ECB European Central Bank : How do banking groups react to macroprudential policies? Cross-border spillover effects of higher capital buffers on lending, risk-taking and internal markets

Giuseppe Cappelletti, Aurea Ponte Marques, Carmelo Salleo, Diego Vila Martín

Working Paper Series

How do banking groups react to macroprudential policies? Cross-border spillover effects of higher capital buffers on lending, risk-taking and internal markets

No 2497 / November 2020

Disclaimer: This paper should not be reported as representing the views of the European Central Bank (ECB). The views expressed are those of the authors and do not necessarily reflect those of the ECB.

Abstract

We study the impact of macroprudential capital buers on banking groups' lending and risk-taking decisions, also investigating implications for internal capital markets. For identication, we exploit heterogeneity in buers applied to other systemically important institutions, using information from three unique condential datasets, including information on the EBA scoring process. This policy design induces a randomized experiment in the neighborhood of the threshold, which we use to identify the eect of higher capital requirements by comparing the change in the outcome for banks just above and below the cut-o, before and after the introduction of the buer. The analysis is implemented relying on a fuzzy regression discontinuity and on a dierence-in-dierences matching design. We nd that, when parent banks are constrained with higher buers, subsidiaries deleverage lending and risk-taking towards non-nancial corporations and marginally expanded lending towards households, with negative eects on protability. Also, we nd that parents cut down on holdings of debt and equity issued by their subsidiaries. Our ndings support the hypothesis that higher capital buers have a positive disciplinary eect by reducing banks' risk-taking while having a (temporary) adverse impact on the real economy through a decrease in aliated banks' lending activity. Therefore, to ensure the eectiveness of macroprudential policy, it is essential that policymakers assess their potential cross-border eects.

Keywords: Macroprudential policy, Capital buers, Lending, Risk-taking, Internal capital markets.

JEL Codes: E44, E51, E58, G21, G28

ECB Working Paper Series No 2497 / November 2020

1

Non-Technical Summary

The global nancial crisis which erupted in August 2007 revealed the limitations of the supervisory framework in ensuring the resilience of the banking system to adverse macro-nancial shocks. In the euro area, this led to changes in the supervisory institutional setting by moving to a centralised banking supervision, while, at the same time, the European Union (EU) built up the macroprudential policy toolkit to address risks of a systemic nature.1 In this study, we are focused on the other systemically important institutions buer (OSII), which aims to reduce moral hazard and misaligned incentives by strengthening the resilience of too big to fail" institutions. Due to the international dimension of the banking sector, domestically oriented macroprudential policies might create unintended cross-border spillover eects. Banking groups constrained with higher capital requirements might restructure, through subsidiaries, their internal capital markets or negatively reduce the local supply of credit.2

In this paper, to explicitly analyse leakages of policy measures, we study the impact of higher capital buers, namely the OSII buers, on banking groups' lending and risk-taking and its further implications on the groups' internal capital markets. For identication, we exploit the heterogeneity in buers applied to other systemically important banks, using the information from three unique condential datasets, including the European Banking Authority (EBA) framework.3 The EBA setting relies on a two-step procedure: i) a scoring process, which automatically qualies a bank, with a score above a predetermined threshold, as systemically important;4 and ii) a supervisory expert judgement, which may qualify some banks below the threshold as systemically important. The EBA scoring process induces for a randomized experiment in a neighborhood of the threshold, allowing to identify the eect of higher capital requirements by comparing the change in the outcome of banks just above and below the cuto, before and after the introduction of the additional capital surcharge. This policy design allows us to implement an exclusive assessment, relying on both a fuzzy regression discontinuity and a dierence-in-dierences matching designs, which exploit the regulatory change and the discontinuity induced by the OSII identication process. The fuzzy regression discontinuity design is the econometric setup to assess the eects of higher capital buers on banking groups' lending, risk-taking and protability and the dierence-in-dierences matching5 is used to assess the implications of higher capital requirements in the internal capital markets of banking groups.

• From a nancial stability perspective, it was also important to mitigate a potential increase of banks' risk-taking due to
  monetary policy easing.
  2Macropudential measures are expressed in ratios, where banking groups can accommodate such higher capital requirements
  by reducing lending and risk-taking in subsidiaries of the group, thus freeing up capital at the consolidated level. 3Under Article 131(3) of the Directive 2013/36/EU ('CRD IV') and the EBA Guidelines (EBA/GL/2014/10).
  4A bank is designated as OSII if the score is equal or higher than 350 basis points. To account for the specicities of each EU member state's banking sector and the resulting statistical distribution of scores, authorities may raise the threshold up to
  425 basis points or decrease it to 275 basis points.
  5This alternative identication strategy is used given the less populated intra-group holdings dataset.

ECB Working Paper Series No 2497 / November 2020

2

In our study we establish two main ndings. First, subsidiaries of banking groups whose parent has been identied as systemic and, subsequently, constrained with a higher capital buer (OSII), reduced credit supply and risk-taking towards non-nancial corporations and marginally expanded lending supply towards households. At the same time, results show a reduction in aliated banks' protability explained by the banks' re-balancing behaviour for lending and risk-taking, i.e. credit shifting towards safer options. Second, lending and holding dynamics within banking groups are also aected when a parent bank is identied as systemically important, since our results indicate that parents cut down on holdings of both debt and equity issued by their subsidiaries.

In terms of nancial stability implications, our results suggest that the implementation of higher capital requirements at the consolidated level leads to a reduction in lending and risk-taking in the local credit mar- kets, particularly towards non-nancial corporations. We observe that this macroprudential policy, aimed at strengthening the resilience of banks, can also trigger an adverse eect in the real economy (as suggested also by Admati et al. (2015) and Cappelletti et al. (2019)).6 Also, our results follow the existent literature on the behaviour of the banking groups' internal markets (Campello (2002), Cetorelli and Goldberg (2012), Mili et al. (2017) and Buch and Goldberg (2017)) where banking groups react to a more stringent requirements by cutting down liquidity towards domestic and cross-border subsidiaries, therefore concentrating it around the parent. At the same time, as cited by Cappelletti et al. (2019), Gersbach and Rochet (2017)7 and Repullo (2004), higher capital requirements can reduce banks' gambling incentives, leading to a prudent equilibrium". Our ndings contribute to this debate suggesting that higher capital buer requirements have a positive disciplining eect by reducing banks' risk-taking, while having at the same time an adverse impact on the real economy via reduction of aliated banks' lending supply to non-nancial corporations and consequent protability of banks. Thus in terms of policy action, as suggested by Hanson et al. (2011) and Gropp et al. (2019), targeting the absolute amount of new capital to be raised8 instead of the capital ratio could mitigate the temporary adverse impact in the real economy, along with the potential optimisation of the risk-weighted-assets. Also, cross-border spillover eects should be factored in when assessing and calibrating macroprudential policy measures to ensure the eectiveness and consistency of macroprudential policy. It is essential that policymakers coordinate potential cross-border eects in the policy measures adopted by national authorities, in order to allow other cross-border authorities to adopt suitable reciprocating macroprudential measures.

• Banks tend to comply with higher capital requirements by dampening down their risk-weighted-assets, i.e. by deleveraging lending and risk-taking. Banks can increase capital ratios by: increasing capital (the numerator of the capital ratio) or by
  decreasing risk-weighted-assets (the denominator of the capital ratio) (Gropp et al. (2019)).
  7Authors show that imposing stricter capital requirement in good states corrects capital misallocation, increases expected
  output and social welfare.
  8As applied in the U.S. stress-tests conducted in 2009 (Hirtle et al. (2009)).

ECB Working Paper Series No 2497 / November 2020

3