We put forward a Merton-type multi-factor portfolio model for assessing banks' contributions to systemic risk. This model accounts for the major drivers of banks' systemic relevance: size, default risk and correlation of banks' assets as a proxy for interconnectedness. We measure systemic risk in terms of the portfolio expected shortfall (ES). Banks' (marginal) risk contributions are calculated based on partial derivatives of the ES in order to ensure a full risk allocation among institutions. We compare the performance of an importance sampling algorithm with a fast analytical approximation of the ES and the marginal risk contributions. Furthermore, we show empirically for a portfolio of large international banks how our approach could be implemented to compute bank-specific capital surcharges for systemic risk or stabilisation fees. We find that size alone is not a reliable proxy for the systemic importance of a bank in this framework. In order to smooth cyclical fluctuations of the risk measure, we explore a time-varying confidence level of the ES.Keywords: systemic risk contributions, systemic capital charge, expected shortfall, importance sampling, granularity adjustment JEL Classification: C15, C63, E58, G01, G21 Non-technical summaryIn the aftermath of the recent financial turmoil, part of the debate on regulatory reforms has focused on the question how the contribution of individual financial firms to systemic risk can be addressed. The final goal of these considerations is to internalise the negative externalities imposed on society in a financial crisis. Two different policy tools that have been put forward for this purpose are a regulatory capital surcharge based on an institution's contribution to systemic risk and a levy that is paid into a resolution / stabilisation fund. In this paper we contribute to the regulatory debate first by introducing a methodology to assess the systemic risk contributions of banks and second by an illustrative implementation a capital surcharge based thereupon.We measure the system-wide risk by means of a portfolio approach that is widely used in the financial industry to manage credit risk at the individual firm level. A key advantage of this approach is that it does not only account for the size and an estimate of the financial soundness of the respective institution but also captures interlinkages between financial firms. We use stock market information in order to gauge market participants' collective evaluation of these difficult to quantify interlinkages which are reflected in the correlation structure of risk factors in a multi-factor model of firms' asset returns. Our methodology can be extended to private firms if proxy variables are used to replace the market price dependent information.In order to assess the contribution of each individual institution to system-wide risk, we apply an allocation technique based on the institution's marginal risk contribution. An important distinguishing feature of this method is the full allocation property, which means...

No abstract

I introduce a novel, hierarchical model of tail dependent asset returns which can be particularly useful for measuring portfolio credit risk within the structural framework. To allow for a stronger dependence within sub-portfolios than between them, I utilise the concept of nested Archimedean copulas, but modify the nesting procedure to ensure the compatibility of copula generators by construction. This makes sampling straightforward. Moreover, I provide details on a particular specification based on a gamma mixture of powers. This model allows for lower tail dependence, resulting in a more conservative credit risk assessment than a comparable Gaussian model. I illustrate the extent of model risk when calculating VaR or Expected Shortfall for a credit portfolio.

M‐PRESS‐CreditRisk is a novel stress testing approach that can help authorities gauge banks' capital adequacy related to credit risk. For the first time, it combines the assessment of microprudential capital requirements under Pillars 1 and 2 and macroprudential buffers in a unified, coherent framework. Its core element is an advanced credit portfolio model—SystemicCreditRisk—built upon a rich, nonlinear dependence structure for correlated bank portfolios. The model is applied to a sample of 12 systemically important German banking groups and delivers measures for systemic credit risk and the banks' contributions to it in both baseline and stress scenarios.

No abstract