OFR Working Paper Shows Impact of Credit Default Swap Stress across the Banking System
Published: March 8, 2016
An OFR working paper published today uses data about the credit default swap (CDS) market to evaluate the impact on banks from default of their largest counterparties. It also takes a macroprudential perspective to consider not only the impacts on individual banks, but on the financial system as a whole.
The authors find that banks’ exposures in the credit default swap market under stress are concentrated in a small number of counterparties. Banks’ indirect exposures to the same counterparties also are potentially much larger than their direct exposures.
The authors — Jill Cetina, Mark Paddrik, and Sriram Rajan — use the same stress test scenarios that the Federal Reserve used for its stress tests in 2013, 2014, and 2015. The Federal Reserve’s test, known as the Comprehensive Capital Analysis and Review (CCAR), stresses the trading books of the largest U.S. bank holding companies (BHCs). In CCAR, banks must consider the default of the counterparty that would owe the bank the most money on credit default swaps during a stress event.
The CCAR stress scenario is relevant given events like those involving American International Group, Inc. (AIG) during the financial crisis. AIG had written hundreds of billions in CDS protection to banks in the United States and Europe. AIG’s failure could have caused material losses for these firms.
The OFR working paper highlights two potential limitations to the approach of considering only the largest counterparty.
First, CCAR considers BHCs’ direct counterparty concentrations, not indirect ones. Access to data on the full U.S. CDS market allowed the authors to stress CDS positions for both BHCs and non-BHCs. The authors compared the direct impact of the default of a BHC’s largest counterparty with the impact of indirect losses of the largest counterparty on the BHC’s other counterparties.
The authors find that indirect effects can be as much as nine times larger than the direct impact on the BHC. As a result, ignoring indirect effects could understate the stress on banks. The authors also find instances when the BHC’s largest counterparty would not be the source of the largest indirect effects.
Second, the authors compared the risks that BHCs face individually to what they face as a group. The authors used a concentration index that quantifies how much each counterparty would owe to the banking system as a whole under the 2015 CCAR stress scenario. They find that those payoffs would be highly concentrated. The index reading is similar to a market exposed to just three counterparties.
The authors also note some changes in the U.S. CDS market. Between 2013 and 2015, BHCs have moved from being net sellers of protection to net buyers. This change suggests a shift of risk from the banking sector to nonbanks. The authors also show that the concentration of banks’ counterparty exposures has increased.
The working paper informs the evaluation of stress tests. By using granular contractual information, OFR researchers provide new insights on economic loss under financial distress. By considering the banking system as a whole, they highlight a macroprudential perspective on stress testing. By analyzing the full financial network, they identify areas where losses may be large and systemic risk concerns may develop.
Stacey Schreft is Deputy Director for Research and Analysis at the Office of Financial Research