Using Kernel Function and Α-Stable Distribution for Determining Value at Risk (Var) for the Companies Included in WIG20 Index at Warsaw Stock Exchange

Abstract

One of the most common from among conducted studies related to capital market includes studies regarding methods intended for suitable description of distribution of rates of return. Familiarity with such distributions (density function) is often necessary condition for usability of other methods, techniques and models describing the respective elements of the capital market such as i.a. risk. One method of risk measurement is evaluation of rate of return variability, exactly based on familiarity of its distributions. From a theoretical point of view, adopting Gauss's axiom is not acceptable, despite its high practicability. This paper presents results of the study related to possible use of α-stable distributions and kernel function for modeling rates of return and using these methods for establishing value at risk (VaR).