| Summary: | Given the weaknesses of the parametric VaR (Value-at-Risk) calculated by normality assumptions, this paper develops a method of parametric VaR calculation considering ten different probability distributions. Specifically, the distribution to be used for the VaR calculation of a specific asset or portfolio is indicated by the Kolmogorov-Smirnov goodness-of-fit test. Additionally, the study compares the normality assumptions applicability for the VaR calculation of both individual assets, and to a large portfolio, in the context of market stability. The experiment makes use of a sample of 15 individual assets traded in the Sao Paulo Stock Exchange and the IBOVESPA index, collected in the Economática® database. The goodness-of-fit tests and VaR calculations are performed by a program developed in MATLAB7.1®. This investigation demonstrates that the assumption of normality brings good risk estimates for large portfolios and individual assets.
|
|---|