Danger Zones for the Financial System

The global financial crisis which began in 2007 has shown how financial turbulences are difficult to detect in time, especially because of the complex and highly dynamic financial connections among institutions and markets. Financial and economic interconnections among sovereigns, banks and corporations, have led to the failure of traditional asset pricing models to assess and correctly price risks for most of financial assets. From the Greek crisis that started in April-May 2010 we have learned the importance of monitoring the feedbacks between sovereign debt and bank risks (IMF, 2010). In many countries such as Ireland the State’s intervention for bailing out the banking system has undermined fiscal sustainability, leading to higher public indebtedness and difficulties in sovereign funding. Similarly, the fall in the market value of sovereign debt has undermined the balance sheets of the banking system. Significant spillovers have also arisen between the banks and the corporate sector as described for example in Chen et al. (2010). Bank and corporate distress, especially when originating in systemically important economies, strongly affected real activity, leading to a deterioration of the quality of loans and aggravating the credit crunch. For example, Dailami (2009) has documented that sovereign default risks at times of market turmoil have significantly affected corporate bond spreads. In other terms, sovereign risk has translated into higher costs of capital for private corporate issuers and therefore directly negatively affected firms’ balance sheets. In such a new and complex economic and financial environment, academics, regulators and practitioners need to reconsider the nature of systemic risk, namely considering the financial system as a complex entity where sovereigns, banks, other financial intermediaries, and corporations are subject to sector-specific and systemic risks, and interact one another through spillover effects. Detecting systemic risk zones is of primary importance: the early detection and causal identification of such phenomena may enable timely intervention and prevent countries, banks, and corporations from moving down towards dangerous risk paths. Thus it is of great interest to provide effective early warnings and a set of leading indicators for signalling impending systemic risks. The objective of this chapter is to present and discuss some of the major results that we obtained on the issue of Early Warning System (EWS), their realization and implementation. We consider sovereign systemic risk, banking systemic risk, and hedge fund systemic risk. Here we try to offer an “helicopter view” of the data mining approach, the one we have chosen in our contributions. This is based on the idea of imposing a minimal amount of structure on the data and allowing uncovering as relevant information on systemic crises from the data. Indeed, in our view one of the main reason why the traditional asset pricing models have failed in detecting systemic risk is the underlying “data-modeling culture”, through which economists, elaborate a theory, and validate the theory through the data (Savona and Vezzoli, 2012). This approach is appropriate to compare and test different theories, but often fails to uncover complex interactions and nonlinear dynamics. The structure of this chapter is as follows. In Section II we discuss the methodologies used to stratify risks within the financial system. Section III presents some empirical model of EWS for sovereigns, banks, and hedge funds. Section IV concludes.