Step-by-step guide to backtesting trading strategies

Stop Guessing: How to Actually Backtest Strategies on TradingView

Most traders fail not because their logic is bad, but because they never test it against market reality. Backtesting is the process of running your trading rules against historical data to see if the strategy makes money in the past. If it doesn’t work in the past, it almost certainly won’t work in the future. TradingView is the most accessible tool for doing this, but it has pitfalls that can give you false confidence. Here is the concrete, step-by-step process to do it right, without wasting time on vanity metrics.

1. Pin Down Your Rules Before You Touch the Mouse

You cannot test a vague idea like "buy when it looks low." TradingView’s strategy tester requires binary, mathematical rules. You must define:

• Long Entry: Exact condition (e.g., RSI < 30 AND Price > Moving Average 50).
• Long Exit: Take profit percentage, stop loss percentage, or a specific indicator cross.
• Short Entry/Exit: Only include these if your logic allows shorting. If you only go long, leave shorts disabled.

2. Use the Strategy Tester, Not the Indicator Panel

A common mistake is using regular indicators to visually scan charts. This is subjective and impossible to quantify. You need to use Pine Script to create a Strategy, not an Indicator.

1. Open the Pine Editor at the bottom of the TradingView chart.
2. Click "New blank file".
3. Type "strategy()" at the top of the file. This initializes the backtesting engine.
4. Add your strategy.entry(), strategy.exit(), or strategy.close() commands based on your rules.

3. The Critical Settings: Slippage, Fees, and Bars

Default settings often make strategies look better than they are. Adjust these before judging performance:

• Slippage: Set this to 100-200 ms or a specific basis point depending on your asset. Slippage is the difference between the price you want and the price you get. Ignoring it overstates your profits.
• Fees and Commissions: Enter the actual transaction costs from your broker. Spreads, commissions, and swap fees eat into net profit. A strategy with a $10 gross profit but $8 in fees is not profitable.
• Price Resolution: Test on at least 1 Hour or 4 Hour charts if you are a swing trader. Testing on 1-Minute charts introduces noise that may not reflect broader market trends relevant to your holding period.

4. Interpret the Metrics That Matter

The Strategy Tester outputs many numbers. Ignore "Profit Factor" alone and look at this combination:

• Net Profit: The absolute dollar amount. A 20% return on a $100 account is $20. A 5% return on a $10,000 account is $500. Context matters.
• Max Drawdown: The largest peak-to-bottom decline. If your strategy loses 50% of the account at any point, most retail traders won’t have the psychological fortitude to keep using it. Aim for drawdowns under 20%.
• Win Rate vs. Risk/Reward: A 40% win rate is fine if you make 3x on wins and lose 1x on losses. A 80% win rate is useless if one loss wipes out 10 wins.
• Expected P&L: This accounts for the probability of each trade. If this is zero or negative, walk away.

5. Out-of-Sample Testing to Prevent Overfitting

This is the step most people skip. If you optimize a strategy on data from 2018–2023, it will look perfect. Then you trade it in 2024, and it fails. You must reserve a period of data that your strategy has never seen. For example, if you optimize on 2015–2020, you must test the exact same parameters on 2021–2023 data. If the strategy performs poorly on this "out-of-sample" data, it is curve-fitted and worthless. If it performs consistently across both periods, it has statistical robustness.

6. Forward Test on Paper

Backtesting is historical. It assumes you could execute every trade perfectly at the close of the bar. In reality, you might not see the signal, or the market might gap against you. After a successful backtest, run the strategy in TradingView’s "Paper Trading" mode for at least two weeks. Do not risk real capital until the paper trades match the backtest results closely.

When You Need More Control

If manual Pine scripting feels too slow or you want to test complex risk management rules without coding every line, there are tools designed to automate the construction of these strategies. The Automated Strategy Builder offers a free demo of a platform that helps generate and structure automated trading strategies, allowing you to test logic with less friction if you prefer a guided environment over writing code from scratch. You can check it out here: Automated Strategy Builder
free demo of a platform that builds automated trading strategies
See it

More guides

• How to Build an Automated Trading Strategy Without Coding