Print
E-MailCarola Frydman is Professor of Finance at the Kellogg School of Management at Northwestern University, and Chenzi Xu is an Assistant Professor of Finance at Stanford University Graduate School of Business. This post is based on their recent paper.
The survey paper is organized around three main lessons learned about banking crises: the importance of leverage as a precursor to crises, the large and negative real impact of crises on various sectors of the economy, and that government and central bank intervention has historically ameliorated these effects. To highlight recent advances, the paper surveys over two hundred empirical studies from the last twenty years (2000-2022) that cover banking crises occurring between 1800 and 1980.
The survey begins by discussing challenges and methodologies in studying historical banking crises empirically. One of the largest is in measurement: what is the right way to capture whether a banking crisis has occurred? The literature has used many, and while each approach is internally consistent, there are disagreements in the chronologies of crises that are available. More recently, the literature has shifted from using retrospective to contemporary sources, from qualitative to quantitative measurements, and towards capturing dimensions of crises beyond just the recognition that they occurred.
With the definition of crises used in the literature in hand, the survey examines the sources and transmission of banking crises, emphasizing that leverage (credit expansion) are critical predictors. Crises occur and escalate with bank runs, influenced by individual depositor behavior. This behavior reveals the significance of social networks, coordination, reputation and trust, and the role of informed and uninformed depositors in propagating financial instability. Heterogeneity in bank fragility also highlights the importance of maturity mismatches and institutional features, such as branching, as additional sources of vulnerability. Transmission of shocks occur both within the banking system, driven by interbank networks transmitting liquidity shocks in presence of moral hazard and asymmetric information; and internationally through exposure to risky assets such as sovereign debt, international bank expansion, and risky capital inflows, especially in emerging markets.
The paper then discusses the real effects of financial crises, which are consistently costly and cause worse declines in output and consumption than other crises. Importantly, the effects vary across time and place. These dynamics are crucial to understand from a policy perspective. One source of heterogeneity in the effects is the pre-crisis leverage levels mentioned above. Other factors affecting the consequences of crises include public sector debt, bank equity, and credit spreads. Although there always exists challenge of establishing causality between financial crises and economic downturns, researchers have used creative methods and unique historical context to isolate credit supply shocks from credit demand changes. For example, during the 1907 Panic, non-financial firms with board interlocks to affected trust companies experienced worse outcomes, accounting for a significant portion of the aggregate decline in corporate investment. Studies have also shown that during the Great Depression, contractionary monetary policy increased bank failures, leading to declines in commercial activity, manufacturing output, and firm exits. Additionally, lack of credit access during the crisis contributed to severe employment contraction in large firms. These results support the idea that disruptions to credit intermediation contribute to the severity of economic downturns and demonstrate the importance of studying financial shocks’ transmission to the real economy.
Historical perspectives on financial crises also have allowed researchers to highlight their long-term real effects. These include, for example, reduced innovation and persistent changes in the patterns of international trade. Moreover, financial crises have also been associated with various political outcomes, including increased political unrest and extremism, a relationship not observed during normal recessions. Notable examples include the rise of anti-Semitic propaganda and Nazi vote share in German towns affected by bank failures during the 1931 crisis and increased labor unrest and Communist Party membership growth in Chinese firms affected by the 1933 U.S. Silver Purchase program. The long-term effects are present in more than just the macroeconomy. Experiencing crises can also shape an individual’s risk preferences, making them less willing to take financial risks and participate in the stock market.
Finally, the paper discusses the role of institutions and interventions that ameliorate or exacerbate the likelihood of crises and their impact. Examining historical events provides insights into institutional features that affect bank stability, such as deposit insurance, bank competition, and prudential regulations. Deposit insurance, meant to prevent bank runs, can also increase bank risk-taking if it reduces depositor monitoring. Historical data from the early 20th century U.S. reveals insured banks attracted more deposits but had increased failures. Bank competition can create larger, more diversified banks that may be more stable but can also become “too big to fail.” Historical evidence suggests deregulation improved bank stability, but increased competition led to riskier behavior and higher likelihood of default.
Historical settings also offer insights into the role of interventions, such as monetary interventions and liquidity injections by the lender of last resort during financial crises. Examining historical events can help identify patterns and recurring issues that inform future policy decisions and the development of more effective safeguards against future crises. However, as central bank balance sheets in the modern era have grown, it is important to keep in mind that aggressive central bank interventions can result in moral hazard, leading to excessive risk-taking and more severe crises. Quantifying moral hazard and understanding the impact of bailout expectations on financial institutions remain challenges. Further research, drawing from both historical and modern perspectives, would help determine the optimal approach for crisis management and prevention of excessive risk-taking.
The survey concludes by highlighting many gaps in the literature for researchers to pursue in order to develop more effective strategies for preventing and mitigating future crises.
The paper is available for download here.
