Choose your data source (NonA internal data or Yahoo Finance) and specify the trading instrument by entering a stock symbol or selecting from the dropdown....
📖
How It Works
1
Select Data Source and Symbol
Choose your data source (NonA internal data or Yahoo Finance) and specify the trading instrument by entering a stock symbol or selecting from the dropdown. The data source determines the depth and granularity of available historical data.
2
Define Date Range and Timeframe
Set the start and end dates for your backtest period, then select the candlestick timeframe (1M, 5M, 15M, 30M, 1H, 2H, 4H, 1D, or 1W). Longer date ranges with smaller timeframes produce more data points but require more processing time.
3
Configure Capital and Position Sizing
Set your initial capital amount and order size to simulate realistic trading conditions. The order size determines how many shares or contracts are traded per signal, directly impacting the equity curve and drawdown calculations.
4
Choose LLM Provider and Model
Select the AI language model provider and specific model that will power your strategy's intelligent components. Different models may produce varying results for AI-assisted signal generation and strategy optimization.
5
Verify Strategy Configuration
Review the complete strategy setup in the execute section, confirming that all parameters, indicators, and conditions are correctly configured before launching the simulation. Check the workflow grid for any incomplete rows.
6
Launch the Backtest
Click the Execute button to submit the backtest to the processing engine. A loading overlay appears showing real-time progress as the backend processes your strategy against historical market data.
7
Monitor and Review Real-Time Results
Once the backtest completes, results stream into the Perspective viewer in real-time via WebSocket. Examine the equity curve, trade markers, and summary statistics directly in the modal. You can also navigate to the Backtest Results page for deeper playback analysis.
ℹ️
Note
Backtest execution simulates your strategy against historical market data. Ensure you have sufficient data points (minimum 50 records) for the selected date range and timeframe combination.
💡
Tips & Best Practices
💡
Tip
Use at least 2-3 years of daily data or 6+ months of hourly data to ensure your strategy is tested across diverse market conditions including bull, bear, and sideways markets
💡
Tip
Set realistic initial capital and order sizes that match your actual trading budget to get meaningful risk metrics like maximum drawdown percentage
💡
Tip
Run the same strategy across multiple symbols to check generalizability before committing to live deployment on a single instrument
💡
Tip
If the backtest returns a data insufficient error, widen your date range or switch to a larger timeframe to ensure the minimum 50-record threshold is met
❓
Frequently Asked Questions
What is the minimum amount of historical data required to run a backtest?
The system requires a minimum of 50 data records for the selected date range and timeframe combination. If your selection yields fewer records, you will receive a DATA_INSUFFICIENT error and need to widen the date range or increase the timeframe.
What data sources are available and how do they differ?
The platform offers NonA internal data and Yahoo Finance data. NonA data may provide deeper historical coverage and higher granularity for certain instruments. Yahoo Finance provides broad coverage of US equities and global indices with standard OHLCV data.
How long does a typical backtest take to complete?
Execution time depends on the strategy complexity, date range, timeframe, and server load. Simple strategies on daily data typically complete in 10-30 seconds. Complex multi-indicator strategies on minute-level data can take several minutes. Progress is shown in real-time via the loading overlay.
Can I run multiple backtests simultaneously?
Each execution session processes one backtest at a time. You should wait for the current backtest to complete or encounter an error before starting a new one. Completed results are preserved and can be reviewed later on the Backtest Results page.
Does the backtest account for transaction costs and slippage?
The current simulation engine focuses on signal accuracy and equity curve generation based on OHLCV data. Transaction costs, slippage, and market impact are not modeled by default. Factor these into your real-world expectations when evaluating strategy viability.
⚠️
Important Notes
⚠️
Warning
Past performance does not guarantee future results. Backtesting results are hypothetical and may not reflect actual trading outcomes due to market conditions, slippage, liquidity, and other factors not captured in simulation.
Was this page helpful? Your feedback helps us improve the documentation.