This project seeks to analyse systemic risk from a network perspective, combining theoretical modelling, empirical analysis, and practical (policy) applications. First, the project aims at contributing to the debate on fundamental questions such as: Is there a resilient architecture to the financial system? Should we put restrictions on institutions that are too big or too connected to fail? Is there a tension between individual banks’ incentives and social welfare? Second, the project will deliver a very practical output by incorporating network effects and positive feedback in policy makers’ day-to-day stress-test methods. The complexity of financial interlinkages calls for a more comprehensive approach:
a) Theoretical modelling, focusing mainly on further development of analytical models to define and measure systemic risk in financial networks. For instance, we aim at a better understanding of risk in financial systems by including the mutual interrelation of systemic and individual risk in complex network systems. More specifically, we will contribute to the methodological enhancement of the analytical tool DebtRank (which was developed by the group and currently used by various central banks) to include further levels of interconnections and study variants that respond to specific features of the financial system, e.g., estimation of network dependencies with limited information available.
b) Empirical analysis, including ad hoc data validation that entails analysis of various large-scale databases, namely Bankscope (covering 31’000 banks or financial institutions), Orbis (130’000’000 enterprises), and Bloomberg (real time financial and macroeconomic data). Extensive work on data collection to validate the models will include a secondment at Bloomberg and/or Olsen, where in-depth data analysis will be carried out.
We expect to answer several questions such as (i) which network architectures are resilient to endogenous and exogenous shocks; (ii) which incentive structure led the system into the current architecture; (iii) how can we anticipate systemic distress by means of simple indicators and stress-tests; (iv) how can we make the financial system more resilient, and what policies could modify incentives in such a direction. Answering these questions is particularly useful for financial supervisory bodies to further improve macro-prudential regulation. Last but not least, the project will disseminate its research findings through academic journals and communicate relevant new insights directly to all stakeholders, including practitioners, policymakers, and regulators through direct interaction (personal meetings, workshops etc.)
Early Stage Resercher working on the project: Chiara Perillo
Supervisor: Professor Stefano Battiston, University of Zurich stefano.battiston(at)bf.uzh.ch