2 Comments
User's avatar
TradingQuant's avatar

I stopped considering Out-of-Sample as a strategic validation tool when I started using it primarily as an invalidation criterion. I find it extremely useful to question and refute some of my more questionable work.

Expand full comment
Nam Nguyen Ph.D.'s avatar

Just like backtest, you still need to use OOS with caution. Incorporating it into a WFA improves the results.

Expand full comment