My guest in this episode is Dr. Ernest Chan, founder of QTS Capital Management.
Investor, researcher, and educator, Ernie is well-known for his blog – which he has been publishing since 2006 – as well as the several books he has authored, including Quantitative Trading, Algorithmic Trading, and Machine Trading.
Our conversation meets at the intersection of tail risk hedging and machine learning. Ernie has a long history with machine learning, having first applied it on Wall Street in the late 1990s. After striking out on his own in 2006, he abandoned it due to the overfitting issues he believed it suffered.
In recent years, however, Ernie has re-adopted machine learning, believing that modern approaches help circumvent the overfitting problems and create robust, reliable models.
Specifically, Ernie applies machine learning as a risk-management layer on QTS’s Tail Reaper program, an intraday trend-following model designed to profit in periods of crisis. We discuss why such a program can be effective as a tail hedge and how the risk management layer can potentially help reduce the premium bleed typically associated with tail programs.
For listeners keen on understanding modern applications of machine learning, this is not one to miss.
