Forward-looking solvency contagion

Abstract

Solvency contagion risk is a key channel through which systemic risk can come about. We introduce a model that accounts not only for losses transmitted after banks default, but also for losses due to the fact that creditors revalue their exposures when probabilities of default of their counterparties change. We apply the model to run a series of simplified stress tests of the UK banking system from 2008 to 2016, based on two datasets of real interbank exposures between the seven major UK banks. We show that risks due to solvency contagion decrease markedly from the peak of the crisis, to the point of becoming negligible. We also characterise the distributions of both vulnerabilities and systemic importances of individual banks, thereby tracking the evolution of risk concentration.

Solvency Distress Contagion Risk: Network Structure, Bank Heterogeneity and Systemic Resilience

“…To systematically analyse the relationship between network structure, institutional characteristics and solvency contagion risk, we use the solvency contagion model introduced by Bardoscia et al (2019). We apply this model of solvency distress contagion to a wide range of simulated networks for both system-wide and idiosyncratic exogenous shocks, including networks calibrated to historical data for the UK interbank network.…”

“…To quantify solvency contagion risk under a set of exogenous shocks we use the model of Bardoscia et al (2019), which is deployed by the Bank of England in regulatory stress testing (see e.g. Bank of England ( 2016)).…”

“…We model solvency distress contagion risk using the model of Bardoscia et al (2019), a realistic model of the contagion mechanism that has been used in regulatory stress testing (see e.g. Bank of England (2016)).…”

“…While most of the existing papers focus on default contagion, more recent work has modelled solvency distress contagion (Battiston et al (2012)) -which recognises that the value of interbank assets can fall, propagating distress, before banks default. Barucca et al (2016) present a general framework for modelling such contagion in a network of interbank exposures; and Bardoscia et al (2019) adapt this approach to produce a model that captures the risks of solvency distress contagion in a network of interbank debt claims, using a structural model to revalue those claims. In this framework, counterparty losses increase non-linearly as solvency falls, more closely reflecting reality.…”

“…We contribute to the literature in several ways. First, we analyse the interaction between network structure, institutional characteristics and contagion risk for this more realistic solvency distress contagion mechanism, using the model of Bardoscia et al (2019)). Second, we systematically analyse how contagion varies both as a function of network connectedness -the density of the network -and as a function of network structurewhether a network has a regular, random, scale-free or core-periphery structure.…”