The best backtesting software for forex depends on what you are testing. If you already have an Expert Advisor, the free MetaTrader 5 Strategy Tester is the reference: it runs the real EA on your broker's own history, with tick-based modelling. If you trade cBots, cTrader's built-in backtesting does the same job. If you want to turn trading rules into a tested strategy without writing MQL or C#, AlgoPuzzle builds the rules from blocks, backtests them in the browser on one-minute data with realistic spreads and commission, adds a Monte Carlo check, and exports the same strategy for MT5, MT4 or cTrader. TradingView's strategy tester suits traders who write Pine Script, and StrategyQuant X suits those who want software to generate and stress-test many strategies. If you trade by hand and want to practise, replay tools such as FX Replay and Forex Tester let you step through history and place trades yourself. The real dividing lines are four: automated rules or manual replay, whether you need to code, what the free version allows, and whether the result can become a strategy you can actually run. This comparison covers all seven, with facts about other products taken from their own sites and published guides as of November 2026.
Disclosure: AlgoPuzzle is our own product. Prices and plans of other tools change often; check their sites before deciding.
Forex backtesting software at a glance
Seven tools, compared on the questions that decide which one fits. "Automated" means fixed rules are replayed over history; "manual" means you replay the chart and decide the trades yourself.
| Software | Type | Runs in | Coding needed? | Free option | Result you can trade |
|---|---|---|---|---|---|
| MT5 Strategy Tester | Automated | MetaTrader 5 (desktop) | MQL5, or an existing EA | Free with MetaTrader 5 | The EA you tested |
| cTrader backtesting | Automated | cTrader (Automate) | C#, or an existing cBot | Free with cTrader | The cBot you tested |
| AlgoPuzzle | Automated | Browser | No, rules from blocks | Free plan: 5 backtests a day, last 3 months | Exported MT5 / MT4 EA or cTrader cBot |
| TradingView strategy tester | Automated | Browser / app | Pine Script, or a published script | Basic testing on all plans | A TradingView strategy, not an MT5 EA |
| StrategyQuant X | Automated, generates strategies | Desktop | No for generation; settings are complex | Trial; paid | Exported strategy code |
| FX Replay | Manual replay | Browser | No | Limited free plan | Your own practice record |
| Forex Tester | Manual replay | Browser (Forex Tester Online); older desktop version | No for manual testing | Paid; check current trial | Your own practice record |
Automated vs manual backtesting: which do you need?
Automated and manual backtesting answer different questions, so the first choice is the type of test, not the brand. An automated backtest asks "would these exact rules have made money?"; a manual replay asks "would I have traded this well?".
| Automated backtesting | Manual replay | |
|---|---|---|
| What is tested | Fixed rules | Your decisions |
| Repeatable | Yes, the same test gives the same trades | No, it depends on you each time |
| Speed | Months of data in seconds or minutes | As fast as you can click through bars |
| Bias risk | Over-tuning the rules to past data | Hindsight and inconsistent decisions |
| Needed for an EA or bot | Yes | No |
| Tools here | MT5 Strategy Tester, cTrader, AlgoPuzzle, TradingView, StrategyQuant X | FX Replay, Forex Tester |
If the end goal is a forex EA that trades for you, you need automated backtesting, because only fixed rules can be turned into code. Backtesting vs. Demo Trading explains where demo trading fits in after that.
The seven tools, one by one
MetaTrader 5 Strategy Tester. Built into MetaTrader 5 and free. It runs a compiled MQL5 Expert Advisor on the history of your broker's server, with modelling from open prices up to "Every tick based on real ticks", plus optimisation of inputs and multi-symbol tests. It is the most faithful test of an EA you will run on MT5, but you need the EA first. Step by step: How to Backtest an EA in the MT5 Strategy Tester.
cTrader backtesting. Part of cTrader Automate and free. It backtests and optimises C# cBots on tick or bar data. Like MT5, it tests a bot that already exists; it does not help you write one.
AlgoPuzzle. An EA builder with a backtester in the browser. You build the strategy from blocks (asset, timeframe, entry condition, lot size, stop loss, take profit, exit rule, trading hours), switch from Build to Backtest, and see the result next to your blocks. The same blocks then export as an MT5 or MT4 Expert Advisor or a cTrader cBot. Details of how it simulates trades are below and on the backtester page.
TradingView strategy tester. Tests strategies written in Pine Script, TradingView's own language, or published by other users. Basic strategy testing is available on every plan; Deep Backtesting over longer history needs a higher plan. A TradingView strategy runs on TradingView: it is not an MT5 or MT4 EA.
StrategyQuant X. Desktop software that generates large numbers of strategies from building blocks and tests their robustness (Monte Carlo, walk-forward and more). It is powerful and paid, and aimed at traders who want the software to search for strategies rather than test one idea. See Best EA Builders Compared.
FX Replay. A browser-based chart replay for manual backtesting: you step through history and place trades by hand. There is a limited free plan and paid plans; it does not run automated rules.
Forex Tester. A long-standing manual backtesting and trading simulator, now offered mainly as Forex Tester Online in the browser, with an older Windows desktop version. It is paid; check its site for the current trial.
How AlgoPuzzle's backtester works
AlgoPuzzle replays your rules on historical one-minute price data with the same logic as the exported Expert Advisor or cBot, so what you test is what you export. Each step of the simulation is fixed and documented:
| Step | What the backtest does |
|---|---|
| Checking the rules | Each rule is checked once per new candle of its timeframe, like the exported EA |
| Opening a trade | At the open of the candle that has just started: buys at the ask (bid plus spread), sells at the bid |
| Stop loss and take profit | Checked minute by minute on one-minute data for as long as the trade is open |
| Both hit in the same minute | The stop loss is assumed to be hit first, the conservative choice, and the results say how often it happened |
| Trailing stop and break-even | Simulated on each minute's open, high, low and close in order |
| Exit rule and trading hours | Follow the same rules as the exported strategy |
| End of the period | Open trades are closed at the last price and marked End of test |
The full method is in Running a backtest, and every number on the results screen is explained in Reading backtest results.
Spreads, costs and data in AlgoPuzzle's backtester
A backtest without trading costs flatters every strategy, so AlgoPuzzle fills in a realistic spread for each market and lets you change it. These are the cost and data settings:
| Setting | How it works |
|---|---|
| Spread | Prefilled with a typical retail spread for the market, editable from 0 to 500 pips. Where the spread recorded in the historical data for a minute was wider, the wider value is used |
| Spread check | If you enter a spread below the typical real spread measured for that market, the run is flagged as optimistic |
| Several markets | Each market is tested with its own typical spread, up to 5 markets in one backtest |
| Commission | $0 to $50 per lot, charged once per closed trade, for raw-spread accounts |
| Starting balance | $100 to $1,000,000, default $10,000 (USD) |
| Market data | Historical one-minute prices, delivered through AlgoPuzzle's own API to logged-in users only |
| Engine check | The engine is cross-checked against the MetaTrader 5 Strategy Tester |
| Not included | Swaps and slippage; see the limits below |
Why the spread matters so much: on a strategy that targets a few pips per trade, a 1 pip difference in spread can change the result completely. How spread raises your break-even win rate shows the arithmetic, and What Is Slippage? covers the cost no backtest fully captures.
Built-in robustness checks
A single backtest result can be luck. AlgoPuzzle adds two checks that most quick backtesters leave out:
- Monte Carlo analysis. Replays the backtest's closed trades 1,000 times by default, resampled or shuffled, and shows a bad-case profit and drawdown next to the backtest's own. It needs at least 30 trades. See Monte Carlo analysis.
- Notes on every run. Too few trades, stop loss and take profit in the same minute, positions still open at the end, indicator warm-up, a spread below the typical real one, a trailing stop tighter than normal one-minute swings. Each one is listed after the run instead of being hidden.
Even with both, a strategy tuned until its backtest looks perfect is likely curve-fitted. Read the results with profit factor, drawdown and expectancy, not just net profit.
What each backtester cannot tell you
Every backtester simplifies reality somewhere. Knowing where is more useful than any ranking:
| Software | Main limitation |
|---|---|
| MT5 Strategy Tester | Only as good as the broker history you download; needs an EA to exist |
| cTrader backtesting | Needs a cBot written in C# |
| AlgoPuzzle | One-minute data, so very tight trailing stops are only a rough guide; one spread setting for the period; swaps and slippage not included; a fixed list of popular markets |
| TradingView strategy tester | Strategies stay on TradingView; deeper history needs a higher plan |
| StrategyQuant X | Generating many strategies makes overfitting easy without strict out-of-sample tests |
| FX Replay / Forex Tester | Results depend on your discipline during replay; cannot test an EA's fixed rules |
The sound workflow uses more than one tool: test the idea quickly, confirm the exported EA in the MT5 Strategy Tester on the same symbol, timeframe and dates with "Every tick based on real ticks", then run it on a demo account.
Free backtesting: what each option includes
| Software | Free version | Main free limit |
|---|---|---|
| MT5 Strategy Tester | Full tester | Needs an MQL5 EA |
| cTrader backtesting | Full backtesting | Needs a C# cBot |
| AlgoPuzzle | Builder, backtest, Monte Carlo, export | 5 backtests and 5 exports a day; backtests on up to the last 3 months; free account |
| TradingView | Strategy tester on all plans | Deep Backtesting on higher plans |
| FX Replay | Free plan | Limited sessions, markets and indicators |
| StrategyQuant X / Forex Tester | Trial or demo at most | Paid products |
AlgoPuzzle's Pro pass, from $12.99 for 30 days and not a subscription, removes the daily limits and opens 6 months, a year or custom dates. Every limit is on the Pricing page.
How to backtest a forex strategy in AlgoPuzzle, step by step
- Build the strategy. Open the builder and set the asset, timeframe, entry condition, lot size, stop loss and take profit. Five simple rule types are a good start, and 19 Trading Indicators Explained covers the indicator blocks.
- Switch to Backtest. The header has a Build | Backtest switch; your blocks stay as they are.
- Choose the period. Last week, month or 3 months on the Free plan; 6 months, a year or custom dates with Pro.
- Set costs. Keep the suggested spread or enter your broker's, plus commission per lot if you pay one.
- Run it and read the results. Net result, return, maximum drawdown, profit factor, win rate, the trade chart and the full trade list.
- Check robustness. Open the Monte Carlo tab and read the notes on the run.
- Export and confirm. Export to MT5, MT4 or cTrader, run it in the platform's tester, then on a demo account.
Which backtesting software should you use?
| If you | Use |
|---|---|
| Have an MT5 EA and want the most faithful test | MT5 Strategy Tester |
| Have a cBot | cTrader backtesting |
| Have rules in mind and want a tested EA or cBot without coding | AlgoPuzzle, then confirm in the MT5 Strategy Tester |
| Write Pine Script and trade from TradingView | TradingView strategy tester |
| Want software to generate and stress-test many strategies | StrategyQuant X |
| Trade by hand and want to practise on history | FX Replay or Forex Tester |
Where to go next
| You want to | Read |
|---|---|
| See the backtester's settings and limits | Forex Backtester in Your Browser |
| Confirm an EA in MetaTrader | How to Backtest an EA in the MT5 Strategy Tester |
| Build a strategy to test | How to Build Your First Algorithmic Trading Strategy |
| Compare EA builders | fxDreema vs EA Builder vs MQL5 Wizard vs AlgoPuzzle |
| See strategies we have already backtested | Strategy of the Week |
| Avoid the usual automation errors | 7 Automated Trading Mistakes to Avoid |
Common questions
What is the best backtesting software for forex?
It depends on how you trade. For an Expert Advisor you already have, the free MetaTrader 5 Strategy Tester is the reference, because it runs the real EA on your broker's history. To build and test rules without writing code, AlgoPuzzle backtests them in the browser and exports them for MT5, MT4 and cTrader. For practising discretionary trading, replay tools such as FX Replay and Forex Tester fit better.
Is there free forex backtesting software?
Yes. The MetaTrader 5 Strategy Tester and cTrader's backtesting are free inside those platforms, TradingView's strategy tester works on its free plan, FX Replay has a limited free plan, and AlgoPuzzle's Free plan includes 5 backtests a day on up to the last 3 months of data with a free account. Each free option has limits on data, features or usage.
What is the difference between automated and manual backtesting?
Automated backtesting runs fixed rules over historical data and records every trade the rules would have taken, so the same test always gives the same result. Manual backtesting replays the chart bar by bar while you decide when to buy and sell, which practises judgement but depends on your discipline. EAs need automated backtesting.
Can I backtest a forex strategy without coding?
Yes. In AlgoPuzzle you build the rules from blocks and backtest them in the browser, then export the same strategy as an MT5 or MT4 Expert Advisor or a cTrader cBot. Manual replay tools also need no code, but they test your decisions rather than fixed rules. The MT5 Strategy Tester, cTrader and TradingView need code or an existing EA, cBot or script.
How accurate is a forex backtest?
A backtest is only as accurate as its data, its cost assumptions and the way it fills orders. Spreads, commission, slippage and swaps change results, and one-minute data cannot show the order of prices inside a minute. Use realistic costs, prefer conservative fill rules, and confirm an EA in the MT5 Strategy Tester and on a demo account before trading real money.
How much historical data do I need to backtest?
There is no fixed amount, but a result built on fewer than about 30 trades says little, and a period that covers only one kind of market can mislead. Test long enough to get a meaningful number of trades, then check the strategy on a period it was not tuned on.