A strategy that loses money live almost always lost money in its backtest first, if anyone ran one properly before going live. The three platforms a broker or prop firm is most likely to compare, MetaTrader 5, cTrader and eTrader, each ship their own way to replay a strategy against historical data before it touches a real account, and the three are built differently enough that "we backtested it" means a different thing depending on which one did the testing.
MT5 Strategy Tester
MetaTrader 5's Strategy Tester runs an Expert Advisor against historical price data inside the terminal itself. It is multi-threaded, so it can test the same Expert Advisor across many parameter combinations at once during an optimization pass, and it includes a visual mode that steps through the historical chart so a developer can watch the Expert Advisor's decisions candle by candle rather than only reading a final report. The result is a report of trades, profit and drawdown for whichever historical window and parameters were tested.
cTrader Automate backtesting
cTrader Automate, Spotware's C# environment for cBots, backtests the same way in spirit: a cBot runs against historical data and produces a report. The data source is a real choice a developer makes before running it. Options include tick data pulled from the server, which cTrader's own documentation describes as the most accurate source, M1 or H1 bar data from the server, or a custom CSV file the developer supplies. cTrader Automate also has a real-time visual mode, where historical data is fed to the cBot at an adjustable playback speed so the developer can watch it trade, alongside a standard non-real-time mode that only shows the final result, and an optimization pass for tuning a cBot's parameters. The finished backtest produces an HTML report with an equity chart, trade statistics and drawdown figures.
Source: help.ctrader.com, "Backtest a cBot".
eTrader Code's cloud backtester
eTrader Code takes a narrower, more deterministic approach. A robot backtests against the firm's own real price history rather than a file the trader has to source and load themselves, and the same bytecode, inputs and candles always produce the same trades on replay. There is no separate export step and no choice of data quality to get wrong. The result shows an equity curve and every trade the robot made over the tested period, and because the editor flags mistakes as the robot is typed, most errors are caught before a backtest is even run rather than discovered afterward in a broken report.
Backtesting compared
| Platform | Language | Data source | Modes | Output |
|---|---|---|---|---|
| MT5 Strategy Tester | MQL5 | MetaTrader's own historical price data | Multi-threaded testing, visual mode, optimization | Trade and profit report |
| cTrader Automate | C# | Tick data, M1 or H1 bars from the server, or a custom CSV | Non-real-time or real-time visual mode, optimization | HTML report with equity chart and statistics |
| eTrader Code | Tide | The firm's own real price history | Deterministic replay, editor flags mistakes as typed | Equity curve and every trade shown |
What determinism actually buys a firm
The practical difference between these three shows up before a single trade is even simulated. On MT5 and cTrader, the developer decides which historical data feeds the test, and that choice matters: tick data and server bars can differ from a broker's own historical spreads and fills, which is one reason backtest overfitting is such a well-known trap on both platforms. eTrader Code removes that decision by testing directly against the firm's own real price history, so the backtest reflects the prices traders on that firm's servers actually saw, not a generic dataset loaded from elsewhere.
Determinism is the second half of it. Because eTrader Code replays the same bytecode against the same candles every time, a robot's backtest result cannot drift between runs the way a floating-point strategy tester with slightly different data or rounding might. That matters most to a firm that wants to sign off a robot once and trust the number it produced, rather than re-run a test twice and get two different answers.
Where MT5 and cTrader still have the edge
Neither MT5 nor cTrader should be written off here. MT5's multi-threaded Strategy Tester and optimization pass are built for scanning a large parameter space quickly, which is genuinely useful when a developer is hunting for the best settings across thousands of combinations rather than confirming one specific version of a robot. cTrader Automate's choice of tick data, M1 or H1 bars, or a custom CSV gives a developer control over exactly how granular the historical simulation is, at the cost of that developer needing to source and manage that data themselves. Both are mature, well-documented tools that plenty of developers already know how to use, and a firm running Expert Advisors or cBots already has that workflow in place.
For a firm comparing all three before choosing where new robots get built, the honest summary is that MT5 and cTrader trade some setup effort for more control over the test itself, while eTrader Code trades some of that control for a backtest that is simpler to run correctly the first time, on the firm's own historical data, with no export and reimport step in between. See our wider look at backtesting tools compared for how these fit alongside other platforms, or book a call to see eTrader Code's backtester on real price history.
"A backtest is only useful if running it twice gives you the same answer. That is the whole point of testing a robot against the same bytecode and the same candles every time."
— The SGHK Team
Key Takeaways
- MT5's Strategy Tester is multi-threaded, with a visual mode and an optimization pass for scanning many parameter combinations.
- cTrader Automate backtests cBots against tick data, M1 or H1 bars from the server, or a custom CSV, with a real-time visual mode and its own optimization tool.
- eTrader Code backtests deterministically against the firm's own real price history, showing an equity curve and every trade.
- The eTrader Code editor flags mistakes as a robot is typed, catching many errors before a backtest even runs.
Frequently Asked Questions
Is MT5's Strategy Tester multi-threaded?
Yes. It can test an Expert Advisor across many parameter combinations at once during optimization, alongside a visual mode for stepping through a single run candle by candle.
What data does cTrader Automate use for backtesting?
A developer can choose tick data from the server, which cTrader's own documentation describes as the most accurate option, M1 or H1 bar data from the server, or a custom CSV file they supply themselves.
Does eTrader Code let a robot backtest against any historical data file?
No. eTrader Code backtests against the firm's own real price history rather than an uploaded file, and the same bytecode, inputs and candles always produce the same trades on replay.
Why does determinism matter in a backtest?
A deterministic backtest gives the same result every time it is run against the same inputs, so a firm can sign off a robot's backtest once rather than worry the number might change on a second run.
Which of the three is easiest to start backtesting on?
eTrader Code needs no data file, since it tests against the firm's own real price history directly, and its editor flags mistakes as the robot is typed. MT5 and cTrader give a developer more control over the historical data source, at the cost of managing that data themselves.
About SGHK
SGHK is a FinTech company that designs and builds its own software for the trading industry: the eTrader trading platform, Launch your Broker and Launch your Prop Firm. Every product is written, hosted and supported in-house and licensed to trading firms, with the CRMs branded to them, all hosted by us in the cloud, managed by each firm and built to scale across clustered servers as our clients grow. Everything is encrypted, and each firm is the only one with access to its data and its clients' data.