16 October 2013

ECB Working Paper: Disentangling the bond-CDS nexus - A stress test model of the CDS market

The model enables the analysis of credit risk transfer mechanisms, includes features of market and liquidity risk, and allows for contagious propagation of counterparty failures. One contribution of the model is that it explicitly incorporates several features proper to OTC derivatives markets, including collateralization, collateral netting agreements and close-out netting procedures in case of counterparty default. It also provides a modelling framework to assess the potentially risk-mitigating or risk-amplifying role of the CDS market in case of a credit event.

The model includes several channels through which an exogenous credit event may affect banks, focusing on the interplay between their bond and CDS holdings. Five transmission channels to banks are featured: (i) direct losses on bond holdings, (ii) write-downs on other (available for sale and held for trading) bond exposures, (iii) direct CDS repayments triggered by the simulated credit event, (iv) increased collateral requirements to cope with higher CDS spreads on other non-defaulted reference entities, and (v) contagious propagation of counterparty failures either through insolvency or illiquidity.

To provide an example with publicly available data, the model is calibrated using balance sheet data released by the European Banking Authority (EBA) on 65 major European banks related to the EU 2011 Capital Exercise. The dataset includes both sovereign bond and CDS holdings at the bank level for 28 European sovereign entities, while banks’ bilateral CDS exposures are estimated and their market values simulated. To analyse the relative magnitude of each transmission channel to banks, we simulate exogenous credit event scenarios for a wide range of recovery rates.

According to the simulation results, banks’ losses due to bond exposures appear to be significantly more important in magnitude than losses due to pure CDS exposures and to counterparty risk on the CDS market. The authors do not find significant failures due to the inability of some banks to honour their CDS repayments. Overall, CDS repayments remain at low levels compared to banks’ liquid asset pools and to capital ratios. In this regard, the usual focus - at least in the financial press - on the large (gross) amounts at stake on the CDS market might be misleading, as it occults another more important source of fragility. According to our results, the largest source of vulnerability for the CDS protection sellers is found to be the sudden increases in collateral to be posted, i.e. the inability of financial institutions to meet collateral calls. Paradoxically, whereas collateral posting and variation margins are counterparty risk mitigation mechanisms, they can turn out to be major drivers of counterparty failures at times of elevated financial stress, i.e. when collateral has to be delivered on multiple positions at the same time.

As regards to contagion, ECB does  not find evidence for significant contagion purely due to failures of counterparties on the CDS market. Potential explanations include the effectiveness of collateral management schemes, the fact that none of the major dealers is found to fail, and that several types of interconnections between banks (interbank loans and deposits, other derivatives) are not accounted for in the model. Moreover, close-out netting of the whole CDS portfolio in case of counterparty failure is shown to play a major risk-mitigating role, as contagion would affect most of the banks active on the CDS market if it were not to be implemented.

Finally, we are not able to document redistributive effects of net CDS repayments in case of a simulated credit event, neither from banks with low exposure to highly-exposed banks nor from highly-liquid banks to banks with lower liquidity.

Full Working paper

© ECB - European Central Bank