State-owned banks in the market for corporate control

Abstract

We study the pre-deal characteristics of state-owned banks acquiring other companies, relative to their private counterparts. We build a unique international data-set of 3682 deals in the years 2003–2013. Econometric results highlight that those state-owned banks that are acting as acquirers have an ex-ante performance similar to their private benchmarks. The results are driven by the role of development banks. This new finding points to the recent evolution of some types of contemporary state-owned financial players.

Keywords:

• state-owned banks
• development banks
• ownership
• mergers and acquisitions

Jel classifications:

• G32
• G34
• L32

Acknowledgments

The research has been supported by the Jean Monnet Network coordinated by Massimo Florio. The authors are grateful to the EU institutions and particularly to the EACEA for their support. The authors thanks two anonymous reviewers, and Miguel Morgado, Director, European Investment Bank, for valuable comments and suggestions. All errors and omissions however remain our own.

Notes

1. These trends are in line with a more general return of state-owned enterprises as players in several industries. In the last decade, 2057 major state-owned enterprises have achieved a combined equity value of almost $2 trillion, more than 6 million employees, operating revenues equal to 19% of global cross-border sales and around 6% of the world GNI. Furthermore, since the 2008 financial crisis, the role of state-owned enterprises in developed economies has been increasing dramatically: the shares of equity holdings by some OECD governments have increased to 20% of their GDP (OECD 2010, 2011). Notably, in 2011, the value of the assets of the targets purchased by state-owned enterprises was equal to 433 billion Euros, corresponding to 17% of the total assets of the targets that have been traded in the M&A arena (Zephyr data).

2. In general, such bailout acquisitions should not be considered for comparison with private-owned banks deals, because they are temporary and governments plan to privatize these banks again as soon as recovered (European Commission 2014, 2015; for a critical analysis of the financial regulation and supervision in the EU after the crisis see Quaglia 2013). However, in our sample we did include them among government-led deals because nowadays such banks are still ultimately owned by governments; besides they represent less than 10% of M&A deals involving commercial state-owned banks.

3. Although we acknowledge that the categorization into commercial and development may hide the fact that some state-owned banks – mainly in developing countries – are engaged both in retail and development lending, a visual inspection of our sample highlight that this “grey area” is limited to a very small number of cases and banks can be properly categorized into the two subsets.

4. See DeYoung, Evanoff, and Molyneux (2009) for a comprehensive literature review of financial institutions’ M&As.

5. https://www.bvdinfo.com. Zephyr is a data-set that contains information about deals, while Bankscope is a database of banks’ financial statements. Given our research question, we matched the two data sets.

6. We considered to be state-owned any bank of which the ultimate owner, defined as the independent shareholder with the highest direct or total percentage of ownership, is a central or local public entity, including public authorities, governments, municipalities and local entities. Further, we considered this independent shareholder to be an ultimate owner (UO) of a bank if it holds more than 25% of the shares

7. To ascertain the pre-deal ownership nature of the target, we looked at the ownership type of the vendor, while we considered the acquirer’s ownership to infer its post-deal ownership. In the rest of the paper, we thus report information on the ownership of targets and acquirers.

8. Our setting is similar to that of Hannan and Pilloff (2009) and Hernando et al. (2009), since we focus on predeal banks’ characteristics and on accounting measures of performance. However, in the multivariate analysis, we use a different econometric specification, since our research questions relate to banks involved in M&A deals rather than banks’ probability of being involved in a deal.

9. We use the ROA, instead of the return on equity (ROE), since our sample is worldwide and the ROA is better equipped for a cross-country analysis of banks with different levels of capitalization and leverage (Rivard and Thomas 1997; Athanasoglou, Brissimis, and Delis 2008).

10. Specifically, in 2009, 39% of development banks had a non-performing loan ratio higher than their national average while 64% of development banks were below the national average; the percentage of development banks with a better ratio was rising compared with previous years. However, it would be interesting to distinguish whether lower non-performing loan ratios are due to their ability to manage risk and to select long-term investment that are both sound and profitable, or whether they underline excessive risk aversion and short-termism, that would be in contrast with their role in fixing market failures.

11. Interestingly, but not reported here to save space, the crucial year of the Great Recession has a positive and statistically significant sign for all the acquirers. Details can be provided by the authors upon request.

12. Details of the estimated time effects can be provided by the authors upon request.

13. The details of the composition of these clusters can be obtained from the IMF or the World Bank website.

14. Brazil, Russia, India, China, South Africa and Turkey.

15. According to the IMF and the World Bank: Hong Kong, Israel, Singapore, South Korea, Taiwan, Qatar, Cyprus, Slovenia, Malta, the Czech Republic, Slovakia, Estonia, San Marino, Croatia, Latvia and Lithuania.