Quantile-based GARCH-MIDAS: Estimating value-at-risk using mixed-frequency information

被引：12

作者：

Xu, Yan ^{[1
]}

Wang, Xinyu ^{[1
]}

Liu, Hening ^{[2
]}

机构：

[1] China Univ Min & Technol, Sch Econ & Management, 1 Daxue Rd, Xuzhou 221116, Peoples R China

[2] Univ Manchester, Alliance Manchester Business Sch, Accounting & Finance Grp, Booth St East, Manchester M15 6PB, Lancs, England

来源：

FINANCE RESEARCH LETTERS | 2021年 / 43卷

基金：

中国国家自然科学基金;

关键词：

Quantile regression; GARCH-MIDAS; Value-at-risk forecast; Error bootstrapping method; MARKET VOLATILITY; REGRESSION; UNCERTAINTY;

D O I：

10.1016/j.frl.2021.101965

中图分类号：

F8 [财政、金融];

学科分类号：

0202 ;

摘要：

Utilizing mixed-frequency data to predict value-at-risk of portfolio returns is promising. Inspired by the GARCH-MIDAS model (Engle et al., 2013), we propose a novel quantile-based GARCHMIDAS model to explain how low-frequency covariates affect the quantile of high-frequency variables, being also an extension of CAViaR (Engle and Manganelli, 2004). We examine the impact of monthly economic policy uncertainty on the daily value-at-risk in the West Texas Intermediate crude oil spot and futures markets from 2000 to 2019 and find that the rise in economic policy uncertainty does drive greater WTI crude oil market risk, and vice versa.

引用

页数：9