Falling under the control of a different type of owner:risk-taking implications for banks

ABSTRACT

European banks have experienced significant changes in the type of entity that owns them (another bank, an individual or a family, a non-financial company, an institutional investor, a government, a foreign entity, a domestic entity…). In this paper, we look at the influence of ownership type changes on risk and profitability. Working with a panel of commercial banks from 17 European countries, we find that although banks that experience a change in ownership type do not exhibit lower or higher risk or profitability than other banks, their risk and profitability is significantly affected after the change takes place. The type of the acquirer plays a significant role in explaining the observed changes. When the acquirer is a non-financial company, the state or an institutional investor, the level of risk increases after the change while the level of profitability remains unchanged. Conversely, when the acquirer is a bank, we find that the level of risk-adjusted profitability decreases. Banks acquired by a different type of owner during the global financial crisis do not perform better or worse than they did before.

KEYWORDS:

• European banking
• ownership change
• type of ultimate owner
• profitability
• risk

JEL CLASSIFICATION:

• G21
• G28

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ We do not consider changes for which the owner’s type remains the same (i.e. a family-owned bank is taken over by another family, a bank controlled by a non-financial firm is controlled by another non-financial firm…). Such changes are not expected to alter risk-taking behavior in our framework. Moreover, ownership changes where the owner type remains the same are not observed in our data.

² Taboada (2011) investigates the impact of changes in bank ownership structure on the allocation of capital by looking into privatization but not the other dimensions of ownership structure.

³ The predicted values of our risk or performance measure (Y) are the fitted values derived from: the regression of the measure on country and years dummies $Y={\alpha }_0+{\delta }_{1}Country+{\delta }_{2}Year+\epsilon$.

⁴ There is some overlap between ownership type change and controlling country change (Foreign vs. Domestic). Therefore, in order to avoid singularity, we remove Ch_bank_private, Ch_privatization, Family_Ch, Ch_Domestic, Ch_Foreign in the estimates of equation (3). That is, we consider only 5 types of ownership changes instead of the 10 types described in Table 2.

⁵ The robust Hausman test indicates that the random effect method is suitable in our panel regression.

⁶ Estimates for the robustness tests which are not shown here are available from the authors on request.

⁷ Due to limited observations, we are not able to run the regressions on the rescued banks solely. The information on rescued banks comes from Petrovic and Tutsch (2009).