Systemic Risk and Foreign Currency Positions of Banks: Evidence from Emerging Europe

Abstract

This article investigates the impact of foreign currency (FX) denominated assets and liabilities on systemic risk by using a unique hand-collected dataset of bank-level FX positions for the period 2005–2012. The sample consists of banks from Central and Eastern Europe with large share of FX exposures on the balance sheet. Empirical findings indicate that systemic risk is dependent on the currency choice of financial institutions. While FX positions denominating in EUR and USD do not pose systemic risk concerns, FX positions denominating in exotic currencies like CHF significantly enhance the banks’ systemic importance. The negative influence of CHF positions might be reduced through prudent corporate governance mechanisms, tight restrictions on banking activity, and more restrictive limits on foreign currency lending in the home countries of the bank holding companies.

Keywords:

• foreign currency assets
• foreign currency liabilities
• governance
• Swiss franc
• systemic risk

JEL Classification:

• G21
• G32
• G34

Acknowledgments

We thank two anonymous referees, the Editor (Josef Brada), and Andreea Stoian for their comments and suggestions. We also thank participants at the INFER Annual Conference, Bordeaux 2017; INFER Workshop on Rethinking Macroeconomic Development and Macroeconomic Policy, Cluj-Napoca 2017; Infinity Conference, Valencia 2017; EEFS Conference, Ljubljana 2017; MFS Conference, Bucharest 2017; EEFS Conference, Amsterdam 2016; and EuroConference, Porto 2016 for helpful comments.

Notes

1. Even in the case of perfectly hedged FX positions on a bank balance sheet, exchange rate risk still exists for banks with FX loans if its borrowers register currency mismatches.

2. While CoVaR estimates the spillovers from banks to the system (contribution to systemic risk), MES estimates the contagion effects from the system to a particular bank. The time-varying linear spillovers between banks and the system are determined using market capitalization returns.

3. In contrast, the methodology proposed by BIS to identify systemically important financial institutions (SIFIs) is based on a system of accounting balance-sheet data with a lower frequency, usually quarterly.

4. The percent corresponds to the CEE region, excluding Russia.

5. The level of foreign ownership varies from around 35% in Ukraine and Slovenia to around 80% in Romania and Serbia.

6. In order to prevent further large uncoordinated withdrawals of bank-holding companies from the CEE region, in January 2009, the European Bank Coordination “Vienna” Initiative (EBCI) was launched. Its aim is to provide a crisis resolution framework, and it involves systemically important groups with subsidiaries in the CEE.

7. A systemically important bank is a financial institution with complex and international activity the collapse of which could negatively affect other banks and jeopardize the financial system stability (BCBS, 2013).

8. The countries are Bulgaria, Croatia, Czech Republic, Hungary, Lithuania, Poland, Romania, Serbia, Slovakia, and Ukraine.

9. System is defined as the total market capitalization of banks from the sample.

10. Value at Risk is the maximum possible loss of a bank’s market equity at a given confidence level, such as 95%.

11. Expected Shortfall is the average loss of a bank’s market equity at a given confidence level, such as 95%, corresponding to the 5% worst outcomes.

12. Adding country-fixed effects does not change the results, since they are perfectly collinear with bank-fixed effects.

13. The governance indices range from 0 to 1. A higher score is associated with a stricter internal governance framework.

14. The higher the value of regulatory indices is, the stricter the external governance framework.

15. In these models, the variable of impaired loans ratio was abandoned, as it is highly correlated with CDS data.

16. Also assessed was whether constraints on loan-to-value ratio, debt-to-income ratio, interbank exposures, and housing-specific loan loss provisioning from Cerutti, Claessens, and Laeven (2017) in home and host countries would affect the relationship among CHF positions and systemic risk. No significant impact has been found.

Additional information

Funding

This work was supported by the Romanian National Authority for Scientific Research and Innovation, BRIDGE GRANT DSS-Direct [PN-III-P2-2.1-BG-2016-0447] and the Romanian National Authority for Scientific Research and Innovation, CNCS – UEFISCDI [PN-II-RU-TE-2014-4-0443].

Notes on contributors

Alin Marius Andrieş

Alin Marius Andrieş is Associate Professor at the Faculty of Economics and Business Administration, Alexandru Ioan Cuza University of Iaşi, Romania.

Simona Nistor

Simona Nistor is Lecturer at the Faculty of Economics and Business Administration, Babeş-Bolyai University of Cluj-Napoca, Romania.