Where the money is: branch network structure and bank profitability in Switzerland

Abstract

This paper investigates the impact of the geographic concentration of branch networks on bank profitability in Switzerland. Leveraging Switzerland's distinctive federalist fiscal structure, where communes set their own income tax rates, we also explore whether the tax burden of branch locations moderates this relationship. Using a hand-collected dataset of branch networks of commercial banks, we find that geographic branch network concentration positively affects bank profitability but only when banks concentrate their branches in low-tax communes. This suggests that the proximity to wealthy clients, who are more likely to reside in low-tax areas, enhances bank profitability.

KEYWORDS:

• Bank profitability
• banking
• branch network structure
• branch network concentration
• income tax burden

JEL Classifications:

• E44
• G21
• G32
• L11

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data is available on request from the authors.

Notes

1 https://www.handelszeitung.ch/unternehmen/jede-funfte-filiale-zu-das-ende-des-bankschalters. Note that bank branch closing is not only a phenomenon in Switzerland. In the U.S., e.g., Capital One and SunTrust shut down 32% and 22% of their branches from 2012 to 2017, respectively (see https://www.wsj.com/articles/banks-double-down-on-branch-cutbacks-1517826601).

2 See Bikker and Haaf (2002) and Schaeck, Cihak, and Wolfe (2009) on the impact on competition and concentration on bank profitability.

3 These correlations have been reported in several studies, such as Schmidheiny (2006), Basten, von Ehrlich, and Lassmann (2017) and Schaltegger, Somogyi, and Sturm (2011), but we find that this negative link between household income and tax-burden also exists during our investigation period.

4 In 2020, Switzerland's financial service activities were approximately 37.9 billion Swiss Francs in value, accounting for 8.5% of the real GDP and 12% of the total tax receipts (Federal Statistical Office 2021a).

5 Schmid and Walter (2012) provide a comprehensive outline of the logic of geographic diversification for financial service firms.

6 The role of branch location in profitability is less likely to derive from the liability side because customer deposits can easily be moved around geographically (i.e., transferred between branches) and banks have several alternatives to customer deposits (e.g., the interbank market and bond issues) to finance their activities.

7 For example, personnel costs for consulting services and costs arising from credit risk (e.g., loan default and interest shortfall).

8 This data is provided from 2010 onwards at https://www.bfs.admin.ch/bfs/en/home/statistics/regional-statistics/atlases/interactive-statistical-atlas-switzerland.html.

9 Note that we focus in this study on household taxation, linking low-tax burden and household income. We forego to incorporate corporate taxation as cantonal and regional banks place a greater emphasis on retail banking and asset management, making them less responsive to corporate taxation from a client perspective. In addition, corporate taxation plays a lower role on the communal level and data is only available on cantonal level.

10 Note that the definition of a commercial bank in Switzerland is not always straightforward as Swiss legislation allows all institutions with a banking license to operate in all business segments, including commercial banking, private banking, asset management, and investment banking.

11 For a detailed description of the bank categories, see Table A.1 in the appendix.

12 Note that Bureau van Dijk stopped Bankscope end of 2016 due to ownership conflicts.

13 Fortunately, this issue arose in our case with respect to only one bank.

14 To ensure that we did not include any aggregated statements, we always compared the balance sheet total reported in a given statement as available on Bankscope with the balance sheet total reported in a bank's annual report.

15 This determination was based on the ratio of net interest income to operating income. As cantonal banks are the phenotype of a commercial bank in Switzerland, we defined the lowest ratios of all 23 cantonal banks with a complete dataset for the 2006–2015 period for the years 2006 and 2015, respectively, as cut-off points. Banks with lower ratios were excluded from the sample. In some cases, however, this ratio can be misleading. To eliminate, for example, consumer credit banks, whose income also stems mainly from net interest income, we further stipulated that a bank's loan positions consist mainly of mortgage loans (mortgage loans over total gross loans). We similarly applied the lowest ratios of the 23 cantonal banks for the years 2006 and 2015, respectively, as cut-off points.

16 If not indicated otherwise, changes (openings and closings) of branches are considered effective as of the end of the reporting year. However, some banks do not present their annual list of branches as of the end of the reporting year and instead report a current list dated close to the release date of the annual report. In such cases and if not indicated otherwise, we consider changes effective as of the end of the year following the reporting year.

17 If, for example, a bank operated three branches in a particular year in a certain commune, the respective tax-burden in this commune is used three times in calculating the average tax-burden for that year. See the robustness section for variations of this measure.

18 A value of $\delta$ close to zero means that profits quickly adjust to their normal (average) levels and therefore implies a fairly competitive market. A value of $\delta$ close to one, on contrary, means that abnormal profits adjust only slowly to their normal (average) levels. See also Athanasoglou, Brissimis, and Delis (2008) for detailed explanations.

19 Monte Carlo studies show that the one-step estimator outperforms the two-step estimator in terms of producing both a smaller bias and a smaller standard deviation of the estimates (see Athanasoglou, Brissimis, and Delis 2008).

20 The variables we treat as endogenous are shown in italics in the result tables.

21 As a robustness check for this outlier, we excluded this specific bank from the sample, but it does not drive our findings as the results do not change.

22 Note that Table A.8 tests for model 1 without the interaction term, while Table A.9 reports the results for Model 2 using the interaction term with the tax rate.

23 This finding also reflects the small-scale spread of the Swiss territory.

24 Results are not shown for reasons of brevity but are available upon request.

25 Note that we also ran the same analysis for 2006. The scatter plot for 2006, which is not shown here for reasons of brevity, similarly suggests little evidence of correlation between BN_CONC and market power.

Additional information

Notes on contributors

Marc Domenic Fuchs

Marc Domenic Fuchs is a Senior Banking Professional. He holds a combined Master's Degree in Banking & Finance from the Lucerne University of Applied Sciences and Arts (Switzerland) and from the zhaw School of Management and Law (Switzerland) as well as a DBA from the Grenoble Ecole de Management (France). Previously, he held several positions in a large regional bank in Switzerland, last as Head of Business Development.

Jojo Jacob

Dr Jojo Jacob is Full Professor of Strategy at Rennes School of Business, France. He holds a PhD in Economics from Eindhoven University of Technology (Netherlands). Previously, he held positions as Associate Professor and Full Professor at Grenoble Ecole de Management (France) and as a Research Fellow at UNU-MERIT (Netherlands).

Florian Kiesel

Florian Kiesel is Associate Professor in Corporate Finance at Free University of Bozen-Bolzano (Italy). He holds a PhD in Economics and Finance from Technical University of Darmstadt (Germany). Before joining Free University of Bozen-Bolzano, he was Assistant and Associate Professor at Grenoble Ecole de Management (France) and Visiting Professor at Zeppelin University in Friedrichshafen (Germany).