Title: Modelling over-dispersion in price jumps arrivals: A comparison between Poisson mixtures and linear Hawkes model
Authors: Ping Chen Tsai - Southern Taiwan University of Science and Technology (Taiwan) [presenting]
Abstract: Price jumps may display some degree of clustering and this feature is typically manifested through a phenomenon known as over-dispersion in count data. The time series of counts will show a variance larger than its mean, and hence rejects a Poisson null hypothesis. Two competing models, namely, mixtures of Poissons and linear Hawkes process, are shown to be able to reproduce the over-dispersion feature, but the former by definition has a clustering parameter equal to zero, whereas the latter has a strictly positive clustering parameter. Different versions of the two models are fitted to price jumps data and the estimation results compared. Special attention is paid to the base intensity of linear Hawkes process, and whether one of the components in mixtures of Poisson corresponds to this base intensity. The issue of many zeros in the count data is also addressed using a zero-inflated model.