152: "Backtesting trading strategies does not work" - John Ehlers
"Backtesting trading strategies does not work!”
Got your attention?
Maybe you agree with this statement.
Perhaps you strongly disagree and you’re currently heading out to the backyard to grab your pitchfork.
Or maybe you’re shaking your head thinking ‘what the heck swanny, have you finally lost it man?’.
Well, the good news is I haven’t lost it (yet). We're going to dive deeper into this statement in the podcast episode today.
But first, I'd like to introduce our guest - John Ehlers.
John is a friend of the show. He’s been a guest multiple times, discussing topics such as cycles, indicators and digital signal processing. In our chat today we’re going to tackle robustness and also intraday trading.
Some of the things you’ll hear on the show today are:
- The startling differences between intraday and daily timeframes and the critical factors you need to watch out for,
- Why it’s more difficult to predict where market prices are going on an intraday basic compared to daily timeframe, and how you need to think about intraday data differently,
- We’ll be digging deeper into this “Backtesting trading strategies does not work” statement to find out what it means,
- How to use a simple Genetic Optimization trick to determine the robustness of a strategy
- How Genetic Optimization can be used to identify suitable ranges for optimization parameters
- Why short walk forward periods could be better than longer,
- Plus a whole lot more.
So lets jump over to my chat now with John Ehlers.