Episodebeskrivelse
Can deep learning outperform traditional quant strategies? In this episode, we explore how a simple neural network model was applied to momentum trading — and how it stacks up against the market.
Inspired by Richard Sutton’s Bitter Lesson, this study puts brute-force computation to the test in financial prediction. We walk through the data setup, model architecture, rolling validation process, and — most importantly — the results. Despite only achieving 52% classification accuracy, the model delivered an annualized return of 12.8% with strong risk-adjusted performance.
We also compare the…
Website: podcasters.spotify.com

