Bayesian forecasting and scalable multivariate volatility analysis using simultaneous graphical dynamic models

TL;DR

This paper enhances scalable Bayesian multivariate volatility analysis using SGDLMs with adaptive predictor selection, demonstrating improved short-term financial forecasting and a new market stress indicator, supported by GPU acceleration.

Contribution

It introduces an adaptive predictor selection method within SGDLMs for on-line Bayesian forecasting, improving scalability and accuracy in high-dimensional financial data.

Findings

SGDLM forecasts volatility and co-volatility effectively

The model provides a leading indicator of financial market stress

GPU implementation significantly speeds up model fitting

Abstract

The recently introduced class of simultaneous graphical dynamic linear models (SGDLMs) defines an ability to scale on-line Bayesian analysis and forecasting to higher-dimensional time series. This paper advances the methodology of SGDLMs, developing and embedding a novel, adaptive method of simultaneous predictor selection in forward filtering for on-line learning and forecasting. The advances include developments in Bayesian computation for scalability, and a case study in exploring the resulting potential for improved short-term forecasting of large-scale volatility matrices. A case study concerns financial forecasting and portfolio optimization with a 400-dimensional series of daily stock prices. Analysis shows that the SGDLM forecasts volatilities and co-volatilities well, making it ideally suited to contributing to quantitative investment strategies to improve portfolio returns. We also identify performance metrics linked to the sequential Bayesian filtering analysis that turn out to define a leading indicator of increased financial market stresses, comparable to but leading the standard St. Louis Fed Financial Stress Index (STLFSI) measure. Parallel computation using GPU implementations substantially advance the ability to fit and use these models.