Activity day:2016-12-26 Published At：2016-12-26 Views:1783 2017-02-12 updated
本系擬於 12/30 (五) 邀請新加坡國立大學的段錦泉教授至本所進行專題演講，相關訊息如下：
時間: 12/30 (五) 10:30-12:00
講題: Banking Network and Systemic Risk via Forward‐Looking Partial Default Correlations
This paper studies systemic risk in a global network with over 1,000 exchange‐traded banks. Network construction follows a methodology comprising three parts: (1) the use of the default correlation model of Duan and Miao (2015) to produce a forward‐looking probability of default (PD) total correlation matrix and then transform it into a partial correlation matrix by applying the CONCORD algorithm; (2) the measurement of banks’ systemic importance hinging on six network centrality indicators based on the partial correlations, which represent the direct connections among banks; and (3) a graphical analysis of the global banking network which can then be partitioned into overlapping bank/group centric local communities. We then specifically study the banks’ systemic importance in 2008 and 2014. Using the 2014 sample, we are able to compare the systemic importance rankings under alternative measures, including G‐SIBs identified by the Financial Stability Board. Our results suggest the Board rankings appear biased towards singling out large institutions as systemic, with connectivity playing a rather minor role.