Global Banking and Macroprudential Policy: New Evidence on U.S. Banks

Abstract:

Cross‐border regulatory arbitrage by global banks remains the main challenge to the effectiveness of macroprudential policy. This article aims to improve the understanding of the extent to which macroprudential initiatives around the world have shaped global banking, focusing on U.S. banks. An ad hoc dataset on the geographical distribution of on and off balance sheet activities of foreign branches allows the investigation of whether regulatory arbitrage has stimulated international leakages of macroprudential regulation. I first focus my attention on the lending behavior in those host countries in which foreign branches are left out of the regulatory perimeter via the institution‐targeted tools implementation. I then investigate whether macroprudential policy leakages due to regulatory arbitrage occur via off‐balance sheet activities. My findings suggest that institution‐targeted macroprudential regulation in host countries increases branches’ local lending, especially when U.S. banks are also exempted from the corresponding tool at home. I further find evidence in support of the fact that stricter local macroprudential policy has increased off‐balance sheet activities of resident foreign branches, as far as interest rate swaps exposures are concerned. From a normative standpoint, my results highlight the importance of reciprocity and international cooperation among macroprudential regulators

Keywords::

• macroprudential policy
• global banks
• foreign branches
• regulatory arbitrage

JEL Classification Codes::

• G21
• F34
• E0

Notes

1 Global banks are known to pro‐actively adjust their activities in response to cross‐country heterogeneity in financial regulation (for instance, see Houston, Lin, and Ma 2012; Bremus and Fratzscher 2015; Ongena, Popov, and Udell 2013).

2 Empirical evidence supports the existence of this channel in several countries. These results stem from country studies, carried out on a coordinated basis by International Banking Research Network (IBRN), see Buch and Goldberg (2017) for details.

3 To date, some efforts from regulators have been made to contain the risk of leakage. In particular, reciprocity on the countercyclical capital buffer is embedded in the Basel III agreement, in effect from 2016. Yet, reciprocity on other macroprudential tools is not mandatory

4 The literature recognizes two further channels of macroprudential regulation transmission due to global banks. Prudential policy stance implemented in a foreign country where global banks have affiliates can affect lending of these latter in their home country via an inward transmission channel (Buch and Goldberg 2017). In this case, stricter foreign borrower‐targeted tools may result in a repatriation of loans extended by global banks, resulting in an increase in domestic credit. Also, when global banks are subject to institution‐based regulation in their country, leakages may arise from a reduction in their activities at a consolidated international level (Kollmann 2013; Kollmann, Zeno, and Muller 2011).

5 See Altunbas, Binici, and Gambacorta (2018) for a very comprehensive discussion on the link between macroprudential policy and systemic risk. For a discussion on shock contagion across banks located in the same host country, see Fiala and Havranek (2017).

6 Foreign claims are defined by the Bank of International Settlements (BIS) as the sum of cross‐border claims of banks of a given nationality vis‐à‐ vis unaffiliated foreigners and local claims of their foreign offices located in a given country. See BIS (2017) for detailed definition of variables included in the international consolidated banking statistics.

7 Systemic externalities arise mainly from fire‐sale, interconnectedness of the financial sector and strategic interactions across banks which lead to build‐ups of leverage, credit growth, and wholesale funding (De Nicolò Favara, and Ratnovski 2012).

8 In a currency union, such as the Euro area, it has, however, been shown by Michal Brzoza‐Brzezina, Kolasa Marcin, and Krzysztof Makarski (2015) that asymmetric macroprudential policies within members of the currency union may be desirable to help stabilize the output of peripheral countries.

9 Through a Freedom of Information Act request.

10 Micro‐data is confidential and thus not released by the FFIEC. Branches contained in the report, also referred as the FFIEC030 report, have assets in excess of $250 million. This reporting threshold of $250m captures large and medium sized branches which constitute more than 90% of the overall population.

11 Obtained by the summation of the relevant policy tools dummies.

12 As highlighted in Udell (2014), borrower‐based macroprudential tools such as LTV ratios can be applied to some types of business and corporate lending (such as factoring and leasing), although this is less common.

13 It is customary to consider lagged variables to attenuate endogeneity problems. Although endogeneity should not be a major problem in my estimation, I correct for this in the robustness checks section via a dynamic GMM estimator.

14 Following Beirne and Friedrich (2017), exchange rates are not included in the empirical analysis as highly correlated with the inflation rates.

15 General Countercyclical Capital Buffer/Requirement (CTC) is not considered here as this tool is seldom applied in my sample host countries (see Table A2).

16 See Michael Brei and Leonardo Gambacorta (2014) for a discussion on leverage ratio measures available in the United States before the crisis.

17 Dynamic GMM panel regressions could alternatively be estimated by a difference estimator consisting in instrumenting the equation in differences with the lagged levels of the explanatory variables. These instruments, however, have been found to be weak in the case of trending variables (Blundell and Bond 1998).

18 Since I am here focusing on the funding structure, on the relative weight of deposits in overall funding in particular, I consider deposits as a ratio of total assets rather than the log of deposits.

19 Positions are vis‐à‐vis BIS reporting banks.

20 See Nadia Benbouzid, Sushanta Mallick, and Ricardo Sousa (2017) for a discussion on how financial structures may affect off‐balance sheet securities.

Additional information

Notes on contributors

Carmela D’Avino

Carmela D’Avino is at the ICN Business School Nancy Metz: Finance, Audit, Accounting and Control, Nancy France, and CEREFIGE, University of Lorraine, France.