Title: A model for jumps in volatility and volume
Authors: Eduardo Rossi - University of Pavia (Italy)
Paolo Santucci de Magistris - Aarhus University (Denmark) [presenting]
Abstract: A multivariate model is proposed for stock returns, volatility and trading volume that allows for idiosyncratic and common jump terms, where the latter is typically associated with important news arrivals, such as significant announcements and/or extreme economic events. In particular, the jump terms are modeled a the sum of random variables where the number elements in the sum is Poisson distributed with time-varying intensity. The interaction between the jump terms makes the model able to provide high flexibility in explaining the contemporaneous presence of jumps and the dynamic dependence between return, volatility and volume. The model is estimated by maximum likelihood and it exploits the information coming from realized volatility measures based on high frequency data. The empirical results testify the ability of the proposed model to fit most of the features that characterize the volatility and the volume relationship.