银行破产：偿付能力和流动性的作用（英）

Sergio Correia|Stephan Luck|Emil Verner Bank Failures:The Roles of Solvency and LiquiditySergio Correia,Stephan Luck, andEmil VernerFederal Reserve Bank of New York Staff Reports, no.1181February 2026 Abstract Bank failures can stem from runs on otherwise solvent banks or from losses thatrender banks insolvent,regardless of withdrawals. Disentangling the relative importanceof liquidity and solvency in explainingbank failures is central to understandingfinancial crises and designing effective financial stabilitypolicies. This paperreviews evidence on the causes of bank failures. Bank failures—both with andwithoutruns—are almost always related to poor fundamentals. Low recovery rates infailure suggest thatmost failed banks that experienced runs were likely fundamentallyinsolvent. Examiners’postmortemassessments also emphasize the primacy ofpoor asset quality and solvency problems. Before deposit JEL classification:G01 1Introduction The U.S. banking system has repeatedly experienced major waves of bank failures. Fig-ure 1 shows the rate of bank failures in the U.S. since 1863, highlighting spikes duringthe Panic of 1893, the 1920s agricultural downturn, the Great Depression, and the GlobalFinancial Crisis. Each crisis reignites the age-old debate about whether bank failures are Disentangling the roles of liquidity and solvency empirically is key to understandingbank failures and crises. It is also central to designing effective financial stability policies.Under the liquidity view, deposit insurance and public liquidity provision may suffice to In this review, we synthesize the empirical evidence on illiquidity and insolvencyas causes of bank failures and discuss what these findings imply for policies meant toaddress the incidence and consequences of financial crises. We narrow our focus alongthree dimensions.First, we focus on bank failures, defined as receiverships or other Notes: This figure plots the rate of bank failures from 1863 to 2024, defined as 100 times the number offailed banks over the total number of banks. For 1863–1934, the failure rate is based on national banks,as reliable data on bank failures for state banks are not consistently available throughout the sample.For 1935–2024, the failure rate is based on Federal Deposit Insurance Corporation (FDIC) member banks. bank runs. Third, our discussion primarily concerns the U.S. banking system, though We organize our discussion around a simple theory of bank runs and failures. Thetheory distinguishes between failures driven by insolvency or by runs on fundamentallysolvent but potentially weak banks.The theory shows that these are related but dis- The theory predicts that both failures driven by insolvency and by runs are morelikely in banks with weak fundamentals. The empirical evidence over the past 160 years of bank failures resoundingly supports this hypothesis. Bank failures are essentially al-ways associated with weak fundamentals, proxied by bank capitalization, profitability,and reliance on expensive funding. Banks that fail with and without runs display simi- The strong relation between weak fundamentals and failures rejects the extreme viewthat runs frequently cause the failure of clearly healthy banks. However, it does not dis-entangle the roles of insolvency and runs in weak but solvent banks.While a bank’ssolvency status can be blurry, especially in the fog of a crisis, theory implies that re- Contemporary bank examiners’ post-mortem assessments of the causes of bank fail-ures provide complementary insights. These assessments tend to emphasize asset losses, Why are bank runs on solvent banks not a common cause of bank failures?First,sleepy or inattentive depositors weaken strategic complementarities among depositorsand reduce the scope for self-fulfilling runs to occur. The high empirical predictability equity injections, depositor relationships, examination, and suspension of convertibilityallow illiquid but solvent banks to avoid costly fire sales. While these responses do not Given that banking distress most commonly stems from fundamental solvency issues,the evidence suggests that policies focused on maintaining and restoring bank solvency,beyond merely providing liquidity support, are necessary for preventing the most ad- 2Liquidity versus Solvency: Theory We begin with a simple illustrative framework that summarizes key themes from thetheoretical literature on bank runs and failures (see, e.g., Diamond and Dybvig, 1983;Morris and Shin, 2003; Rochet and Vives, 2004; Goldstein and Pauzner, 2005). We usethis framework to derive empirically testable predictions. The framework distinguishes SetupThere are two datest∈ {1, 2}. Att=1, a bank holds a mix of liquid risk-freecashCand illiquid risky loansL, financed by depositsDand equityE. Total initial assetsare thusA=C+L=D+E. For simplicity, we assume that this initial capital structure to pay all depositors at once in the initial period. There is a continuum of risk