The Chinese stock market crash in June 2015 has demonstrated necessary to improve understanding of systemic risk from the perspective of financial network. This study constructs a tail risk network to present overall systemic risk of Chinese financial institutions, given the macroeconomic and market externalities. Employing the Least Absolute Shrinkage and Selection Operator (LASSO) method of high-dimensional models, our results show that firm’s idiosyncratic risk can be affected by its connectedness with other institutions. The risk spillover effect from other companies is the main driving factor of firm-specific risk, comparing with macroeconomic state, firm characteristics and historical price movement. The number of connections between institutions significantly increases during June 2014 to June 2016. Moreover, we utilize the Kolmogorov-Smirnov statistic to test significance of systemic risk beta based on tail risk and further rank the systemic risk contribution. Regulators could detect those firms that are most threatening to the stability of system.

Introduction

With the development of financial innovation and globalization over the past few years, the co-movement between financial institutions’ assets and credit exposure tends to increase, which gives rise to risk spillover through the networking of firms2 (Adrian and Brunnermeier, 2016). Since the outbreak of the global financial crisis in 2007, and the dramatic effects of the Lehman collapse in 2008, systemic risk has become a matter of great concern for policy makers and financial institutions. This crisis reminds us that systemic risk can arise through interconnections across the individual firms, and individual firm’s failure may have repercussions on the entire financial system (FSB/IMF/BIS, 2009; IOSCO, 2011). As the second largest market in the world, the Chinese financial system has drawn growing worldwide attention after a series of liberalization policies in China after 20103 .Many financial institutions are connected by their mutual asset holdings and forming of various financial networks. At present, the cooperation between financial institutions has result in tremendous unprecedented progress both in depth and breadth that also provides more possible risk contagion. Particularly in the Chinese stock market crash of 2015, thousands of A-shares hit either upward or downward price limits, accounting for about one third of the total market. Given this condition, firm-specific risk cannot be appropriately assessed in isolation without accounting for potential risk spillover effects from other firms (Hautsch et al., 2014). Thus, characterization of systemic risk across financial institutions in China is a key problem. Yet to the best of our knowledge, this issue has barely been touched upon in the literature. This study constructs a tail risk network to investigate the systemic risk across Chinese financial institutions. The firms selected are classified into categories of Commercial Bank, Brokerage, Insurance, and Other categories according to the classification catalogue proposed by the company of Shenyin &Wanguo (SW) Securities. The industry classification of SW Securities, one of the largest research institutions in China, has been unanimously recognized by investors and widely utilized in the Chinese market. From the sample period of January, 2010 to November, 2017, we measure systemic risk contribution using the conditional value-at-risk developed by Adrian and Brunnermeier (2016) when the incremental risk of one firm — that is conditional on another — experiences a crash. Considering the spillover effects, we employ the Least Absolute Shrinkage and Selection Operator (LASSO) method (Belloni and Chernozhukov, 2011), which is a statistical shrinkage technique and standard for high-dimensional models. The potential factors are macroeconomic variables, firm-specific characteristics, lagged return, and the influence of other firms. We further use the Kolmogorov-Smirnov (K-S) statistic extended by Abadie (2002) to test for the significance of systemic risk beta, a measurement based on tail risk and finally rank the systemic risk contributions.