Adrian Reid warns that many impressive backtested trading systems are over‑optimized to historical data, yielding strong past performance but failing in live trading. He emphasizes avoiding the pursuit of perfect equity curves, the danger of excessive fine‑tuning, and the need to learn proper testing and validation methods (taught in his Trade Success System).
Any amazing trading system you see online is over optimized. And what that means is it looks amazing on historical data, but it performs terribly in the future. And that's really critical because we need to know the difference between a well optimized system and an over optimized system. And that's part of knowing how to test well. And that's a big thing. I teach that in the trade success system.
But you know, the key is stop looking for perfection. Because when you look for perfection, that amazing system with the laser sharp equity curve that has tiny drawdowns and makes 30 return per year and all that, when you look for that, you end up over optimizing. And that then doesn't work in the future. So what happens is you have a strategy that looks amazing and then you start trading it and it loses money. You're like, what happened? Well.
You fine tuned it so much to the past data to make it look amazing on the back. Test that in real time trading, it doesn't work.
So that's absolutely a thing. That's one of the biggest challenges with systematic trading, which is why you need to learn to test for yourself and learn to test properly.
Sat.
