Open Access
American Research Journal of Business and Management
ISSN (Online): 2379-1047
DOI: 10.46568/arjbm
A New Robust Estımator for Value at Rısk
Abstract
We consider a common risk measuring method namely Value-at-Risk (VaR). The easiest and the most
prevalent method of calculating VaR is the variance-covariance method. This method is based on normal
distribution assumption. However, there are a lot of inferences in literature that non-normal distributions are
much more common than the normal distribution. Because of economic growth and political and financial issues,
there can be possible higher or lower prices than normal ones in economic data, which are named outlier in
statistic theory. In order to handle these data anomalies and distribution differences, robust estimation and testing
methods have been determined and studied for last decades. In this study, we propose a new robust variance
covariance estimator for calculating VaR value of a given portfolio. Simulation results show that the proposed
estimator is more robust than the corresponding normal theory solutions. Also, a real data for different
economical markets are analyzed to show the performances of the proposed estimators.
