University of Groningen
Faculty of Economics & Business
MSc International Economics and Business
Master Thesis
The Effect of merger and acquisition on Bank Performance
Case of the EU(15) countries, 2002-2008
Author:Yifei LI
Student Number: 1940562
Email: Andrea868@163.com
Supervisor: dr. Micheal Koetter
Co-assessor: dr. Dirk Bezemer
The Effect of Merger and Acquisition on Bank Performance
Case of the EU(15) countries, 2002-2008
Yifei Li
ABSTRACT
This paper investigates the effect of merger and acquisition on bank performance. By examining the merger and acquisition deals incurred in EU(15) between the year 2002 and 2008 I analyze whether the M&A could lead to performance improving. For the performance measurements, this paper chooses to use both accounting ratios and X-efficiency to test the three hypothesizes. From the empirical result I find that acquirer bank usually get higher average ROAA and ROAE ratios and little more efficient than target bank. Non-M&A banks generate lower ROAE and ROAA ratios but higher cost efficiency scores than M&A banks. All hypothesize testing are insignificant statistically which mean the there are no significant difference between the two comparison parts in the three hypothesizes. The results reflect that M&A deal may change bank performance and have impact on the bank operating ratios but the effect is really little. .
Table of content
1. Introduction ...4
2. Literature review ...7
2.1 Corporate governance effects of bank M&A ...7
2.2. Performance effects of bank mergers and acquisitions ...8
3. Methodology...12
3.1 Accounting ratios for performance comparison ...12
3.2 Cost efficiency for performance comparison ...12
3.2.1 Stochastic frontier analysis ...13
3.2.2 Overview of bank cost function...14
3.3 Data and sample ...15
3.4 Cost efficiency measure and model ...16
4. Empirical Result ...19
4.1 Efficiency result ...19
4.2 Results on M&A effects ...20
1. Introduction
This paper aims to use stochastic frontier analysis to study the effect of mergers and acquisitions (M&A) on bank performance. Last decade has witnessed the unprecedented wave of mergers and acquisitions in banking industry at global level especially for European regions (Rezitis, 2006). Table 1 shows the changes of number of credit institutions and M&A in EU(15) countries in the year of 1999 and 2009. According to this statistics of European Central Bank (ECB), except Ireland, the number of credit institutions declined dramatically with an average rate of 16% from 1999 to 2009. In the meantime mergers and acquisitions among financial institutions increased visibly with an average growth rate of 75%. These two trends can also be seen from figure 1 and figure 2. In figure 1 the number of credit institutions presents obvious decrease while M&A deals in EU(15) countries rise significantly in figure 2.
of the study in performance effect of bank M&A are focus on US banking market. Small numbers of research in EU banking sectors are analyzing M&A impact in early years and their results are mixed. For example, Beccalli and Frantz have collected EU acquirers and world wide targets from 1991 to 2005 to study performance effect of bank M&A. Koetter (2004) investigated bank efficiency changes before and after M&A using data of Germany banks during 1993 to 2003. Both the two studies show efficiency improving after M&A. While Altunbas et al. (1997) find little changes of bank efficiency in post-M&A period. Considering the trends showed in figure 2 and the reason mentioned above, it is worth to study whether M&A can lead to performance improving for banks in most recent years. Moreover it is also worth to compare the post-M&A performance of bank with that of non-M&A banks. Only in this way could we get a complete overview of M&A effect in the whole banking industry. Given the above explanation this paper intends to get insight into the effects of mergers and acquisitions involving European banks during 2002 to 2008.
This paper will use parametric approach of stochastic frontier analysis (Aigner et al., 1977; Battese and Corra, 1977; Meeusen and Broeck, 1977; Battesse and Coelli, 1995) to calculate the cost efficiency of the sample banks and compare with the banks not involved in merger and acquisitions. This technique not only provides results of efficiency rankings but also is in accordance with the result of standard non-parametric measures of performance (Weill, 2004). The hypothesizes of the paper are
H1: The average cost efficiency and ROAE and ROAA of acquirer bank in EU(15) is higher than that of target banks before 2 years of merger and acquisition.
H2: The average cost efficiency and ROAE and ROAA of the acquirer bank after 2 years of the merger and acquisition in EU(15) is improved compared with that before 2 years of merger and acquisition
banks not involved in merger and acquisition.
The first two hypothesizes compare the performance effect of bank which involved in M&A deal. Test of H1 aims to exam if the acquirer bank perform better than target bank before M&A. H2 tests whether the bank improve performance by M&A. The third hypothesize compares performance of M&A bank in post-M&A period with non-M&A banks.
Beside hypothesizes test, this paper also describes the comparison performance result by different bank types and countries. This could give a detailed view of the whole sample and explain the impact of bank type and country on bank performance.
2. Literature review
Scholars analyze the effects of bank M&A mainly in two aspects which are 1) the effect on corporate governance, 2) the operating performance of both parts of firms before and after M&A.
Given this paper tend to study the impact of bank mergers and acquisition on bank performance, the following part will give a brief literature overview on the first aspect and the last aspect will be explained specifically in section 2.2.
2.1 Corporate governance effects of bank M&A
2.2. Performance effects of bank mergers and acquisitions
The debate that whether mergers can have a beneficial impact on banking and the public as a whole has become more and more important following the rapid pace of consolidation (Peristiani, 1996). A question arising from this debate is whether mergers are an effective restructuring tool. There are mainly three approaches to evaluate bank performance before and after mergers and acquisitions. One uses measurement of accounting profitability like return on equity and cash flow return. According to the literature by Delong and DeYoung (2007), accounting ratios measure actual financial performance during certain time period and provide researchers a way to study crucial elements of financial performance and also get overview of the overall financial performance. Therefore Delong and DeYoung (2007) evaluate the long-run change in financial performance after mergers by seven factors which are ROA, ROE, interest margin, cost efficiency, loans-to-assets, core deposits-to-assets and noninterest income ratio. They found from these measurements that banks financial performance improved after merger activities. Altunbas and Marques (2007) use the change between the merged banks’2 years average ROE after the acquisition and weighted average of ROE of the merged banks’2 years before acquisition as dependent variable of the model. Pilloff (1996) measured financial effect of acquisitions on bank performance by comparing profitability and efficiency in pre-merger period to that of post-merger period. Comett and Tehranian (1992) examine the pre and post-merger performance of 20 large holding-firm mergers during 1982 to 1987 and show the cash flow returns are better improved than a national group of publicly traded non-merger banks. However these accounting ratios are historical figures and often neglect current market values ( Pilloff, 1996).
and Ryngaert (1994) investigate the weighted sum of acquirer and target abnormal returns when merger occurred and found no wealth creation. However Cornett and Tehranian (1992) show positive relationship between merger and abnormal returns. Delong and DeYoung(2007) adopt an event study methodology to capture the initial stock market reaction to 216 merger deals. They use cumulative abnormal returns and show the market reaction to bank M&A is neither more favorable nor less favorable for the sample. Nevertheless market reaction evaluation only rely on market expectations and do not capture the actual gains after mergers and acquisitions (Pilloff , 1996).
The third approach tests the changes of X-efficiency of banks before and after mergers and acquisitions. Given the explanation of Pilloff (1996) that the main motive for bank mergers and acquisitions is cost reduction thus most of the empirical studies examine X-efficiency to demonstrate whether bank performance improved after mergers and acquisitions. According to Leibenstein (1966) technical-efficiency is the effectiveness of the firm to produce an output under a given set of inputs. X-inefficiency happens under the situation that technical-efficiency can not achieve because of lack of competitive pressure. It measures the distance of the position of equilibrium of each bank form the optimal operative frontier (Beccalli and Frantz, 2008). X-efficiency includes cost and profit efficiency. Cost-efficiency is most commonly specified efficiency criterion to evaluate how close a bank’s cost is to a best-practice bank’s cost would produce with the same bundle of outputs. Similarly profit efficiency measures how close a bank is to the realization of the maximum possible profit under certain level of input and output.
only way to determine whether mergers and acquisitions are in public interest. Society can gain from M&A if they improve X-efficiency. In addition cost and profit efficiency separate improvement in efficiency and market power effects and it can also give prediction whether mergers and acquisitions could achieve efficiency gains. Thirdly, many researches have showed that technical and allocative inefficiencies may dominate scale and product mix economies. Demsetz (1973) asserts that bank behavior is driven by cost efficiency rather than market power. Moreover banks will be forced to achieve maximum X-efficiency considering the competition pressure and technological advances (Vennet, 2002). In conclusion, I plan to use the cost efficiency to examine the impact of mergers and acquisitions on bank performance given the above reasons.
X-efficiency estimation follows two main approaches which are parametric and non-parametric approaches. Bauer et al. (1988) provide three parametric approaches (SFA, DFA and TFA) and one non-parametric approach (DEA) to estimate X-efficiency. For comparison of these different approaches, Bauer et al. (1988) found that the result following parametric approach is more consistent with standard measure of performance than DEA. However Weill (2004) argues that this finding is relevant for USA banking data. Resti (1997) found similar results using both SFA and DEA after examine 270 Italian banks. Drake and Weyman-Jones (1996) investigate British building societies and get high correlation between the efficiency scores which derived by SFA and DEA. Sheldon (1994) gets no relationship between SFA and DEA scores.
period and find that there are a large percentage of mergers that decrease profits and efficiency. Huizinga et al (2001) analyze 52 EU bank mergers from 1994 to 1998 and found that cost efficiency of post-merger banks is higher than per-merger and profit efficiency improves only marginally. Beccalli and Frantz (2008) also find the improved performance after merger and acquisitions by examining large simple over the period 1991 to 2005. Considering specific region analysis, Koetter (2008) generates all Germany savings and cooperative bank mergers during 1993 and 2005. He finds that mergers can improve cost and profit efficiency especially the latter one. However Anthony N. Rezitis(2007) analyzes the bank mergers and acquisitions in Greece and found the negative effects of M&A on technical efficiency and total factor productivity. Sufian (2007) use the Data Envelopment Analysis (DEA) to investigate the effects of mergers and acquisitions on Singapore domestic banking groups’ efficiency. This study finds the mean of overall sample’s post-merger efficiency is higher than that of pre-merger period. Viverita examines the impacts of merger on commercial bank’s performance in Indonesia during 1997 to 2006 and gets the efficiency improving results.
Given other literatures, profit efficiency is also be evaluated for analyzing bank performance before and after M&A. Humphrey and Pulley (1997) believe that only cost efficiency can not generate bank performance. Therefore Hackethal et al. (2004) use alternative profit model to get the bank profit efficiency. Koette (2008) also generate bank profit efficiency model and compares to what extent both cost and profit efficiency improve pre and post-merger and acquisitions.
3. Methodology
3.1 Accounting ratios for performance comparison
This paper uses two important accounting ratios to capture the performance changes before and after M&A: ROAA (return on average asset) and ROAE (return on average equity). ROAA mainly reflects a bank’s profit relative to size and it is equal to net profit divided by average total asset. This ratio measures how efficient a bank can exploit its assets and it is commonly used for performance comparison among banks. ROAE refers to net income over the average shareholders equity. This reflects a bank’s profitability and presents how much profit a bank could make using the investment of shareholders. In this paper the accounting ratios and CE in the year of M&A will not be compared because of the strong impact of the deal on bank accounting operation. Instead I will capture ROAA and ROAE in both one year and two years before and after M&A.
3.2 Cost efficiency for performance comparison
3.2.1 Stochastic frontier analysis
This paper follows the parametric approach to investigate the cost efficiency of M&A and non-M&A banks. As mentioned in section 2, cost efficiency is wildly used concept for measuring the bank performance and it not only gives information on the production wastes and also on whether the firm gets optimal outputs under certain inputs (Weill, 2004). The method used to generate cost efficiency is stochastic frontier approach. The main stochastic frontier model specifications contain an error components specification which both allowed time-varying efficiencies and directly affected by certain variables (Battese and Coeill, 1992, 1995).
The stochastic frontier model can be referred to as a “composed error” model which means that a firm’s cost can be explained by an error term that includes two components, one is a random noise Vs’ and the other is technical inefficiency Us’. The standard model for estimation cost efficiency is
ci=c(yi,wi,;β)exp(vi+ui) (3.1) Equation (3.1) is the cost function of an industry where
ci is the total costs of the firm yi is the vector of the firm output wi is the price of the inputs βis the unknown parameter vi is random variable
ui is the non-negative random variable
In this equation vi is assumed to be iid N(0,σv2) and ui is assumed to calculate for cost inefficiency of production. In order to make the noise vector captures noise specifically, it is better assume vi and ui are independent from each other.
If the equation transforms to log-linear Cobb-Douglas function, then it can be written as
The model can be estimated by maximum likelihood method under these assumptions for ui and vi. Following the maximum likelihood method in the study of Kumbhakar and Lovell (2000) the parameter β and ui can be estimated by two steps. First step use OLS to evaluate the slop parameters and the second step use the maximum likelihood to calculate the intercept and the variances of the two random variables. Then the β and ui can be got if the likelihood function is maximized.
Given the parameters can be estimated using the maximum likelihood method then the average error term which is equal to average of the efficiency term ui can be computed. Then the cost efficiency can be obtain with the equation
CEi=exp(ui) (3.3)
3.2.2 Overview of bank cost function
Most of the literatures focus on cost and profit efficiency study of banking sector. The comparison of these two efficiency items rely on the assumption that whether the market is perfect (Cavallo and Rossi, 2002). In order to analyze the research question , this paper choose to use cost function approach and make the assumption that European Union banking industry markets is perfectly competition markets. Because the cost function approach can estimate multiple outputs model while the production function approach only make assumption of single output using stochastic frontier analysis.
institutions as a whole. In addition, frontier efficiency analysis can be better used under intermediation approach given it excludes interest payments.
3.3 Data and sample
The data set is obtained from two sources: Zephyr database for data on the M&A operations; Bank scope for balance sheet data and profit and loss data of M&A and non-M&A banks.
The sample comprises European Union (15) banks (commercial, saving and cooperative banks) involved in M&A deals between 1/01/2002 and 31/12/2008 and the comparison sample contains non-M&A EU banks from 2002 to 2008. The sample includes three types of banks which are commercial, saving and cooperative banks. Commercial banks offer both commercial and retail services and they are usually profit oriented. Therefore this type of bank always focuses on the investment and wholesale services. While saving and cooperative banks emphasize on retail business. They mainly serve for the customers and small and medium sized companies ( Agarwal and Elston, 2001) and most of their funds is consumer deposits.
3.4 Cost efficiency measure and model
For the inputs and outputs indicators I follow the intermediation approach and refer to Ferrier and Lovell (1990), Allen and Rai (1996), Eisenbeis et al. (1999) and Koetter (2005) to compute the result. The three outputs are total securities (y1), off-balance sheet items (y2) and total customer loans (y3) and the first input is fixed assets (x1) followed by total assets (x2) and total borrowed funds (x3). The input prices are w1, price of fixed assets (=Non-interest expenses /Fixed assets); w2, price of labor (=Personnel expenses/Total assets); w3, price of loanable funds (=Interest expenses/Total borrowed funds). The indicators are presented below.
Input(X) X1 Fixed assets X2 Total assets
X3 Deposit
Output(Y) Y1 Total securities
Y2 Off-balance sheet items Y3 Total customer loans
Price(W) W1 Other operating expenses /Fixed assets W2 Personnel expenses/Total assets
W3 Interest expenses/Deposit
Total cost(TC)
TC Total operating expenses
The model I use refers to the study by Weill. It is a translog cost function associated with input cost share equations
i i m i m m i n i i i m i mi i n ni i m n m n nm k n n k nk m j j mj n m n n m m v u t Z t y t w w t Z y Z w w Z y w w w w w w y y w w y W TC
t
i
m m + + + + + + + + + + + + + + + + = ∑
∑
∑
∑
∑
∑∑
∑∑
∑∑
∑
∑
ln ln ) ln( 2 1 ln ln ln ) ln( 2 1 ln ln ) ln( ) ln( ) ( ln 2 1 ln ln 2 1 ) ( ln ln ln 2 3 2 1 0 3 2 1 0 3 3 3 3 0 3)
(
)
(lnZ
δ ζ κ τ τ ι η δ δ γ β α β α βWhere TC is total cost, ym is the mth bank output (m=1,2,3), wn is nth imput price (n=1,2,3), w3 is price of borrowed funds. This equation can be estimated by maximum likelihood method and I refer to the method of Greene(2005) and follow the bank stochastic panel frontier model with time-variant inefficiency. In the equation I use the Lang and Welzel (1996) to scale input prices by the price of borrowed funds. The error term εi concludes two components which are random noise vi and inefficiency item ui. In year t the bank could not get optimal cost because of these two components. In this equation the random noise is assumed to be i.i.d. with vi~N(o,σv2) and it is independent of the explanatory variables. The inefficiency term is also i.i.d with ui~N(0.σu2) and follow the half-normal distribution. In addition ui is independent from vi. Bank-specific efficiency scores are computed using the conditional expectation of ui considering εi. The point estimation of cost efficiency are got from E(ui|εi). Therefore cost efficiency is between 0 and 1 where a bank with cost efficiency of 1 means it is fully efficient. If a bank’s cost efficiency is 0.8, it represents 80% of observed cost could have been sufficient to produce an identical output.
reinstate with time went by. From this point of view the cost efficiency is bank-specific given it is affected by fixed effect.
Considering the above analysis we can illustrate the three hypothesizes by figure 1, FITURE1
As shown in figure 1, assume that the locus of cost efficiency presents an increasing positive trend for all banks over observed time period. Bank A represents the banks not involved in merger and acquisition in the observed period and its cost efficiency is rising continuously. The less efficient Bank B acquired bank C in year t3 and become a new bank D. Then the cost efficiency of Bank D improved fast due to the merger. Considering the M&A deals may have intense impact on the accounting operations for the acquirer and target banks. Consequently the year when the deal happened will not be analyzed. In order to study the cost efficiency changes in pre- and post-M&A period I compare the differences of cost efficiency surrounding two-year of the deal. Because two year before the deal could reflect relatively more exact efficiency conditions for both the acquirer and target banks in pre-M&A period. After one and two year of the deal, the bank may go through adaptive phase and the CE at that time could be more representative.
According to figure 1 the three hypothesizes could be illustrated as
H1: B1>C1 H2: B1<D1 H3:A1<D1
For the comparison test, I choose to use T-test first to estimate the whether the average bank performance indicators is significantly different before and after M&A or between different type of M&A banks. However T-test assumes the mean of the comparison variables are normally distributed. Considering the variables that need to be compared are not really normal distributed (see Figure 3-5). From figure 3 and 4, the distribution of ROAA and ROAE is clearly not normally distributed. In addition from figure 5 the cost efficiency is apparently half-normal distributed. Therefore beside T-test I choose sign test which is nonparametric analysis for testing the equality of matched pair of variables（Snedeco and Cochran, 1989）. This method only requires the comparison pair have the same distribution and not limit to normal distribution. The hypothesizes could be better test in that case.
4. Empirical Result
4.1 Efficiency result
Table 3 summarizes the key variables for estimating cost efficiency. It is clear that the standard deviation of the three output terms is really large. This indicates that banks data in the sample is spread out over a large range of values. There are very large banks with over million Euros total asset while there are still small banks with little assets. Because of the large difference in outputs and inputs among the sample banks, input prices also have great deviations.
term εi in cost function concludes two components which are random noise vi and inefficiency item ui. The inefficiency item Insig2u in table 4 shows statistic significant at 99% confidence interval level. This means the banks in the sample is not fully efficient. We could successfully compute the cost efficiency number and see to what extent could banks produce output using one unit of input.
4.2 Results on M&A effects
4.2.1 Comparison of bank performance by different bank types
and countries
Table 5.1 shows M&A and non-M&A banks performance by different countries. For non-M&A banks Netherlands and Luxembourg generate highest ROAA and ROAE respectively. Denmark gets higher cost efficiency than other banks. Netherlands also have higher ROAA and ROAE among acquirer banks. UK acquirer banks are more efficient than other banks. Target banks from Denmark and Portugal have largest ROAA and ROAE ratios. Netherlands banks operate more efficient than other country’s banks.
4.2.2 Hypothesizes tests
Given the three hypothesizes I firstly examine the banks that involved in merger and acquisition deals. According to table 6.1, the performance of the acquirer is better than target bank before two years of M&A deal. As illustrated in table 6.1 the acquirer bank generate higher ROAA and ROAE ratios. The ROAE of acquirer bank is significantly higher than that of target bank during the observed time period. Considering the comparison of cost efficiency between acquirer and target bank we can clearly see from the table that the acquirer bank is more cost efficient than target bank. However the difference between averages cost efficiency of acquirer and target is not statistically significant. It can be seen that there is little differences between the acquirer and target. This result do not support the first hypothesize and shows that though acquirer bank is more efficient and profitable than target bank before merger and acquisition from the value of ROAA and ROAE , the banks which play different roles during M&A deal show little difference in cost efficient .
one- and two-year of the M&A, Nevertheless both of the cost efficiency differences are not statistically significant.
Last but not least, the test of the third hypothesize could represents the comparison of M&A bank with peer group. Because non-M&A banks take the majority part in banking industry therefore this comparison can test whether banks involved in M&A is more efficient than banks not engaged in the deal. Table 6.3 demonstrates the result of the third hypothesize test and it is clear that the ROAA and ROAE of acquirer bank are not significantly higher than that of non-M&A banks after both one- and two-year of the deal. While for the test of cost efficiency, the average cost efficiency of acquirer bank are apparently lower than that of non-M&A group and the differences are very significant. Therefore the third hypothesize is not fully supported by the data analyze.
In conclusion the empirical result unfortunately do not support hypothesizes 1,2 and 3. From the three hypothesize test we can see that M&A may have impact on bank performance. However whether the impact is good or bad for the bank remains not clear. Given the comparison of M&A bank, the acquirer are the banks with higher ROAE and ROAA ratios and they are relatively more cost efficient. For the acquirer bank, M&A could improve banking operation and increase cost efficiency in post-M&A period. Although acquirer banks also generate higher ROAE and ROAA, they do not show higher cost efficiency ratios when comparing with peer non-M&A bank in the same industry.
5. Conclusion
This paper tends to get insight into the impact of merger and acquisition on bank performance. In order to analyze the effect of merger and acquisition I collect all the M&A deals happened during 2002 and 2008 in EU(15) as the sample to compare both accounting ratios and cost efficiency changes. There are three hypothesizes to be examined. The first two hypothesizes mainly study the M&A banks and last one relate to the comparison of M&A bank with non-M&A banks. The empirical result shows that acquirer bank generate higher ROAE and ROAA ratios and also little more cost efficient than target bank. M&A may lead to little performance improving for acquirer bank. The non-M&A bank’s ROAE and ROAA are lower than acquirer banks while the cost efficiency of bank not involved in M&A are higher than M&A banks. All the three hypothesizes are not significant tested especially for cost efficiency of the banks. This may interpret that M&A do have impact on bank performance but the effect is not significant surrounding 2 years of the deal happens.
6. Acknowledgement
First, I would like to offer my sincerest gratitude to my supervisor, Dr. Micheal Koetter, who always gave me valuable suggestions and advises during the thesis writing period. I also appreciate for my co-assessor Dr. Dirk Bezemer, who would kindly give useful valuation for this paper.
Secondly, I want to express my gratitude to my friends, who company me through the whole steps of thesis writing.
7. References
Aigner, D.J., Lovell, C.A.K., Schmidt, P. (1977), “Formulation and Estimation of Stochastic Frontier Production Function Models”, Journal of Econometrics 6, 21-37.
Altunbas Y, Evans L, Molyneux P (2001), “Bank ownership and efficiency”, J. Money, Credit Bank 33(4), 926–954 (November)
Allen Linda, Cebenoyan, A. Sinan (1990), “Bank Acquisitions and Ownership Structure: Theory and Evidence”, Journal of Banking and Finance 15, 425-48 Allen Linda, Rai Anoop (1996), “Operational Efficiency in Banking: An International
Comparison”, Journal of Banking and Finance 20, pp. 655-72
Altunbas Y, Gambacorta Leonardo, Marques David (2007), “Securitization and the bank lending channel”, European Central Bank, Working Paper Series: 838 AN: 0948805
Amel Dean (2004), “Consolidation and Efficiency in the Financial Sector: A Review of the International Evidence”, Journal of Banking and Finance 28, 2493-1519 Anthony N. Rezitis (2008), “Efficiency and productivity effects of bank mergers:
Evidence from the Greek banking industry”, Economic Modelling 25, 236-254 Bettese, George E, Corra, Greg S (1977), “Estimation of a Production Frontier Model:
With Application to the Pastoral Zone of Eastern Australia”, Australian Journal of Agricultural Economics, 21, 169-79
Berger Allen. N, Humphrey David. B, (1997), “Efficiency of Financial Institutions“, Working Papers -- Financial Institutions Center at The Wharton School, 78
Cavallo, Laura, Rossi, Stefania P.S.(2002), “Do environmental variables affect the performance and technical efficiency of the European banking systems? A parametric analysis using the stochastic frontier approach”, European Journal of Finance 8, pp123-146
Delong, Gayle, Deyong, Robert (2007), “Learning by Observing: Information Spillovers in the Execution and Valuation of Commercial Bank M&As”, Journal of Finance 62, 181-216
Dimitris K. Christopoulos, Sarantis E.G. Lolos, Efthymios G. Tsionas (2001), “Efficiency of the Greek Baking System in View of the EMU: a heteroscedastic stochastic frontier approach”, Journal of Policy Modeling 24: 813-829
Drake Leigh, Weyman-Jones, Thomas. G, “Productive and Allocative Inefficiencies in U.K. Building Societies: A Comparison of Non-parametric and Stochastic Frontier Techniques”, Manchester School of Economic and Social Studies
Eisenbeis Robert A., Ferrier Gary D., Kwan Simon H. (1999), “The Informativeness of Stochastic Frontier and Programming Frontier Efficiency Scores: Cost Efficiency and Other Measures of Bank Holding Company Performance”, Working Paper Series (Federal Reserve Bank of Atlanta) 23, pp1, 40p, 8 Charts, 1 Graph; (AN 2732297) E.Fiorentino, A.De Vincenzo, F.Heid, A.Karmann, M.Koetter (2009), “The effects of
privatization and consolidation on bank productivity: comparative evidence from Italy and Germany”, Bance D’Italia Eurosistema 722
Elena Beccalli, Pascal Frantz (2008), “M&A operation and performance in banking”, Journal of Financial Service Research 36(2-3), 203-226
Farrell, M. J. (1957), “The measurement of productive efficiency”, Journal of the Royal Statistical Society 120, 253–281
Ferrier, G. D., Lovell (1990), “Measuring cost efficiency in banking”, Journal of Econometrics 46, 229–245
Gary D.Ferrier, C.A.Knox Lovell (1990), “Measuring Cost Efficiency in Banking Econometric and Linear Programming Evidence”, Journal of Econometrics 46, 229-245
Gunter Lang, Peter Welzel (1996), “Efficiency and Technical Progress in Banking Empirical Results for a Panel of German Cooperative Banks”, Journal of Banking & Finance 20, 1003-1023
325-332
Greene, William (2005), “Reconsidering Heterogeneity in Panel Data Estimators of the Stochastic Frontier Model”, Journal of Econometrics 126, 269-303
Gunter Lang, Peter Welzel (1999), “Mergers among German Cooperative Banks: A Panel-based Stochastic Frontier Analysis”, Small business Economies 13, 273-286 Guest, Paul (2007), “The Impact of Mergers and Acquisitions on Executive Pay in the
United Kingdom”, ESRC Centre for Business Research working paper, AN: 0938900
Gintschel, Andreas, Hackethal, Andreas (2004), “Multi-bank loan pool contracts: enhancing the profitability of small commercial banks”, Applied Financial Economics 17, 1239-1252
Hannan Timothy H, Wolken John D (1989), “Returns to Bidders and Targets in the Acquisition Process: Evidence from the Banking Industry”, Journal of Financial Services Research 3, 5-16
Houston .Joel. F, Ryngaert Micheal D (1994), “The Overall Gains from Large Bank Mergers”, Journal of Banking and Finance 18, 1155-76
Humphey David. B, Pulley Lawrence. B (1997), “Banks' Responses to Deregulation: Profits, Technology, and Efficiency.”, Journal of Money, Credit & Banking 29, 73-93
H.P.Huizinga, J.H.M.Nelissen, R.Vander Vennet (2001), “Efficiency Effects of Bank Mergers and Acquisitions in Europe”, Faculteit Economie EN Bederijfskunde J.W.B.Bos, M.Koetter, J.W.Kolari, C.J.M.Kool (2009), “Effects of Heterogeneity on
Bank Efficiency Score”, European Journal of Operational Research 195, 251-261 Kumbhakar, Subal C., Lovell, C. A. Knox (2000), “Stochastic Frontier Analysis”,
Cambridge; New York and Melbourne: Cambridge University Press AN: 0535786 Laura Cavallo, Stefania P.S. Rossi (2002), “Do environmental variables affect the
performance and technical efficiency of the European banking systems? A parametric analysis using the stochastic frontier approach”, The European Journal of Finance 8, 123-146
Comparison”, Journal of Banking & Finance 20, 655-672
Laurent Weill (2004), “Measuring Cost Efficiency in European Banking A comparison of Frontier Techniques” Journal of Productivity Analysis 21(2), 133-152
Leibenstein Harvey (1966), “Allocative Efficiency vs. “x-efficiency””, American Economic Review 56, 392
Meeusen, W., Van Den Broeck (1977), “Technical Efficiency and Dimension of the Firm: Some Results on the Use of Frontier Production Functions”, Empirical Economics 2, 109-122
Meon, Pierre-Guillaume, Weill, Laurent (2005), “Can Mergers in Europe Help Banks Hedge against Macroeconomic Risk?”, Applied Financial Economics 15, 315-26 Micheal Koetter (2008), “An Assessment of Bank Merger Success in Germany”,
German Economic Review 9(2), 232-264
Micheal Koetter (2006), “Measurement Matters-Alternative Input Price Proxies for Bank Efficiency Analyses”, Journal of Financial Service Research 30, 199-227 Micheal Koetter (2005), “Evaluating the German Bank Merger Wave”, Tjalling C.
Koopmanns Discussion Paper No. 05-16
M.Koetter, J.W.B.Bos, F.Heid, J.W.Kolari, C.J.M.Kool, D.Porath (2007), “Accounting for distress in bank mergers”, Journal of Banking & Finance 31, 3200-3217
Peristiani. S (1996), “Do mergers improve the x-efficiency and scale efficiency of U.S. banks?: Evidence from the 1980s”, Federal Reserve Bank of New York, Research paper: 9623 AN: 0702774
Pillioff. Steven.J (1996), “Performance Changes and Shareholder Wealth Creation Associated with Mergers of Publicly Traded Banking Institutions”, Journal of Money, Credit & Banking 28, 294-310
Rezitis, Anthony N (2006), “Productivity Growth in the Greek Banking Industry: A Non-parametric Approach”, Journal of Applied Economics 9, 119-138
Sheldon George (1994), “Nichtparmetrische Messung des technischen Fortschritts im Schweizer Bankensektor.”, Schweizerische Zeitschrift fur Volkswirtschaft und Statistik/Swiss Journal of Economics and Statistics 130, 691-707
Vander Vennet, R. (1996), The Effect of Mergers and Acquisitions on the Efficiency and Profitability of EC Credit Institutions, Journal of Banking and Finance 20, 1531-1558.
Viverita Brown Kym, Skully Micheal (2007), “Efficiency Analysis of Islamic Banks in Africa, Asia and the Middle East”, Review of Islamic Economics 11, 5-16
Waddock, Sandra A., Bodwell, Charles, B.Graves, Samuel (2002), “Responsibility: The new business imperative”, Academy of Management Executive 16, 132-148 Willian Greene (2005), “Reconsidering heterogeneity in panel data estimators of the
stochastic frontier model”, Journal of Econometrics 126, 269-303
8. Appendix
Table 1 Banks number changes and M&A
Number of bank M&A
Country N1999 N2009 ∆N Growth rate N1999 N2009 ∆N Growth rate
Austira 878 790 22 2.51 12 27 15 125.00 Belgium 117 104 -13 11.11 24 19 -5 -20.83 Denmark 210 164 -56 -26.67 11 41 30 272.73 Germany 2996 1948 -1048 -34.98 76 54 -22 -28.95 Greece 57 66 -9 -15.79 18 33 16 88.89 Finland 345 349 4 1.16 4 13 9 225.00 France 1163 712 -451 -38.78 25 32 -7 -28.00 Ireland 80 498 418 522.50 6 14 8 13.33 Italy 894 801 -93 -10.40 15 21 6 40.00 Luxembourg 209 147 -62 -29.67 3 8 5 166.67 Netherlands 615 295 -320 -52.03 24 28 4 16.67 Portugal 223 166 -57 -25.56 0 2 2 200.00 Sweden 149 180 31 20.81 17 21 5 29.41 Spain 386 352 -34 -8.81 10 8 -2 -20.00 United Kingdom 502 389 -113 -22.51 40 62 22 55.00 Total 8824 6961 -1781 285 383 86 Average -16.40 75.66
Sources: ECB (2000, 2009) and Zephyr
Table 2: number of M&A and non-M&A banks (by country and by M&A year)
Table 3 Summarize of variables for cost efficiency
Variable Mean Std. Dev. Min Max
y11 2,320.69 27,650.83 0.03 1,637,583.00 y22 3,003.42 39,415.38 0.02 3,132,690.00 y33 6,433.61 60,895.40 0.02 3,172,226.00 w14 8.38 65.61 0.02 4,354.00 w25 0.02 0.02 3.0300E-07 0.62 w36 12.11 367.03 5.8600E-04 31,940.00 h57 504.88 4,408.91 0.01 204,870.60 Observations:22.405
Table 4 Parameters and descriptive statistics for cost efficiency Dependent variables lnTC/w3 Dependent variables lnTC/w3 ln(w1/w3) -0.0348*** lny3 0.910*** (-0.012) (-0.0207) ln(w2/w3) 1.170*** lnz 0.0347 (-0.014) (-0.0237) lny1lny1 0.0846*** lnzlnz 0.186*** (-0.00129) (-0.00762) lny2lny2 0.0142*** ln(w1/w3)lnz 0.0351*** (-0.00182) (-0.00307) lny3lny3 0.200*** ln(w2/w3)lnz -0.0300*** (-0.0036) (-0.00312) lny1lny2 -0.00301** lny1lnz -0.00915*** (-0.0013) (-0.00279) lny1lny3 -0.0509*** lny2lnz 0.0187*** (-0.00237) (-0.00296) lny2lny3 -0.0273*** lny3lnz -0.168*** (-0.00242) (-0.00481) ln(w1/w3)2 0.0190*** t2 0.0309*** (-0.00187) (-0.00111) ln(w2/w3)2 0.0692*** Ln(w1/w3)t -0.00305*** (-0.00207) (-0.00101) ln(w1/w3)ln(w2/w3) -0.0554*** Ln(w2/w3)t 0.00191* (-0.00181) (-0.00107) ln(w1/w3)lny1 0.0141*** lny1t 0.00763*** (-0.0011) (-0.00103)
ln (w1/w3)lny2 4.09E-05 lny2t 0.00427***
Table 5 Comparison of bank performance by country and type Table 5.1 Bank performance by country
Non-M&A Acquirer Target
Country ROAA ROAE CE ROAA ROAE CE ROAA ROAE CE
Austria 0.427 5.237 0.609 0.614 10.255 0.495 0.688 7.343 0.584 Belgium 0.951 6.977 0.603 0.663 11.282 0.661 0.415 7.979 0.669 Denmark 1.087 7.355 0.831 1.021 4.336 0.717 1.646 11.317 0.899 Germany 0.281 4.287 0.734 0.163 2.927 0.720 0.030 0.890 0.690 Greece 0.719 6.595 0.610 0.770 7.353 0.735 0.265 3.341 0.718 Finland -0.285 0.190 0.613 0.812 14.596 0.681 0.898 17.983 0.740 France 0.963 9.532 0.704 0.536 8.707 0.566 0.411 7.322 0.664
Ireland 0.242 -3.658 0.323 N.A. N.A. N.A. N.A. N.A. N.A.
Italy 0.770 6.957 0.729 0.716 7.405 0.704 0.385 5.722 0.704 LU1 0.763 11.752 0.326 0.822 21.381 0.291 0.483 8.534 0.223 NL2 1.678 10.262 0.508 8.700 21.887 0.254 7.577 10.030 0.904 Portugal 0.178 3.125 0.523 0.673 9.906 0.549 0.651 12.690 0.488 Sweden 1.045 7.351 0.740 0.966 14.884 0.795 1.109 9.681 0.783 Spain 0.741 9.382 0.657 1.003 12.982 0.684 -0.188 -0.035 0.736 UK3 1.307 8.894 0.523 1.569 13.060 0.865 0.967 12.283 0.432 1 Luxembourg 2 Netherlands 3 United Kingdom
Table 5.2 Different types of banks performance
ma Type Obs ROAA Obs ROAE Obs CE
Mean Std Dev. Mean Std Dev. Mean Std Dev. 0 1 7880 0.9374 3.8981 7863 8.2445 20.2921 2658 0.6016 0.2474 0 2 15961 0.4330 0.5701 15961 5.4485 11.0037 12249 0.7266 0.1087 0 3 7615 0.3357 0.9579 7614 4.1042 7.0470 5500 0.7271 0.1022 1 1 1047 1.1125 3.8663 1047 9.3606 26.6070 810 0.6254 0.2250 1 2 438 0.4863 0.4600 438 7.0516 7.4054 401 0.7159 0.1386 1 3 312 0.6065 0.7818 312 7.8739 6.8895 306 0.7124 0.1113 2 1 456 0.7412 3.8987 456 6.4793 18.9519 309 0.6413 0.2067 2 2 159 0.3355 0.6979 159 4.0267 10.3460 149 0.7374 0.1383 2 3 143 0.4336 0.7819 143 2.5829 22.0541 140 0.6904 0.1294
Type 1 2 3 indicate commercial, cooperative and saving banks respectively
Table 6 Hypothesize test
Table 6.1 H1: The average cost efficiency and ROAE and ROAA of acquirer bank in EU(15) is higher than that of target banks before 2 years of merger and acquisition.
ma t ROAA ROAE CE Mean Std Dev. Mean Std Dev. Mean Std Dev. 1 B2 0.5101 0.1740 8.3750 2.1446 0.6795 0.0610 2 B2 0.4077 0.1785 6.0983 2.8943 0.6700 0.0746 Sign-test P-value 0.0078 0.0078 0.7734 T-test 0.1025*** 0.0685 2.2767*** 1.3569 0.0095 0.0303 1 B1 0.5632 0.1847 8.4305 2.3847 0.6675 0.0438 2 B1 0.3190 0.3557 4.5241 3.7559 0.6654 0.0330 Sign-test P-value 0.0078 0.0078 0.7734 T-test 0.2442*** 0.2405 3.9064*** 2.0188 0.0021 0.0193
Table 6.2 H2: The average cost efficiency and ROAE and ROAA of the acquirer bank after 2 years of the merger and acquisition in EU(15) is improved compared with that before 2 years of merger and acquisition
ma t ROAA ROAE CE Mean Std Dev. Mean Std Dev. Mean Std Dev. 1 B2 0.4359 0.1321 7.2820 1.2928 0.6659 0.0222 1 A2 0.4986 0.2479 7.7219 4.3588 0.6776 0.0738 Sign-test P-value 0.5000 0.5000 0.1875 T-test -0.0627 0.3648 -0.4399 5.3051 0.0117 0.0574 1 B1 0.5632 0.1847 8.4305 2.3847 0.6654 0.0330 1 A1 0.4747 0.2548 2.1147 4.9405 0.6654 0.0220 Sign-test P-value 0.7734 0.7734 0.7734 T-test 0.0885 0.3463 2.1147 6.3042 0.0000 0.0275
Table 6.3 H3: The average cost efficiency and ROAE and ROAA of acquirer banks involved in merger and acquisition after 2 years of merger and acquisition in EU(15) is higher than that of the banks not involved in merger and acquisition.
0 A2 0.4454 0.0727 5.9683 1.2129 0.7101 0.0080
Sign-test P-value 0.5000 0.5000 0.1875
T-test 0.0532 0.2697 1.7536 3.1639 0.0441***
0.0228
Figure 1
Change trends of No.of credit institution
0 500 1000 1500 2000 2500 3000 3500 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year N o .of cre d it inst i tution Austria Belgium Germany Denmark Spain Finland France United Kingdom Greece Ireland Italy Luxembourg Netherlands Portugal Sweden Figure 2
Change trend of EU(15) M&A
Figure 3 Distribution of ROAA -100 -50 0 50 1 00 R O AA 0 10000 20000 30000 Frequency
Figure 4 Distribution of ROAE
