Banking and Financial Integration

Elena Carletti (Bocconi University and CEPR), Agnese Leonello (European Central Bank), Robert Marquez (University of California, Davis)

The economic crisis brought about as a result of the ongoing Covid-19 pandemic has led to the intensive use of loan guarantees as a tool to stimulate bank lending. To give an order of magnitude, in Europe more than 320 billion euro in new loans were provided under PGSs in the four major European countries (France, Germany, Italy and Spain) as of September 2020 (ECB, 2020). These guarantees partially insure the bank against default by borrowers, paying either up to a fixed amount (a first loss position for the government) or sharing losses between the bank and government in specific proportions. While loan guarantees represent a useful policy tool for other instances of crisis, particularly when disruptions are believed to be short-lived as in the case of Covid-19, their use raises a number of important questions in terms of their implications for banks’ underwriting processes and thus, ultimately, for financial stability at large. In this project we plan to build a model that incorporates bank monitoring and financial fragility in the prevention of bank runs, and to study how the use of guarantees affects both the probability of bank runs and banks’ underwriting incentives. The results will shed light on the potential longer-term implications of loan guarantees and will thus contribute to the debate on the exit from public measures introduced as response to the Covid-19 outbreak.The economic crisis brought about as a result of the ongoing Covid-19 pandemic has led to the intensive use of loan guarantees as a tool to stimulate bank lending. To give an order of magnitude, in Europe more than 320 billion euro of new loans were provided under PGSs in the four major European countries (France, Germany, Italy and Spain) as of September 2020 (ECB, 2020). These guarantees partially insure the bank against default by borrowers, paying either up to a fixed amount (a first loss position for the government) or sharing losses between the bank and government in specific proportions. While loan guarantees represent a useful policy tool for other instances of crisis, particularly when disruptions are believed to be short-lived as in the case of Covid-19, their use raises a number of important questions in terms of their implications for banks underwriting processes and thus, ultimately, for financial stability at large. In this project we plan to build a model that incorporates bank monitoring and financial fragility in the sense of bank runs and to study how the use of guarantees affect both the probability of bank runs and banks’ underwriting incentives. The results will shed light on the potential longer-term implications of loan guarantees and will thus contribute to the debate on the exit from the public measured introduced as response to the Covid-19 outbreak.

Brunella Bruno (Bocconi University), Immacolata Marino (University of Naples Federico II and CSEF), Giacomo Nocera (Audencia, Nantes)

In this project, we empirically investigate how banks respond to a deterioration in their loan portfolios. This is a policy-relevant issue because high level of NPLs may increase systemic risk and impair banks’ ability to provide credit to the real economy. NPLs in Europe amounted to €1 trillion in 2015, as an effect of the subsequent global financial and euro sovereign crises. Policy makers around the world are now concerned that NPLs may potentially grow to even higher levels than in the past due to the COVID-19 outbreak.

Despite this relevance to policy, however, the existing academic research has been unable to establish a credible causal link between asset quality deterioration and an adverse effect on lending. In this paper, we fill this gap by exploiting the first Asset Quality Review (AQR) carried out by the European Central Bank as a quasi-experiment. The review determined an unexpected increase in NPLs in a subset of European banks due to the application of new and stricter classification criteria, with large heterogeneity among the reviewed banks. We find that a shock to NPLs has a negative impact on asset and lending growth. This effect is not linear, being stronger in banks located in high-NPL countries. The banks affected the most are undercapitalized, suggesting that NPLs influence the credit supply via a capital channel.

The policy implications of our findings are important. A negative macro shock determining a surge in NPLs threatens the banks' ability to provide credit. To cope with this requires a combination of measures, e.g., micro-prudential initiatives to recapitalize banks and comprehensive actions to resolve problem loans and return them to more normal levels, especially in those countries already affected by NPL issues.

Filippo De Marco (Bocconi University), Marco Macchiavelli (Federal Reserve Board), Rosen Valchev (Boston College).

The home bias - the surprisingly large portfolio share devoted to domestic assets – is a prominent feature of the data. In this paper, we go beyond the classic home bias problem and provide new evidence that international portfolios are related to investors' beliefs about the future prospects of foreign countries. Using data on the foreign sovereign debt holdings of European banks matched with their forecasts on future bond yields, we find that investors optimally exploit comparative advantages in information production. We rationalize the results in a model where partial information specialization arises endogenously by introducing some degree of "unlearnable uncertainty" about asset payoffs.

Brunella Bruno (Bocconi University), Immacolata Marino (University of Naples Federico II and CSEF), Giacomo Nocera (Audencia, Nantes)

This research project aims to assess the effect of the introduction of a stricter supranational supervisor and more harmonized rules on bank opacity. We plan to address this issue using the ‘asset quality review’ (AQR) exercise carried out by the European Central Bank in 2014 as a quasi-natural experiment in a ‘difference-in-difference’ framework. We will exploit the time variation induced by the change in the ECB regulation (before/after the AQR exercise) and the cross-bank variation (banks included vis-à-vis banks not included in in the AQR exercise) to identify the causal impact of stricter supervision on bank opacity. We will also explore the mechanisms that might have led to a change in bank opacity, by looking at any change in the contributions of the various sources of opacity occurring after the AQR. We will especially focus on the impact of non-performing loans (NPLs) by investigating whether the relation between NPLs and bank opacity has changed in response to the stricter scrutiny induced by the European single supervisor.