Analyzing systemic risk in financial networks using network tools: for the case of financial break down of Turkish money market in 2000

Abstract

This thesis analyzes performance of several network centrality measures in assesing systemic risk relying on data from the Turkish money market during financial crisis 2000. In order to gain clearer picture of network topology of Turkish money market various network investigation tools such as volume, transactions, links, connectivity and reciprocity are employed. The main borrower role of Demirbank in the collapse of banking system has been studied with several centrality measures. It is shown that those measures can be used as strong signals much before crisis. In an ex-post analysis of crisis, it is shown that centrality measures perform well in detecting systemically important agent, which Demirbank in our case. Network topology of Turkish interbank market satisfies the scale free property which enables us to monitor response to the failure of the central agent. This thesis adds to the new and growing literature on network topological analysis of interbank money markets.