We examine the impact of economic policy uncertainty (EPU) on bank liquidity hoarding. We create a comprehensive measure of bank liquidity hoarding that takes into account asset-, liability-, and off-balance sheet activities. Using over one million bank-quarter observations, we find that in response to EPU, banks hoard liquidity overall and through all three components. This behavior is more pronounced for banks with less liquidity, more peer-bank spillover effects, and more EPU exposure. Additional analyses of interest rate spreads on several bank products suggest that our findings reflect at least in part bank choices, rather than just the reactions of customers.

While operational risk is generally perceived as idiosyncratic with limited systemic implications, we document that operational risk threatens financial stability. Using supervisory data on large U.S. Bank Holding Companies (BHCs), we find operational losses increase systemic risk through a direct channel that impairs market values of loss-experiencing BHCs as well as a channel of correlated losses that impact multiple institutions simultaneously. Findings are driven by tail events, more pronounced for systemically important and closer-to-distress BHCs, and vary by business lines, event types, and financial/economic environments. Our results extend the operational and systemic risk literatures and have key policy implications.

Bank prudential supervision is designed to enhance financial stability, but we are unaware of extant empirical research linking this supervision to financial system risk. In particular, there are no prior findings on how supervision enforcement actions (EAs) – major tools of supervisors – affect systemic risk. We empirically investigate relations between EAs and banks’ contributions to systemic risk. We find significantly smaller bank contributions to systemic risk after EAs than before them, suggesting that EAs are associated with enhanced financial stability. The data also suggest that the primary channel behind this relation is reduced leverage, but lower portfolio risk also plays a role. We also find that the magnitude of our findings is greater during financial crises than normal times, and that more severe EAs and EAs against banks are more effective in systemic risk reduction than that are those less severe and those against individual bank managers, respectively.