Knowledge that Transforms

Ongoing financial innovation raises the specter of banking and payment crises. Little aggregate evidence exists on the repercussions of substantial suspensions of payments. State-level experiments fill this gap. Four times in the last forty years, U.S. governors suspended payments from state-insured depositories. Rhode Island’s deposits crisis (1991), which was large, prolonged, and occurred during a recession, substantially lengthened and deepened the downturn. Deposits freezes in Nebraska (1983), Ohio (1985), and Maryland (1985), which were short and occurred during expansions, had little macroeconomic impact. Data sparsity inhibits analysis of these events with standard methods. To perform inference, we develop a novel Bayesian method for synthetic control, which generates output useful for policymakers and theorists. Our findings suggest policies that ensure institutions continue to process payments on a business-as-usual basis at all times have substantial value.

We estimate the effect of the spring 2023 failures of Silicon Valley Bank and Signature Bank on the “connectedness” of US bank stock return volatilities using the forecast error variance decomposition framework of Diebold and Yilmaz (2012, 2014) and Lastrapes and Wiesen (2021). Using split-sample and time-varying VAR methods, we find that those failures significantly increased spillovers across a sample of surviving regional banks, but had only small and temporary effects on spillovers across systemically important too-big-to-fail banks. Our main findings imply that regulatory policy toward systemically important banks has been credible but that additional oversight of regional banks should be considered.

I use data on unsecured consumer loans from Lending Club to study how peer-to-peer lending markets respond to monetary policy shocks. I find that both loan supply and demand decrease following unexpected increases in the federal funds rate. The contraction in supply is smallest for risky borrowers, while the decline in demand is largest for these borrowers. In contrast, both demand and supply increase following surprise LSAP contractions, with the increases being largest for risky borrowers. These findings suggest that peer-to-peer lending dampens the effectiveness of monetary policy transmission in unsecured consumer credit markets while increasing risk-taking.

We study the relationship between the behavior of a CEO’s ancestors and firm performance. To do so, we collect detailed information on emigrants from Italian municipalities during the Age of Mass Migration (1892–1924) from Ellis Island ships lists. We adopt an epidemiological approach complemented with an instrumental variables strategy and find that Italian firms managed by a CEO who belongs to a family with past emigration experience tend to perform better and to be more productive. In line with an inter-generational transmission of attitudes hypothesis, we show a positive relationship between the emigration experience of a CEO’s ancestors and alternative measures of corporate risk-taking. In addition, we find a positive relationship between having an ancestor who emigrated during the Age of Mass Migration and FDI to the United States. We also provide evidence that these CEOs have better managerial practices.

We analyze the relationship between cross-listing and innovation using a sample of firms from 40 countries spanning 1980–2016. We measure innovation through both the number of patents granted and citations received. Our results reveal a positive association between cross-listing and innovation, with this effect being more pronounced for firms from countries with poor legal environments and less developed financial systems. Overall, our findings align with bonding theory, suggesting that managers of cross-listed firms seek to bind themselves by adhering to the high legal and regulatory standards demanded by U.S. markets.

Despite the significance of social networks in influencing firm behavior, research on their impact on corporate tax behavior is limited. In this paper, we construct social networks of CFOs from U.S. companies based on their employment history, education, and non-professional activities. We find that firms with more socially connected CFOs have lower effective tax rates (ETR) compared to firms with less socially connected CFOs. This effect is more pronounced when corporate governance is weaker and managers have higher incentives. Furthermore, a firm's ETR decreases as CFO centrality increases. We do not observe similar results regarding the connectedness of boards of directors. Additionally, firm pairs exhibit similar ETRs when their CFOs are socially connected, suggesting an exchange of tax-related information among CFOs through their social networks. We also find that the past ETRs of firms with central CFOs predict the ETRs of firms with non-central CFOs. This indicates that less socially connected CFOs tend to follow the tax planning strategies of their more socially connected counterparts. Overall, our findings indicate that more socially connected CFOs possess more relevant information and resources regarding tax planning, leading to the adoption of more aggressive tax strategies compared to their less socially connected counterparts.

In recognizing the uniqueness of their business model, the FDIC launched a new community bank definition in 2012 (reaffirmed in 2020) that changed its approach to identifying this bank group. This paper examines the impact of this re-defined community bank status on bank performance. Using a quasi-difference-in-differences approach, the study finds that banks that obtain the community bank status exhibit greater financial stability and lower risk, with lending and deposit structures mediating these effects. These findings offer new insights into a "warm glow" effect brought by the re-classification, affecting the performance of these institutions. By assigning the community bank status, the FDIC may have tapped into the social and emotional significance tied to the word "community" for various stakeholders.

Using a large sample of European listed banks, we investigate the relationship between a bank’s business model and systemic risk between 2005 and 2020, a period which includes various episodes of instability. Our findings indicate that, during tranquil periods, banks with different business models exhibit similar sensitivity to systemic risk. However, during periods of instability, the type of business model becomes critical: investment banks contribute more to and are more exposed to systemic risk. Distinguishing between endogenous and exogenous crises, our results reveal that market-oriented banks contribute more to systemic risk when instability is endogenous to the financial sector. Conversely, focused retail banks consistently show lower contributions and exposures to systemic risk. Additionally, our findings highlight the importance of business model migrations in reducing systemic risk. Banks transitioning from diversified to more retail-oriented models reduce their systemic risk, whereas migrations in the opposite direction do not exhibit the same benefit. These findings underscore the importance of maintaining diverse business models in the banking sector to enhance financial stability. • The study investigates the relationship between a bank’s business model and systemic risk. • During tranquil periods, banks with different business models exhibit similar sensitivity to systemic risk. • During periods of instability, investment banks contribute more to and are more exposed to systemic risk. • Market-oriented banks contribute more to systemic risk when instability is endogenous to the financial sector. • Focused retail banks show lower contributions to and exposures to systemic risk across different crisis periods.

We study the relationship between bank failure and dedicated institutional ownership (hereafter IO) employing a logit model. We focus on dedicated institutional investors (hereafter IIs) as defined by Bushee (2001) and Bushee and NOE (2000) because they are stable shareholders and have large investments in the investee companies. Four results are obtained. First, based on the instrumental variable approach, a greater proportion of dedicated IO is associated with reduced probability of bank failure. This result is robust to the propensity score matching technique. The rationale is that dedicated IIs collect information on the investee banks by holding stable and concentrated positions in these banks, monitor them, and reduce their ownership in cases of trouble earlier than other IIs do. This effect has a larger magnitude in banks with greater organizational complexity and larger size. Second, after controlling for the sell herding effect of other IIs, we find that the dedicated IO proportion still has a negative and significant coefficient, indicating that dedicated IIs trade on fundamental information rather than herding with other IIs. Third, three potential channels of collecting information, (i) placing representatives on the board as directors, (ii) greater capacity in analyzing financial statements through cross-ownership in other banks, and (iii) higher monitoring incentive due to more stable and concentrated ownership, are investigated. We find evidence in favor of the effect of cross-ownership in the banking industry, ownership stability and concentration. Fourth, the ownership of dedicated IIs is significantly larger in banks acquired by other banks than those filing for Chapter 7 liquidation, ascribing a constructive role for dedicated IIs.

In this paper, we propose asset class liquidity risk indicators constructed by aggregating financial, monetary and credit variables. We measure the presence of liquidity in six highly representative markets such as the Equity Europe, Long Term Italian Government Bond, Short Term Euro Government Bond, Equity US, Bond Corporate Investment Grade USD, Short Term US Government Bond markets over the period January 2007–January 2023. Our approach allows for a time-varying measure of the relative contribution of the raw drivers to the asset class indicators. We use endogenous Markov-switching models to identify episodes of financial distress which have characterized the behaviour of assets over the last two decades. Finally, we map the Markov-switching regimes with bubble episodes identified via recursive testing procedures.