TENET: Tail-Event driven NETwork risk

Abstract

This is the accepted version of the paper.This version of the publication may differ from the final published version. Abstract CoVaR is a measure for systemic risk of the networked financial system conditional on institutions being under distress. The analysis of systemic risk is the focus of recent econometric analyzes and uses tail event and network based techniques. Here, in this paper we bring tail event and network dynamics together into one context. In order to pursue such joint efforts, we propose a semiparametric measure to estimate systemic interconnectedness across financial institutions based on tail-driven spillover effects in a high dimensional framework. The systemically important institutions are identified conditional to their interconnectedness structure. Methodologically, a variable selection technique in a time series setting is applied in the context of a single-index model for a generalized quantile regression framework. We could thus include more financial institutions into the analysis to measure their tail event interdependencies and, at the same time, be sensitive to non-linear relationships between them. Network analysis, its behaviour and dynamics, allows us to characterize the role of each financial industry group in 2007 -2012: the depositories received and transmitted more risk among other groups, the insurers were less affected by the financial crisis. The proposed TENET -Tail Event driven NETwork technique allows us to rank the Systemic Risk Receivers and Systemic Risk Emitters in the U.S. financial market. Permanent repository link:

Time Varying Quantile Lasso

“…It has uncovered the cross-sectional dependencies among financial institutions to be important when determining the risk on the market. Adrian and Brunnermeier (2016), Hautsch et al (2015) and Härdle et al (2016), just to mention a few, dealt with evaluating systemic risk according to the relevance of each financial institution itself. This inspired us to connect the Lasso parameter λ with the systemic risk, since it depends not only on the volatility but also on the size of model parameters and the correlation structure of the design matrix.…”

“…The Standard & Poor's 500 stock market index (S&P500) moves in opposite direction of λ, which can also provide some information about behaviour of λ in connection to the situation on financial markets. Another systemic risk measure is CoVaR presented by Adrian and Brunnermeier (2016) and extended by Härdle et al (2016), where a single index model for generalized quantile regression instead of linear quantile regression was employed.…”

“…The dynamic behaviour of the parameter λ can be observed when the model is assumed to vary over time and therefore the fitting is performed with the use of moving windows. The proposal of investigating time series of λ and its dependency on model characteristics was brought into focus by Härdle et al (2016), which was a foundation of FinancialRiskMeter (http://frm.wiwi.hu-berlin.de). Following the ideas behind the two aforementioned projects, we use the derivation of the formula for the penalization parameter λ as a result of the optimization problem.…”

Time Varying Quantile Lasso

“…Lasso in quantile regression has been used by Härdle et al (2016) to model tail event dependencies among U.S. financial companies. Based on the penalization parameters the FinancialRiskMeter (FRM), http://frm.wiwi.hu-berlin.de, was developed, see Figure 1.…”

Industry Interdependency Dynamics in a Network Context

“…Depending on their network construction, they study the systemic risk of over 1000 exchange-traded banks in the global network framework and argued that connectivity hasn't been paid enough emphasis in Financial Stability Board. Härdle et al (2016) propose a nonlinear semiparametric quantile regression method on CoVaR to construct a tail-event driven network in order to study the systemic risk among different financial sectors and conclude that the interconnectedness is growing during financial crisis period with largest systemic risk receivers and emitters being the most systemically 3 important.…”