Financial networks could become fragile during periods of economic and financial distress, since interconnectedness among participating firms could transmit and amplify adverse shocks. Relying on balance sheet data, complemented with information on interbank exposures, this paper analyses interconnectedness in an advanced emerging market economy, using two complementary approaches. The first approach focuses on the financial network topology, and finds that the financial system resembles a highly clustered small world network, with in and out-degrees, connectivity and exposures exhibiting a (double) heavy tail behaviour, which favours the formation of strong community structure and preferential attachment in the network. Bank size is Pareto distributed, and highly correlated with centrality. The second approach focuses on how the network topology contributes to the transmission of shocks by modelling default contagion, using balance sheet network analysis. It finds that direct counterparty credit exposure poses less risk to the banking system than fire sale losses triggered by liquidity shocks. Both approaches, either using a topological or induced system losses perspective, identify systemically important financial institutions (SIFIs) consistently by accounting for both macro and microprudential risk dimensions.
Read full abstract