Bespoke Realized Volatility: Tailored Measures of Risk for Volatility Prediction
Date/Time: Wednesday 10 May, 2 pm -3 pm
Location: 110 Finance Decision Lab, 4 Eastern Rd / Online via Zoom
Abstract: Standard realized volatility (RV) measures estimate the latent volatility of an asset price using high frequency data with no reference to how or where the estimate will subsequently be used. This paper presents methods for “tailoring” the estimate of volatilityto the application in which it will be used. For example, if the volatility measure willbe used in a specific parametric forecasting model, it may be possible to exploit that information and construct a better measure of volatility. We use methods from machine learning to estimate optimal “bespoke” RVs for heterogeneous autoregressive (HAR) andGARCH-X forecasting applications. We apply the methods to 886 U.S. stock returns and find that bespoke RVs significantly improve out-of-sample forecast performance. We find that the bespoke RV places more weight on data from the end of the trade day, and thatthe resulting volatility forecasts are more responsive to news than benchmark forecasts.