You’re backtesting an intraday trading strategy that trades new highs.
You find a column/filter that improves the results.
It eliminates about 1/3 of the trades – say, 300 of the original 900 in the trade set – but the average profit goes up… a lot.
So you apply the filter and re-run the backtest, expecting to see a trade set with 600 trades (900 – 300, of course.)
But to your surprise, you find there are over 700 trades in the results.
When you look closer, you see some trades you thought would be removed, but they weren’t.
What’s going on? Is the backtest broken?
Have you experienced this common problem?
Any guesses about what’s going on here? Hit reply and let me know.
-Dave
