SI391: Why Trend Following Works... the Evidence ft. Richard Brennan

The podcast explores the structural properties of financial markets that enable trend following, questioning why the patterns trend followers depend on exist. It highlights that markets possess memory, returns are not bell-shaped, and volatile periods cluster together. The hosts use the Hurst exponent to demonstrate long-range dependence in markets, indicating conditions that sustain trends persist once established. The discussion also covers how extreme daily moves occur far more frequently than standard models predict, revealing power law distribution and fat tails. An agent-based simulation shows that a threshold of divergent participation is needed for trends to emerge, and the recent crude oil spike is analyzed in this context.