Title: Monitoring financial networks with online Hurdle models
Authors: Shawn Mankad - Cornell University (United States) [presenting]
Kamran Paynabar - Georgia Institute of Technology (United States)
Mostafa Reisi Gahrooei - Georgia Institute of Technology (United States)
Samaneh Ebrahimi - Georgia Institute of Technology (United States)
Abstract: Financial trading data can be used to assess the stability of the financial system. We address this important problem through a network modeling approach that incorporates the complex structure found within high-fidelity trading datastreams. For instance, activities within financial markets can occur in varying market conditions, at different prices, and be of different types and sizes. Furthermore, typically regulators can identify the parties involved in any transaction, allowing for integration with other data, such as information on each counter-party, market announcements, etc. Building on the financial networks literature, which has gained popularity to model these complex dynamics, we create a novel network monitoring system to detect changes within a sequence of sparse networks constructed from an interbank lending market in the European Union. Our approach combines a state space model with the Hurdle model to capture temporal dynamics of the edge formation process, which is modeled as a function of the node and edge attributes and estimated using an extended Kalman Filter. Statistical process control charts are used to monitor the network sequence in real time in order to identify changes in trading patterns that are caused by fundamental shifts in market conditions. We show that the proposed methodology would have raised alarms to the public prior to key events and announcements by the European Central Bank during the 2007-2009 financial crisis.