Market-Informed Networks for Modeling and Forecast Evaluation of Financial Extremes

RESUMO

Modeling the joint distribution of extreme values in high-dimensional financial time series is challenging. We introduce a time-dependent network Huesler-Reiss model in which market-informed adjacency matrices determine how strongly observations contribute to the estimation, including the Joint Extremes Adjacency Matrix (JEAM). Covering one-minute stock returns from three sectors of the S&P 100, JEAM achieves the best out-of-sample log scores for both tail directions, improving scores by 11-15%.