VaR BASED RISK MANAGEMENT

Abstract

In this paper we discuss the Value-at-Risk concept and we analyse the market risk by using EWMA approach. EWMA (exponentially weighted moving average) forecasting technique is a popular measure of various risks in financial risk management. We will compare standard EWMA, robust EWMA and skewed EWMA forecast of VaR. JP Morgan standard EWMA is derived from Gaussian distribution. Robust EWMA is based on Laplace distribution and skewed EWMA is a new approach derived from an asymmetric Laplace distribution. Asymmetric Laplace distribution takes into account both skewness and heavy tails in return distribution and the time varying nature of them in practice. Skewed EWMA VaR is a generalization of the standard EWMA method. Using these approaches we will analyse selected financial series (three European market indexes and one exchange rate). We have found and confirmed that skewed EWMA forecasting of VaR outperforms the standard EWMA method.

Estimating Value-at-Risk Based on Non-Normal Distributions

“…According to Tsay (2010) and Bohdalová & Greguš (2013), VaR is simply the (1 − p)-th quantile of the CDF of the loss function L(h). We note that CDF F h (x) is the focus of the econometric modelling.…”