A Network-Based Spectral Decomposition Approach to Systemic Risk

Macro-prudential policy is concerned with mitigating system wide risks arising from negative externalities from monetary and financial activities. The need to design a tax scheme to control systemic risk from financial intermediaries (FIs) has come to the forefront after extensive tax payer bail outs in the aftermath of the 2007 financial crisis. We propose a network-based spectral decomposition approach called “eigen pair” that provides theoretical fundation of the stability of the financial system as well as a criteria to identify Systemically Important Financial Institutions (SIFIs), those contributing to the instability of the system. We show that by defining an appropriate expected cost function, the eigen pair approach can be used to monitor the stability of financial systems under specific macro-prudential regulations, such as the capital and liquidity pillars of Basel III. Finally, we proposed an ex-ante bail in tax schemes on SIFIs to interanlize systemic risk and mitigate ex-post tax payer bail outs.We argue that it is a mistake to focus on a preconceived list of systemically important FIs (SIFIs) and use ad hoc surcharges to stabilize the financial system. Governance issues surrounding "bail ins" are discussed.