In this paper we study systemic risk for the US and Europe. We show that banks’ exposures to common risk factors are crucial for systemic risk. We come to this conclusion by first showing that relations between US and European banks are smaller than within each region. We then show that European banks react more strongly to the onset of the financial crisis than US banks. Regarding the consequences of systemic risk, we show that dependence between the banking sector and a wide range of real sectors is limited. Our results imply that regulators and supervisors should address international bank dependencies arising from common risk factors, while recessions in real sectors due to bank defaults should be a secondary concern.

We examine the dynamics of bond correlation using a broad sample of US corporate bonds, and document that bond correlation varies heavily over time. We attribute this variation in bond correlation to variation in risk factor correlation reflecting time-varying flight-to-quality behavior of investors. We show that risk factor correlation increases when investor sentiment worsens, i.e., corporate bond investors exhibit stronger flight-to-quality when their sentiment is bad. Thus, bad investor sentiment leads to flight-to-quality behavior and, ultimately, high bond correlation. Very good sentiment, in contrast, can cause risk factor correlation and bond correlation to be negative.

Based on individual CDS transactions cleared by the Depository Trust & Clearing Corporation, we show that illiquidity strongly affects credit default swap premiums. We identify the following effects: first, transaction direction affects prices, as buy (sell) orders lead to premium increases (decreases). Second, larger transactions have a higher price impact. This finding stands in stark contrast to corporate bond markets. Third, traders charge higher premiums as a price for liquidity provision, not as compensation for asymmetric information. Fourth, buy-side investors pay significantly higher prices than dealers for demanding liquidity. Finally, inventory risk seems to matter little in explaining liquidity premiums.