Automated Volatility Forecasting
75 Pages Posted: 31 Jan 2021 Last revised: 20 May 2023
Date Written: May 19, 2023
Abstract
We develop an automated system to forecast volatility by leveraging over one hundred features and five machine learning algorithms. Considering the universe of S&P 100 stocks, our system results in superior out-of-sample volatility forecasts compared to existing risk models across forecast horizons. We further demonstrate that our system remains robust to different specifications and is scalable to a broader S&P 500 stock universe via hyperparameter transfer learning. Finally, the statistical improvement in volatility forecasts translates into an enhanced annual return around 8.5% from a cross-sectional variance risk premium strategy.
Keywords: Automation, Machine Learning, Volatility Forecasting, High-Frequency Data, Transfer Learning
JEL Classification: C13, C14, C52, C53, C55, C58
Suggested Citation: Suggested Citation