Bank market concentration and bank efficiency in the European Union: a panel Granger causality approach

Abstract

The relationships between bank market consolidation and bank efficiency are of particular relevance in the European Union (EU), but they remain controversial. Using a panel Granger causality approach, this paper contributes to the literature, testing not only the causality running from bank market concentration to bank efficiency, but also the reverse causality running from efficiency to concentration. The results obtained confirm the relative complexity of these causality relationships, although they generally point to a negative causation running both from concentration to efficiency and from efficiency to concentration. These findings are in line with the Structure Conduct Performance (SCP) paradigm and the suggestions that the increase of the banks’ market power will contribute to inefficiency, since these banks will face less competition to obtain more output results with less input costs. Our results suggest that within this panel of all 27 EU countries over a relatively long time period, from 1996 to the onset of the 2008 financial crisis, the more cost-efficient commercial and savings banks operated in less concentrated markets.

Appendices

Appendix I

Table 4 Yearly number of banks by EU country

Appendix II

Table 5 Concentration measure C3 = percentage share of the total assets held by the three largest banking institutions

Table 6 Helfindhal-Hirschman Index (HHI) = sum of the squares of all the country’s banking institutions’ market shares

Appendix III-Data Envelopment Analysis (DEA)

DEA was originally presented in Charnes et al. (1978), assuming constant returns to scale, which can be accepted as optimal but only in the long run. Later, Banker et al. (1984) introduced an additional convexity constraint (λ) and allowed for variable returns to scale. Following also Coelli et al. (1998), Thanassoulis (2001) and Thanassoulis et al. (2007), we can assume that at any time t, there are N decision-making units (DMUs) that use a set of X inputs (X = x₁, x ₂, …, x_k) to produce a set of Y outputs (Y = y₁, y₂, …, y_m), thus obtaining the DEA input-oriented efficiency measure of every i DMU, solving the following optimisation problem:

$$ \begin{array}{*{20}c} {{\min_{{\theta, \lambda }}}{\theta_i}} \hfill \\ {\begin{array}{*{20}c} {s.t.} \hfill & {\sum\limits_{{\mathrm{r}=1}}^{\mathrm{N}} {\mathrm{y}_{\mathrm{mr}}^{\mathrm{t}}\lambda_r^t\geq } y_{mi}^t} \hfill \\ {} \hfill & {\sum\limits_{{\mathrm{r}=1}}^{\mathrm{N}} {\mathrm{x}_{\mathrm{kr}}^{\mathrm{t}}\lambda_r^t\leq {\theta_i}} x_{ki}^t} \hfill \\ {} \hfill & {\lambda_r^t\geq 0} \hfill \\ {} \hfill & {\sum\limits_{{\mathrm{r}=1}}^{\mathrm{N}} {\lambda_r^t=1} } \hfill \\ \end{array}} \hfill \\ \end{array} $$

The DEA approach provides, for every i decision-making unit (DMU, here every country’s banking sector), a scalar efficiency score (θ_i ≤ 1). If θ_i = 1, the DMU lies on the efficient frontier and will be considered an efficient unit. On the contrary, if θ_i < 1, the DMU lies below the efficient frontier and will be considered an inefficient unit; moreover, (1-θ_i) will always be the measure of its inefficiency.

In the present study, the data are sourced from the IBCA-BankScope 2008 CD and the sample comprises annual data from the consolidated accounts of the commercial and savings banks of all EU countries between 1996 and 2008.

For the DEA estimates, we define the outputs and the input prices of the cost function using the following variables:

Dependent variable = Total cost (TC) = natural logarithm of the sum of the interest expenses plus the total operating expenses

Outputs:

• 1. Total loans = natural logarithm of the loans

• 2. Total securities = natural logarithm of the total securities

• 3. Other earning assets = natural logarithm of the difference between the total earning assets and the total loans

Inputs:

• 1. Price of borrowed funds = natural logarithm of the ratio interest expenses over the sum of deposits

• 2. Price of physical capital = natural logarithm of the ratio non-interest expenses over fixed asset

• 3. Price of labour = natural logarithm of the ratio personnel expenses over the number of employees

Appendix IV

Table 7 Yearly Data Envelopment Analysis (DEA) cost efficiency measures of the EU member states

Appendix V

Table 8 Panel unit root tests