Effectiveness of copula-extreme value theory in estimating value-at-risk: empirical evidence from Asian emerging markets

Abstract

Traditional Monte Carlo simulation using linear correlations induces estimation bias in measuring portfolio value-at-risk (VaR) due to the well-documented existence of fat-tail, skewness, truncations, and non-linear relations of return distributions. In this paper, we evaluate the effectiveness using copula-extreme-value-based semi-parametric approaches in assessing portfolio risks in six Asian markets based on their different return distribution shapes. We incorporate extreme value theory (EVT) to model the tails of the return distributions and various copulas to build the joint distribution of returns. The backtesting analysis of the Monte Carlo VaR simulation suggests that the Clayton copula-EVT has the best performance regardless of the shapes of the return distributions, and in general the copulas with the EVT perform better estimation of VaRs than the traditional copulas. This concludes the economic significance in incorporating the down-side shock in risk management.JEL classifications: G15, F31, C46

Review of Recent Legacy Airline Mergers in the U.S.: An Empirical Study from Investors’ Perspectives

“…Thus, our data distributions became semi-parameter distributions with the uniform distribution (no parameter) to model the distribution center and the generalized Pareto distribution (with parameters) to model the distribution tails. As indicated by Hsu, Huang, and Chiou (2012), the above i.i.d. sequence, y i,t , can be described as follows:…”

“…where q i,t stands for the continuous change of stock trading volume of airline stock i on day t and w i,t and w i,t-1 represent the trading volume of airline stock i on date t and date (t-1), respectively. The GARCH-EVT-Copula model suggested by Hsu, Huang, and Chiou (2012) was adopted in this study. Copula models are commonly used in financial risk modeling to fit non-normal distributed data and have been used in studying the dependence between block delays and gate arrival delays at major U.S. airports (Diana, 2011).…”

Forecasting portfolio-Value-at-Risk with nonparametric lower tail dependence estimates

“…1 Thus yielding the standard copula-GARCH approach to estimating and forecasting multivariate portfolio VaR (Hsu et al, 2012;Nikoloulopoulos et al, 2011;Weiß, 2011).…”

A wavelet based approach to measure and manage contagion at different time scales

“…Additionally, triggered by the fact that linear correlation does not capture tail dependence and nonlinear transformations of the marginal return distribution, the application of copulas (introduced to financial time series by Embrechts et al [8]) has become a widely accepted tool to capture nonlinear dependences such as tail dependence between financial time series (see Refs. [1,[9][10][11]7]). …”