AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Walk Forward Test Robustness
I would describe as in sample and out of sample, which is exactly what I do. Andre Arunga used to teach was he would take the complete data set. He had a very good idea of whether something was over-optimized or not. In all honesty, I don't really use it. We've got software, multi-charts does it automatically. That's helpful. But walk forward analysis is completely different. It's taking chunks and altering the settings on predetermined time period,. Like every year, looking, optimizing for the last, say, four years; then doing that in chunks.