Systemic risk, financial markets, and performance of financial institutions

Edward M. H. Lin, Edward W. Sun, Min-Teh Yu

Research output: Contribution to journalArticlepeer-review

23 Scopus citations


This paper studies the exposure and contribution of financial institutions to systemic risks in financial markets. We employ three popular indicators of a financial institution's exposure to systemic risks: the systemic risk index (SRISK) and marginal expected shortfall (MES) of Brownlees and Engle (Volatility, correlation and tails for systemic risk measurement, Social Science Research Network, Rochester, NY, 2012) and the conditional Value-at-Risk (CoVaR) of Adrian and Brunnermeier (2011). We use a primary database of Taiwan financial institutions for our empirical study. A panel contains data of stock market returns and balance sheets of 31 Taiwan financial institutions for 2005-2014. We focus on systemic risk analysis so as to understand the dynamics of volatility, interdependency, and risk during the recent financial crisis. We then report the time series dynamics and cross sectional rankings of these systemic risk measures. The main results indicate that although these three measures differ in their definition of the contributions to systemic risk, all are quite similar in identifying systemically important financial institutions (SIFIs). Moreover, we find empirical evidence that systemic risk contributions are closely related to certain institution characteristic factors. The results of the Granger causality tests prove that a systemic risk measure is a great alternative tool for monitoring early warning signals of distress in the real economy.
Original languageEnglish
Article numberSI
Pages (from-to)579-603
Number of pages24
JournalAnnals of Operations Research
Issue numberSI
StatePublished - Mar 2018


  • Systemic risk; MES; SRISK; CoVaR; Financial crisis

Fingerprint Dive into the research topics of 'Systemic risk, financial markets, and performance of financial institutions'. Together they form a unique fingerprint.

Cite this