Discovering SIFIs in Interbank Communities

Nicolo' Pecora, Alessandro Spelta, Pablo Rovira Kaltwasser

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)


This paper proposes a new methodology based on non-negative matrix factorization to detect communities and to identify central nodes in a network as well as within communities. The method is specifically designed for directed weighted networks and, consequently, it has been applied to the interbank network derived from the e-MID interbank market. In an interbank network indeed links are directed, representing flows of funds between lenders and borrowers. Besides distinguishing between Systemically Important Borrowers and Lenders, the technique complements the detection of systemically important banks, revealing the community structure of the network, that proxies the most plausible areas of contagion of institutions’ distress.
Original languageEnglish
Pages (from-to)N/A/-N/A
JournalPLoS One
Publication statusPublished - 2016


  • Community detection, non-negative matrix factorization, financial networks


Dive into the research topics of 'Discovering SIFIs in Interbank Communities'. Together they form a unique fingerprint.

Cite this