of a session is in the numbers, not in the memory of the trades: automated trading statistics (profitable and losing trades, drawdowns, max drawdown, consecutive wins and losses and more), profit chart (balance, equity, margin, drawdown), export of statistics, trading history and profit chart to Excel. A manual backtest is a hypothetical result: it shows how a set of rules would have been traded on past data, under the costs set in the simulation, and says nothing certain about the future.

Automated testing and the Strategy Optimizer

The same software tests Expert Advisors (EAs): built-in examples, custom EAs in .dll format, or EAs made without coding in Easy Forex Builder, a separate tool included for buyers. An EA can run on several pairs and timeframes, on real tick data with floating spread, next to other EAs, and manual and automatic trading can be combined. The Strategy Optimizer searches an EA’s parameters with a full grid, available even without a paid data subscription, or with the AI-based Neural Network and XGBoost methods. Optimized parameters fit the past by construction, so an optimized result needs unseen data before it means anything.

The demo version

What the numbers mean in bars

A forex chart prints about 120 one-hour bars a week (24 hours, 5 days). The free EURUSD series spans about 23 years and 5 months: roughly 146,000 one-hour bars, 36,600 four-hour bars or 6,100 daily bars (approximate, holidays ignored). In a manual test every bar is a moment to look at, so the timeframe decides the workload: a year of one-hour data is about 6,200 bars, a year of daily data about 260.

The demo month has 21 weekdays, about 500 one-hour bars: enough to learn the controls and the testing modes, not enough for a sample of trades on a daily chart, where the same month is about 21 bars.

GMT without Daylight Saving Time matters for session rules. The London open, 08:00 local time, is 08:00 GMT in winter and 07:00 GMT in summer, so on this data “enter at the London open” means a different bar in the two halves of the year.