Flirting with Models
Dr. Ernest Chan - Tail Reaper (S3E5)
Episode guests
Podcast summary created with Snipd AI
Quick takeaways
- Machine learning can be effectively used as a risk management layer on tail reaper programs to reduce premium bleed and capture market movements during crisis periods.
- Advancements in machine learning techniques like random forests and dropout have helped overcome overfitting issues and improved the effectiveness of tail risk hedging strategies.
Deep dives
Applying Machine Learning to Tail Risk Hedging
Dr. Ernest Chan, founder of QTS Capital Management, discusses the intersection of tail risk hedging and machine learning. He explains his history with machine learning and how modern approaches can help overcome overfitting issues. Dr. Chan applies machine learning as a risk management layer on QTS’s tail reaper program, an intraday trend following model designed for crisis periods. By using an intraday breakout strategy on S&P E-mini futures, he captures market movements during periods of large volatility. The risk management layer uses machine learning to determine the probability of a profitable trade. Dr. Chan highlights the effectiveness of this approach in distributing tail risk and emphasizes the importance of appropriate hedge ratios.
Remember Everything You Learn from Podcasts
