A0153
Title: Bayesian analysis of multiple networks for financial risk management
Authors: Mike So - The Hong Kong University of Science and Technology (Hong Kong) [presenting]
Abstract: A network is a common tool for studying relationships or links among a set of nodes, like individuals or companies. The relationships can change over time and consist of interactions in multiple types of links or views. This kind of multi-view data can be observed in financial markets where the nodes in networks can be listed companies in a stock market and the links among the nodes represent various co-movement of returns and other financial random variables. To analyze the relationship among listed companies or stocks, we propose a model using observed covariates to explain interactions in different links. We also introduce a latent space to display the unobserved relationship on a plot and provide insights into financial risk management.