Modeling Heavy-Tailed Stock Index Returns Using the Generalized Hyperbolic Distribution

In the present study, we estimate the parameters of the Generalized Hyperbolic Distribution for a series of stock index returns including the Romanian BETC and indexes from other two Eastern European countries, Hungary and the Czech Republic. Using different econometric techniques, we investigate whether the estimated Generalized Hyperbolic Distribution is an appropriate approximation for the empirical distribution computed by non-parametric kernel econometric methods. The main finding of the analysis is that the probability density function of the estimated Generalized Hyperbolic Distribution represents a very close approximation (at least up to the 4th order term) of the empirical probability distribution function.

Keywords: Generalized Hyperbolic Distribution, heavy-tailed returns, non-parametric density estimation
JEL Classification: C13, C14, C16, G10