On Bank Credit Risk: Systemic or Bank Specific? Evidence for the United States and United Kingdom

Abstract

We develop a multivariate credit risk model that accounts for joint defaults of banks and allows us to disentangle how much of banks’ credit risk is systemic. We find that the United States and United Kingdom differ not only in the evolution of systemic risk but, in particular, in their banks’ systemic exposures. In both countries, however, systemic credit risk varies substantially, represents about half of total bank credit risk on average, and induces high risk premia. The results suggest that sovereign and bank systemic risk are particularly interlinked in the United Kingdom.

How Much of Bank Credit Risk Is Sovereign Risk? Evidence from Europe

“…A possible explanation is that their systematic intensity reflects the impact of the U.S. financial crisis on Europe more than the evolution of the tensions related to Europe sovereign debt crisis . More fundamentally, our estimated systematic‐sovereign intensity is also remarkably different from the systemic‐ bank intensity presented by Li and Zinna (), which is estimated separately on a panel of seven large U.S. and UK banks, over a similar time period (2008–13). In fact, their intensity reaches its peak during the 2008‐09 global financial crisis, and takes much lower values during the Europe sovereign debt crisis.…”

“…An alternative and more general specification to model default risk premiums is the extended market price of risk proposed by Cheridito, Filipovic, and Kimmel (), in which both mean reversion parameters and unconditional mean parameters are allowed to change under and . However, the use of such parametric form in multivariate credit risk models poses some difficulties due to the restrictions required to rule out arbitrage opportunities (see Feldhutter and Nielsen , Li and Zinna , for a discussion of this point). Furthermore, both sovereign and bank exposures—the main object of interest of this study—are not affected by the choice of the market price of risk, given that they are modeled under the measure.…”

Measuring sovereign credit risk using a structural model approach

“…Chesney and Morisset (1992), Claessens and Pennacchi (1996), Karmann and Maltritz (2003), Oshiro and Saruwatari (2005), Huschens et al (2007), and Gray et al (2007) used a structural model to predict default events. Duffie et al (2003) applied a reduced-form approach to measure Russia's default risk during the Asian financial crisis of the 1990 s. Arnold (2012), Li and Zinna (2013), and Acharya et al (2014) investigated the relationship between the sovereign credit risk and credit risk of banks.…”

“…Using data in different countries, Arnold (2012), Li and Zinna (2013), and Acharya et al (2014) showed the sovereign risk and credit risk of banks are related. Ang and Longstaff (2013) used CDS of U.S. Treasury, some U.S. states, and major Euro countries to show the systemic sovereign credit risk is more related to financial market variables than macroeconomic conditions.…”

Official Demand for U.S. Debt: Implications for U.S. Real Rates

“…The slice‐sampling method was developed by Neal (), and since then has been applied by a number of asset‐pricing studies (e.g., Li and Zinna , , Zinna ).…”