Abstract

JIMENEZ GOMEZ, Andrés Eduardo  and  MELO VELANDIA, Luis Fernando. Conditional Modeling of Skewness and Kurtosis on Colombian Financial Series. Desarro. soc. [online]. 2016, n.76, pp.273-321. ISSN 0120-3584.  https://doi.org/10.13043/DYS.76.7.

Traditional methodologies used to calculate the value at risk and conditional value at risk usually model the first and second moments of the series, assuming that the third and fourth moments are constant. This paper uses the methodology proposed by Hansen (1994) to model the first four moments of the series, in particular, several parametric shapes are used to model the skewness and kurtosis. The traditional measures of VaR, CVaR and proposals are calculated for the Representative Market Rate, TES, and the IGBC for the period between January 2008 and February 2014. Overall, it was found that measures of market risk have better performance when conditional skewness and kurtosis of the series is modeled.

Keywords : VaR; conditional value at risk; backtesting; skewness; kurtosis.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )