A dynamic factor semiparametric model for VaR and expected shortfall driven by realized measures

TL;DR

This paper introduces a semiparametric joint VaRES model driven by realized measures, capturing dynamic tail risk and outperforming traditional models in risk prediction tasks.

Contribution

It develops a novel dynamic factor semiparametric model that incorporates high-frequency realized measures to better capture tail risk dynamics.

Findings

Model outperforms benchmarks across multiple loss functions.

Effectively captures time-varying tail severity.

Separates risk level changes from tail risk intensification.

Abstract

This paper proposes a semiparametric joint VaRES framework driven by realized information, mo tivated by the economic mechanisms underlying tail risk generation. Building on the CAViaR quantile recursion, the model introduces a dynamic ESVaR gap to capture time-varying tail sever ity, while measurement equations transform multiple realized measures into high-frequency risk innovations.These innovations are further aggregated through a dynamic factor model, extracting common high-frequency tail risk factors that affect the quantile level and tail thickness through dis tinct risk channels. This structure explicitly separates changes in risk levels from the intensification of tail risk.Empirical evidence shows that the proposed model consistently outperforms quantile regression, EVT-based, and GARCH-type benchmarks across multiple loss functions, highlighting the importance of embedding high-frequency information directly into the tail risk generation layer