2014/12/04 by Stefano Gurciullo, Gurciullo, Stefano
Computer Science · Economics, Econometrics and Finance · Physics and Astronomy · #Banking stability, regulation, efficiency #Complex Systems and Time Series Analysis #Global Financial Crisis and Policies #cs.SI #msc:91B74 #msc:91B82 #msc:91G99 #physics.soc-ph #q-fin.GN #q-fin.RM
paper · pdf · doi:10.48550/arxiv.1412.1679
50 pages, Paper presented as part of the coursework for the PhD Transfer Viva
arxiv created 2014/12/04 · arxiv updated 2014/12/05
This work proposes an augmented variant of DebtRank with uncertainty intervals as a method to investigate and assess systemic risk in financial networks, in a context of incomplete data. The algorithm is tested against a default contagion algorithm on three ensembles of networks with increasing density, estimated from real-world banking data related to the largest 227 EU15 financial institutions indexed in a stock market. Results suggest that DebtRank is capable of capturing increasing rates of systemic risk in a more sensitive and continuous way, thereby acting as an early-warning signal. The paper proposes three policy instruments based on this approach: the monitoring of systemic risk over time by applying the augmented DebtRank on time snapshots of interbank networks, a stress-testing framework able to test the systemic importance of financial institutions on different shock scenarios, and the evaluation of distribution of systemic losses in currency value.