Credit networks and systemic risk of Chinese local financing platforms: Too central or too big to fail? ¬リニ: ¬タモbased on different credit correlations using hierarchical methods

The accelerating accumulation and risk concentration of Chinese local financing platforms debts have attracted wide attention throughout the world. Due to the network of financial exposures among institutions, the failure of several platforms or regions of systemic importance will probably trigger systemic risk and destabilize the financial system. However, the complex network of credit relationships in Chinese local financing platforms at the state level remains unknown. To fill this gap, we presented the first complex networks and hierarchical cluster analysis of the credit market of Chinese local financing platforms using the ‘‘bottom up’’ method from firm-level data. Based on balance-sheet channel, we analyzed the topology and taxonomy by applying the analysis paradigm of subdominant ultra-metric space to an empirical data in 2013. It is remarked that we chose to extract the network of co-financed financing platforms in order to evaluate the effect of risk contagion from platforms to bank system. We used the new credit similarity measure by combining the factor of connectivity and size, to extract minimal spanning trees (MSTs) and hierarchical trees (HTs). We found that: (1) the degree distributions of credit correlation backbone structure of Chinese local financing platforms are fat tailed, and the structure is unstable with respect to targeted failures; (2) the backbone is highly hierarchical, and largely explained by the geographic region; (3) the credit correlation backbone structure based on connectivity and size is significantly heterogeneous; (4) key platforms and regions of systemic importance, and contagion path of systemic risk are obtained, which are contributed to preventing systemic risk and regional risk of Chinese local financing platforms and preserving financial stability under the framework of macro ✩ This work was supported by the National Natural Science Foundation of China (Grant No. 71473179) and Peking University-Lincoln Institute Fund (Grant No. DS10-20150901-CX). ∗ Corresponding author at: 1211 Tongji Building A, Siping Road 1500, Shanghai 200092, China. E-mail address: [email protected] (X. Chen). http://dx.doi.org/10.1016/j.physa.2016.05.032 0378-4371/© 2016 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). F. He, X. Chen / Physica A 461 (2016) 158–170 159 prudential supervision. Our approach of credit similarity measure provides a means of recognizing ‘‘systemically important’’ institutions and regions for a targeted policy with risk minimization which gives a flexible and comprehensive consideration to both aspects of ‘‘too big to fail’’ and ‘‘too central to fail’’. © 2016 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).