Title: A self-exciting model of mutual fund flows
Authors: Ran Sun - University of Paris-Dauphine (France) [presenting]
Abstract: Dynamics of mutual funds flows are studied with a unique private database provided by several french asset management firms. This database covers a large range of funds with different characteristics. Compared to the other public available databases of fund flows, our database gives more information which allow us to understand better the mutual funds clients investing decision making. One important characteristic is the frequency of fund flows. We use a discrete counting process to model this flow frequency. Some stylized facts of the data are discovered; among them the self exciting property is a substantial one. In line with literature, this self exciting property indicates the correlation nature of funding liquidity. In a financial point of view, we give an analysis of client risk of mutual funds; in a statistical point of view, our model estimates more accurately tail risk of fund flows.