Short-Term Risk and Adapting Covariance Models to Current Market Conditions
9 Pages Posted: 26 Sep 2014 Last revised: 8 Apr 2021
Date Written: September 3, 2015
Covariance models of stock returns appear throughout the investment process, e.g., forecasting portfolio risk, hedging, constructing mean-variance optimal portfolios, and algorithmic trading. Typically built from historic time-series, they estimate the past but–because markets and regimes continually change–not the present and future.
There are non-time-series, instantaneous ways to infer or predict volatility, e.g., option implied volatility, intra-period trading range, machine learning on alternative data. This paper describes a conceptual framework and algorithm for incorporating these inferences into any linear factor covariance model so that it represents one's beliefs about future behavior instead of echoing the past.
Keywords: Covariance, Factor Models, Nowcasting, Projection Algorithms
JEL Classification: C00, C11, C53, G19
Suggested Citation: Suggested Citation