Group backtest: one strategy on a whole list of symbols

The group backtest runs one AlertaScript strategy with the same settings on every symbol of a list. It shows how each symbol did and what the whole list did as one capital, so you can tell a strategy with an edge apart from one that only fits a single chart. It is a Pro Lab tool, part of the Pro and Elite plans.

Running a test

1. On a computer, press Pro Lab on the right-hand toolbar; on a phone, open the analysis centre from the chart's bottom bar and choose Pro Lab. Then choose Group backtest.
2. Choose the strategy: a ready template, one of My scripts, or strategy code you paste; pasted code is not saved. Only scripts with a strategy() call have trades to test.
3. Choose the symbols: one of your watchlists or a ready list (top crypto on spot, top perpetuals, US large caps, or FX and indices). Up to 40 symbols are tested.
4. Choose the timeframe (1 hour, 4 hours or 1 day) and 500, 1,000 or 2,000 bars per symbol, then start the test. Results fill in as the symbols finish, and stopping keeps what is done.

Reading the results

• The first line says on how many symbols the strategy made money and on how many it beat buying and holding.
• When one symbol brings 50% or more of all the profit, the tool says so: the group result leans on that symbol.
• The chart is the whole list as one capital, split equally across the symbols at the start and never rebalanced, against buying and holding the same list. The deepest drawdown is read from that curve.
• The table gives every symbol's net profit, buy and hold, trades, win rate, profit factor, maximum drawdown and a small equity curve. Sort by any column and open a symbol on the chart from its row.

Good to know

• Each symbol runs the strategy with its saved settings, on the same timeframe and the same number of bars, through the script engine the Strategy Tester uses.
• Symbols that could not be tested, for example with too little history, are listed with the reason and left out of the totals.
• A strategy that works on most of a list is more likely to have an edge than one that shines on a single symbol. Past results still promise nothing.