Reconstructing and Stress Testing Credit Networks

44 Pages Posted: 12 Dec 2017

See all articles by Amanah Ramadiah

Amanah Ramadiah

University College London - Financial Computing and Analytics Group, Department of Computer Science

Fabio Caccioli

University College London

Daniel Fricke

Deutsche Bundesbank; University College London; London School of Economics & Political Science (LSE) - Systemic Risk Centre

Date Written: December 8, 2017

Abstract

Financial networks are an important source of systemic risk, but often only partial network information is available. In this paper, we use data on bank-firm credit relationships in Japan and conduct a horse race between different network reconstruction methods in terms of their ability to reproduce the actual credit networks. We then compare the different reconstruction methods in terms of their implied systemic risk levels. In most instances we find that the observed credit network significantly displays the highest systemic risk level. Lastly, we explore different policies to improve the robustness of the system.

Keywords: network reconstruction; systemic risk; bipartite credit network; aggregation level

Suggested Citation

Ramadiah, Amanah and Caccioli, Fabio and Fricke, Daniel, Reconstructing and Stress Testing Credit Networks (December 8, 2017). Available at SSRN: https://ssrn.com/abstract=3084543 or http://dx.doi.org/10.2139/ssrn.3084543

Amanah Ramadiah

University College London - Financial Computing and Analytics Group, Department of Computer Science ( email )

Gower Street
London, WC1E 6BT
United Kingdom

Fabio Caccioli

University College London ( email )

Gower Street
London, WC1E 6BT
United Kingdom

Daniel Fricke (Contact Author)

Deutsche Bundesbank ( email )

Wilhelm-Epstein-Str. 14
Frankfurt/Main, 60431
Germany

University College London ( email )

Gower Street
London, WC1E 6BT
United Kingdom

London School of Economics & Political Science (LSE) - Systemic Risk Centre

Houghton St, London WC2A 2AE, United Kingdom
London

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
114
Abstract Views
755
rank
277,308
PlumX Metrics