Comparing the generalised hyperbolic and the normal inverse Gaussian distributions for the daily returns of the PSI20
The presence of non-normality, fat-tails, skewness and kurtosis in the distribution of the returns necessitates the fitting of distributions that account for this phenomenon. We fit the Generalized Hyperbolic Distribution and the Normal Inverse Gaussian to the daily returns from the Portuguese Stock...
| Autor principal: | |
|---|---|
| Outros Autores: | |
| Formato: | workingPaper |
| Idioma: | eng |
| Publicado em: |
2019
|
| Assuntos: | |
| Texto completo: | http://hdl.handle.net/10400.3/5047 |
| País: | Portugal |
| Oai: | oai:repositorio.uac.pt:10400.3/5047 |
| Resumo: | The presence of non-normality, fat-tails, skewness and kurtosis in the distribution of the returns necessitates the fitting of distributions that account for this phenomenon. We fit the Generalized Hyperbolic Distribution and the Normal Inverse Gaussian to the daily returns from the Portuguese Stock Index, the PSI20. We use the EM algorithm for estimating the parameters of the Normal Inverse Gaussian while those of the Generalized Hyperbolic distribution are estimated using the Nelder-Mead algorithm. We find that the Generalized Hyperbolic is a better fit than the Normal Inverse Gaussian as it better estimates the probabilities at the left tail where the losses are concentrated. |
|---|