Dynamic copula models for multivariate high frequency data in Finance 2003
Alexandra Dias and Paul Embrechts
The stylized facts of univariate high-frequency data in finance are
well known; see Dacorogna et al. (2001). In Breymann et al. (2003)
we analyzed bivariate high frequency forex data as a function of the
sampling frequency, however treating the data as iid. In the present
paper, using the data from Breymann et al. (2003), we model the
dynamics as GARCH type processes and investigate the stylized facts
of the bivariate residuals. As a function of the sampling frequency, we
test for tail-dependence and ellipticity. We also investigate clustering
of extremes and change-points.
Kategorie:
Wydawnictwo:
Department of Mathematics ETH Zurich
Język:
english
Strony:
42
Plik:
PDF, 2.10 MB
IPFS:
,
english0