This guide explains how to use the in-platform cloud backtesting function that allows you to backtest your Profectus strategies without having to leave the platform.
Backtesting explained
Backtesting is the process of running your strategy on historical data. The idea is to gather information about the reliability of your strategy and adjust your strategy based on the information that is presented in the backtest.
Backtesting is the most time-consuming activity of algo trading because you need to collect a lot of data before you can verify that an idea actually works. We have a lot of experience with the backtesting process, and we know how annoying it is to switch between the Profectus platform and your backtesting environment.
You want to be able to quickly see if an idea has the potential to become something successful. A quick backtest with some general parameters can already give you some insights into the potential. This is the exact process that we wanted to make more efficient, and we have been able to do this.
How to use the Profectus Backtester
In the upper-right corner of your builder canvas, you find 2 new buttons:
Test Strategy - Backtests
With your strategy open, click TEST STRATEGY (top right).
In the dialog, set:
Project — which strategy to test (defaults to the one you have open).
Asset — the instrument, e.g. GBPJPY.
Timeframe — e.g. H1.
Date range — any window between 2015 and 2026.
Click START BACK-TEST.
The test will now start running in the cloud, and you can just continue to build or do whatever you like. As long as you keep Profectus open in a tab on your browser, the backtest will run in the cloud uninterrupted.
On average, a back-test takes between 2 and 15 minutes, depending on the current demand.
As a free user, you have a limit of 5 back-tests you can run per month.
Reading your backtest results - The Backtests panel
Every run is saved under its project in the right-hand Backtests panel, showing at a glance:
Net Profit (as a %)
PF (profit factor)
Win Rate
You can keep multiple runs per project, useful for comparing a tweak against the previous version. Each run has a VIEW button (opens the full report) and a delete button. You can also expand a run to see its asset, timeframe, date range, and the inputs used with a SEE CHART shortcut.
The full report (VIEW)
Clicking VIEW opens the multi-page Backtest Report, the same branded analysis the Profectus Analyzer produces. For a full explanation of every metric, see the Report Analyzer metrics guide.
You can Download the report as a PDF.
The interactive chart (SEE CHART)
SEE CHART opens a candlestick chart of the test period with every trade drawn on it:
▲ / ▼ marks each entry (buy / sell)
Green dots are winning trades, red dots are losing trades
A dashed line traces each trade's path from entry to exit price, with the pip result labelled
It's the trade-by-trade view, useful for seeing where on the chart your strategy did well or badly, not just the aggregate numbers. It shows a maximum of 200 trades.
Data and accuracy
Backtests run against real historical data for 32 instruments, spanning 2015 to 2026:
AUDCAD AUDCHF AUDJPY AUDNZD AUDUSD CADCHF CADJPY CHFJPY EURAUD EURCAD EURCHF EURGBP EURJPY EURNZD EURUSD GBPAUD GBPCAD GBPCHF GBPJPY GBPNZD GBPUSD NZDCAD NZDCHF NZDJPY NZDUSD SP500 US30 USDCAD USDCHF USDJPY XAGUSD XAUUSD
Available timeframes: M1, M5, M15, M30, H1, H4, H6, H8, D1, W1, MN1.
The data right now is Dukascopy 1-minute bar data, but we’re shifting to a different data source and will clearly notify you when this happens.
Important note on XAUUSD & XAGUSD: XAUUSD & XAGUSD contract specifications tend to differ a lot between brokers. As a result, the risk per trade might be different from what you put in your buy/sell trading blocks. To be safe, always run your strategy with the lowest lot size or % of risk (0.01) to see the impact in your backtest reports. Apply the same process before you ever want to put the trading system on a demo or live trading account.
FAQ
Q: What data is used for the Profectus backtester?
A: We are currently using Dukascopy 1 min OHLC data. In the near future, we might switch to Darwinex data, and we are also looking for the option to offer granular tick data backtesting.
Q: My backtest is stuck. What do I do?
A: On certain rare occasions, you might find that your backtest gets stuck and won’t finish. It will also not return a report in this case. You should see an error message if this happens, and you can just start a new backtest at any time.
Q: How accurate are my backtest results?
A: Always be very careful with the interpretation of your backtest results. Currently, the backtests are running on 1 min OHLC bars, and this is not the highest quality data. It’s fast and good enough to give you a good impression if you are on the right track and can pursue the idea you are working on.
Q: Can I run multiple tests at the same time?
A: Yes! The backtesting engine works with a queuing mechanism that simply queues all the jobs it receives. If you want, you can backtest multiple systems simultaneously, and they will simply return your reports in chronological order.
Q: Is my intellectual property safe when I use the backtesting feature? Are you stealing my project code?
A: Your intellectual property remains yours, and we are not storing or taking your code. The execution file is used to run the backtest, and we store the branded reports in our databases.