Backtesting data granularity?

Trying to determine if tick data is necessary for backtesting, or would minute data be granular enough? Anyone tried both?

1 Like

Hi AlgoTrader2018. It will really depend on the strategy that you are using. If your strategy does not execute more than one time every few minutes then minute data should work just fine. If it is going in and out of trades multiple times a minute then you will want to consider tick data for backtesting. Tick data can give you more precision in your trade entries and potentially help you get into trades sooner and find potential reversals before they occur on the one-minute chart.

1 Like

If you have an intrabar strategy, then tick data is definitely the way to go. But even with a close bar strategy, there are times when you may need to tick data as well.

Imagine backtesting a strategy with a stop/limit attached to each trade (like most strategies do). What if during a backtest, there is a candle where the High and Low prices would have hit both your limit and your stop on an existing trade? Which order was hit first? Did the trade make money or lose money? If you only have candlestick data, there is no way of knowing.

Tick data, however, would tell you whether price hit your limit first at a profit or price hit your stop first at a loss. I consider tick data vital when using stops and limits in a backtest.