Platforms
Backtesting
Backtesting is running a set of trading rules on historical price data to see how those rules would have performed in the past.
Backtesting applies fixed rules to old data. You define the entry, exit, stop-loss and quantity, run them over months or years of candles, and the software lists every hypothetical trade with its profit, loss and drawdown. AmiBroker has a full backtester driven by AFL. TradingView has the Strategy Tester for Pine Script strategies.
It matters because it replaces opinion with a record. You learn how often a method trades, how long its losing streaks run, and how deep the drawdown gets. For Indian intraday traders that last number is the useful one. It tells you whether you could have sat through the bad month without abandoning the rules.
A back-test is only as honest as its assumptions. Indian intraday results change a great deal once you add brokerage, statutory charges and slippage on every trade. Futures data needs correct handling of expiry and rollover. Clean historical data for option strikes is hard to get. A test on a continuous index chart with zero costs says little about what a real account would have done.
The common mistakes are curve fitting and repainting. Curve fitting is tuning parameters until the past looks perfect; the rules then fail on new data. Repainting is worse: an indicator that redraws its signals after the fact will show entries no live trader could have taken. Always hold back a stretch of data the rules have never seen, and test on it last.
Worked example
What costs do to a back-test
Illustrative numbers. A 5-minute Nifty futures rule-set shows 400 trades in a year averaging 6 points gross, or 2,400 points. Assume costs and slippage of 4 points per round trip. The net average is 2 points and the year shrinks to 800 points, one third of the headline. At 7 points of cost the same rules lose 400 points. The signals did not change. Only the assumptions did.
In WizeBuySell
WizeBuySell plots signals on your chart. It is not a backtesting report, and we do not publish back-test results or accuracy figures. Such numbers depend on costs, slippage, quantity and discipline that differ for every trader. Signals are confirmed on candle close and do not repaint. Any back-test you run is your own work and responsibility. The 2-day free trial lets you watch signals form live first.