Multivariate range Value-at-Risk and covariance risk measures for elliptical and log-elliptical distributions

TL;DR

This paper introduces multivariate range Value-at-Risk and covariance measures tailored for elliptical distributions, providing explicit formulas, properties, and applications in portfolio optimization with real stock data.

Contribution

It proposes new multivariate range risk measures, derives explicit formulas for elliptical distributions, and applies them to portfolio selection with empirical data.

Findings

Range-based risk measures have desirable properties for risk management.

Explicit formulas are derived for common elliptical distributions.

Range-based efficient frontiers align with theoretical expectations.

Abstract

In this paper, we propose the multivariate range Value-at-Risk (MRVaR) and the multivariate range covariance (MRCov) as two risk measures and explore their desirable properties in risk management. In particular, we explain that such range-based risk measures are appropriate for risk management of regulation and investment purposes. The multivariate range correlation matrix (MRCorr) is introduced accordingly. To facilitate analytical analyses, we derive explicit expressions of the MRVaR and the MRCov in the context of the multivariate (log-)elliptical distribution family. Frequently-used cases in industry, such as normal, student-$t$, logistic, Laplace, and Pearson type VII distributions, are presented with numerical examples. As an application, we propose a range-based mean-variance framework of optimal portfolio selection. We calculate the range-based efficient frontiers of the optimal portfolios based on real data of stocks' returns. Both the numerical examples and the efficient frontiers demonstrate consistences with the desirable properties of the range-based risk measures.