Exploring the Success of Regional Bank Mergers: Financial Versus Non-Financial Performance

2.1 Effects of Mergers on Bank Performance

M&As’ success in the banking industry has been of academic interest for decades. Many empirical results have been generated since and provide mixed evidence regarding success and determinants of success (see the reviews by Amel et al. 2004; DeYoung, Evanoff, and Molyneux 2009; Kolaric and Schiereck 2014). Approaching these questions, the first challenge is to define ‘M&A performance’ (Zollo and Meier 2008, 55), revealing a wide range of methods. Kolaric and Schiereck (2014), for example, categorize the most prominent ways to measure success into analyses of shareholder reactions (event studies of stock returns), bank efficiency (preferrably frontier methods), and operating (accounting) performance (performance studies).

Due to the regulatory and structural differences between the U.S. and European banking markets, we focus on the success of European bank mergers. The corresponding results are inconclusive and highly sensitive to the time period under investigation (Kolaric and Schiereck 2014) and the geographical focus of the study. A pan-European sample of bank mergers has, for example, been analyzed by Hagendorff and Keasey (2009), who find small performance gains in the post-merger period, and provide evidence that banks follow a cost-cutting strategy in the years following a merger. Many of the respective (cross-country) studies, such as Ismail, Davidson, and Frank (2009), Lozano-Vivas et al. (2011), and Galariotis et al. (2021), cover only large, mostly listed banks, which can hardly be compared to the stakeholder-oriented regional banks in this study.

Merger analyses of individual countries with a significant share of stakeholder-oriented banks are more directly relevant. For example, for savings banks in Spain, Bernad, Fuentelsaz, and Gómez (2013) report significant post-merger profitability increases. More recently, Coccorese and Ferri (2020) found no average efficiency gains from the mergers of Italian cooperative banks; only banks merging for the third time (at least) were reported to be able to improve efficiency. For the German market, the results tend to be neutral and potentially positive. For a sample of cooperative banks, Lang and Welzel (1999) observed efficiency gains if (but only if) branches closed in the wake of the merger. Koetter (2008) uses profit and cost efficiency estimates of German savings and cooperative banks. The results indicate that every second merger is a success, whereby, “on average, profit efficiency improvements are larger than those of cost efficiency” (Koetter 2008, 258). Auerbach (2009) employs a dynamic panel regression methodology to examine the impact of bank mergers on savings and cooperative banks. Focussing on cost-income ratios and profitability to assess post-merger effects, the author concludes that, on average, the cost-income ratio is higher for merging banks in the year of the merger than for non-merging banks. However, this ratio gradually improves over time, and by the third and fourth years following the merger, the average cost-income ratio for merging banks surpasses that for non-merging banks. Regarding mergers’ profitability effects, Auerbach (2009) reports inconclusive results. Behr and Heid (2011) employ a matching strategy and find a neutral, potentially positive effect of German bank mergers on profitability (return on assets) and cost efficiency (cost-income ratio).

Finally, the Austrian market was analyzed by Hahn (2007), who reported significant gains in productive efficiency for mergers between 1996 and 2002. Similarly, the results of Egger and Hahn (2010) indicate the positive effects of bank mergers on banking performance, especially cost performance.

Empirical findings on the factors associated with merger success are also diverse. Following Kolaric and Schiereck (2014), the determinants can be categorized into the characteristics of partner companies, transaction characteristics, and environmental factors. Koetter (2008) observes that larger differences in pre-merger efficiency indicate more successful mergers. The relatively higher efficiency of the acquirer signals that management skills and cost-saving potential are more easily identified if the target’s position is relatively worse ex ante. Likewise, Auerbach (2009) proposes that mergers with small banks with weak financial profiles are particularly successful. Kolaric and Schiereck (2014) also list merger experience among success-driving factors, which is somewhat confirmed by Coccorese and Ferri (2020). Lang and Welzel (1999), as mentioned above, argue that, in their sample, efficiency gains only emerge if branches are closed. Many studies, such as Drees, Keisers, and Schiereck (2006), Braun, Parise, and von Nitzsch (2013), and Gindele et al. (2019), have highlighted the importance of ‘soft’ factors for mergers to be successful. In practice, mergers often suffer from a lack of clear communication and strategy or an underappreciation of necessary efforts and costs, such as technical and personnel integration or customer retention.

The variation in the results of merger success might be due to country-specific differences and changes over time, but methodological problems also play a role. Merging banks non-randomly self-select into treatment for (perhaps unobserved, strategic) reasons, and not accounting for this, for example, by comparing with the average non-merging bank, may lead to simultaneity bias in the results (Egger and Hahn 2010). Interestingly, only a few studies have used matching methods, such as Behr and Heid (2011) and Egger and Hahn (2010). In addition, Yu (2019) used propensity score matching (PSM) to examine more than 2000 credit union mergers in the USA, with no observable improvement in profitability and cost efficiency. Additionally, Coccorese and Ferri (2020) checked their results for robustness using PSM.

The mixed results could also be explained by the different motives for bank mergers. The common opinion is that merging banks strive for efficiency gains and market share, broaden their customer base, enhance revenue, and diversify or alter their risk profiles (Huhtilainen, Saastamoinen, and Suhonen 2022). In reality, other motives may dominate, especially for small non-listed banks in regional markets. These banks may consider a merger as a possible solution for management succession, to exit declining markets and avoid financial problems (Elsas 2007), or to avoid necessary (pending) investments in compliance systems, IT, and workforce training (Braun, Parise, and von Nitzsch 2013; Dunaway 2023; Thompson 2019). This pressure drives banks with lacking or exhausted internal optimization potential to seek consolidation (Braun, Parise, and von Nitzsch 2013), and smaller inefficient units with modest growth prospects merge into larger banks that operate in more favorable business environments (Huhtilainen, Saastamoinen, and Suhonen 2022). However, obtaining hands-on evidence is difficult. Kjellman, Tainio, and Kangas (2014), for example, propose that a loss of confidence of Finnish bank managers in being able to successfully run the bank in the future may have led them to seek a merger. Additionally, a survey of 29 managers of German cooperative banks by Gindele et al. (2019) provided valuable insights. Manager retirement (often without replacement) was the main motive or trigger, which was stated about twice as often as the economic situation. Long-term sustainability is the main strategic goal that mergers should support, increased productivity comes second, and synergies rank seventh (only stated twice). Additionally, the institutional environment must be considered. In Austria, bank mergers are not possible across sectors (or ‘pillars’), and, due to regional principles applied, most mergers are among locally near banks (as in Germany, see Braun, Parise, and von Nitzsch 2013). The performance potential of bank mergers that are defensive and restricted in nature may not be huge, which might explain some common empirical findings.

The paper at hand focusses on stakeholder-oriented banks, namely credit cooperatives and savings banks, which constitute an important and integral part of European banking systems (Kalmi 2017). Unlike shareholder-oriented banks, which focus on maximizing shareholder interests, stakeholder-oriented banks have broader business objectives and balance the interests of a wider group of stakeholders (including employees, customers, the regional economy and society). For savings banks and credit cooperatives, improving non-financial objectives (such as providing local access to financing or community engagement) may represent an additional goal of a merger. Furthermore, because bank performance should be evaluated against pursued objectives (Ahn and Le 2014; Brown 2006), measures of merger success should also consider the consequences for stakeholders.

2.2 Consequences of Bank Mergers for Stakeholders

As diverse study findings are on performance effects, DeYoung, Evanoff, and Molyneux (2009) argue that it seems equally difficult to reach a conclusion on whether customers benefit from bank mergers or on how consolidation affects societal risks. They also report growing evidence that mergers may adversely affect a variety of borrowers, depositors, and other external stakeholders (DeYoung, Evanoff, and Molyneux 2009). Mergers are often feared to result in the possible loss of (former) banks’ regional ties (Drees, Keisers, and Schiereck 2006). Coccorese and Ferri (2020) argue that increased consolidation might harm borrowers that are more likely served by small banks, with potential negative impact on regional development and equality. Such a development may violate (cooperative) banks’ ethics and the mission of favoring the financial inclusion of marginal customers and countering the credit rationing of small businesses (Coccorese and Ferri 2020).

Few studies have dealt with stakeholder perceptions of mergers. In telephone interviews, Urban and Pratt (2000) find that U.S. consumers’ expectations of service quality differ according to several socio-economic variables^[1] and that there is much uncertainty about prospective impacts before mergers, which must be addressed through appropriate communication. According to Mylonakis (2006b), bank employees in Greece consider mergers a threat to their jobs and benefits, which is confirmed to some extent by Mylonakis (2006a). Álvarez-González and Otero-Neira (2020) polled bank employees in Spain about their opinion on merger consequences for customers to infer about effects on bank-client relationships. They report improvments in the products and services offered but also that bank mergers negatively affect prices and branch proximity. Álvarez-González and Otero-Neira (2023) discuss results from a customer survey on factors being decisive for their loyalty to the merged bank. Managers of credit cooperatives interviewed by Gindele et al. (2019) reported that customers feared branch closures in the wake of a merger, while municipal officials would have to face the associated tax revenue losses. Furthermore, a deterioration in close relationships with the bank is surmised for the cooperative members, customers, and employees.

Most empirical studies on the post-merger consequences for stakeholders focus on the impact on small business lending (SBL), starting with the presumption that small local banks (still) may have some competitive advantage in soft information collection and formation of strong relationships (Berger, Bouwman, and Kim 2017; Mkhaiber and Werner 2021; Nguyen and Barth 2020). However, increased regulatory costs and technological burdens impair small banks’ ability to maintain SBL levels (Hughes et al. 2019). Mergers might negatively affect bank lending to local small and medium-sized enterprises through branch closures or centralized decision making. The latter possibility spurs fear that mergers might disrupt relationship lending, for example, when large banks acquire community banks, as argued by Jagtiani, Kotliar, and Maingi (2016). However, their results did not confirm this perception or common sense. Minton, Taboada, and Williamson (2021) even find relatively more small business lending originations (in counties where the target bank is located) when acquirers are small and from the same (U.S.) state as the target. Avery and Samolyk (2004) also report a positive effect of (U.S.) community banks’ consolidation activity on SBL.

However, other studies have reported a decline in SBL after consolidation activity that is partly offset by other banks or non-bank institutions (Berger et al. 1998; Bonaccorsi di Patti and Gobbi 2007; Craig and Hardee 2007).^[2] Empirical studies often provide evidence that mergers negatively affect the ties between banks and their customers (Allen, Damar, and Martinez-Miera 2016) because relationships, especially with the borrowers of target banks, may be disrupted or terminated (Degryse, Masschelein, and Mitchell 2011; Karceski, Ongena, and Smith 2005). According to Jagtiani, Maingi, and Dolson (2022), SBL is often redirected towards the acquirer’s areas of operation (counties), particularly in mergers involving banks with non-overlapping networks. Credit gaps in counties where only the target was present before the merger may have resulted from this, and Jagtiani, Maingi, and Dolson (2022) find no evidence of other banks making up for these gaps.

If passed on, the merger-induced cost savings may lead to lower lending rates. For commercial bank mergers in the USA, Erel (2011) reports such an effect, which is the more reduced the higher the increase in market power for the consolidated entity is. Barros et al. (2014) also find more favorable loan rates after bank mergers, whereas Yu (2019) does not. For a sample of corporate customers of savings banks whose mergers were forced by county reforms in Germany, Koetter et al. (2018) report reductions in the cost of borrowing, leading to more investment and employment in these firms. Dunaway (2023), on the other hand, argues that the trend of U.S. community banks merging into larger entities has mostly harmed consumers by increased pricing and reduced credit supply.

Bank mergers may also negatively affect financial access and inclusion by amplifying branch closures. Mergers reduce branch density in the U.S., especially in rural areas (Calzada, Fageda, and Martínez-Santos 2023), with the potential to negatively affect access to financing, as suggested by Nguyen (2019) and Jagtiani, Maingi, and Dolson (2022).

Bank customers may also experience changes in their deposit rates after bank mergers. Lower rates of deposits in U.S. markets where significant (local) mergers have taken place have been reported, for example, by Prager and Hannan (1998). Dinger (2015) confirms mostly negative effects of consolidation for U.S. depositors. Bord (2018) argues that with banks becoming larger through mergers, fees and minimum required balances increase, driving off especially low-income households.

After consolidation, regional banks may also alter their risk attitudes and behaviors, with potential effects on financial stability. However, studies of these effects at the banking level are scarce. Knapp and Gart (2014) report that the variability of loan charge-offs and rates of non-performing loans of U.S. bank holding companies significantly increase after mergers, due to changes in the loan mix. Yu (2019) also reports a post-merger shift towards more risky loan categories for U.S. community banks, leading to increased credit risk. For German savings banks forced to merge due to county reforms, Koetter et al. (2018) observe increased risk-taking (lower capital, more non-performing loans, and fewer loan loss provisions).

Literature on merger-related reductions in personnel is also scant, with the exception of Mylonakis (2006a), who reports significant job losses in merged or acquired institutions in the Greek banking market. Sherman and Rupert (2006) argue that merging banks may keep personnel necessary for customer retention. If additional personnel are required during the integration process, employment may increase temporarily. The emergence of merger-induced workforce expansion is confirmed for the European sample of Hagendorff and Keasey (2009).

Academic assessments of whether consolidation changes banks’ local support activities through sponsorship and donations are, to the best of our knowledge, nonexistent.

Francis, Hasan, and Wang (2008) report a positive effect of bank mergers on new business formation, although branch closures (confirmed to occur increasingly after mergers) are generally presumed to impair start-up activity (see, e.g. Ho and Berggren 2020, or Prieger 2023). Firms in the acquirer’s region are found to reduce their investments if the respective banks experience a distressed merger (Dinger, Schmidt, and Theissen 2021).

4.1 Basic Results on (Average) Merger Success

Table 3 reports the results of the average development of the five variables of interest, from the pre-merger year to three years after consolidation (as also in Behr and Heid 2011). The first panel shows basic evolution. For the second panel (naive comparison), we deduct the mean of the respective variables from the control group (in the respective year). The third panel presents the matching procedure outcomes. The term ‘merger success’ is used for the average change in banks’ performance variables from the pre-merger year to the end of the evaluation period.

The immediate negative consequences of the average merger (in the consolidation year) can be observed from all variables in (the naive) comparison with never-merging banks. An initial worsening of profitability, cost-income ratio, and loan growth is also observable in the results of the matching procedure. Until the end of the examined post-merger period, however, profitability and efficiency improved significantly in the naive comparison to the control group. Yet, from matching, we infer that these effects are much smaller and statistically insignificant. This also applies to the estimated final merger success after three post-merger years, except for the (weakly significant) improvement in social efficiency.

Although a merger-induced increase in social efficiency (SE), in principle, would represent a desired effect, it pays to take a closer look at the evolution of individual components. SE seems to improve mainly owing to decreases in operating costs overcompensating increased branch closings (which represent a reduction in the proposed benefits). Even though we are not able to evaluate the qualitative nature of these changes – personnel reduction may be ‘only’ attrition, and other cost reductions do not necessarily negatively affect the service quality or local suppliers, caution should be exercised when interpreting an increase in social efficiency as an improvement in regional stakeholder welfare.

Thus, there is no significant average improvement in profitability and efficiency in the post-merger period when selection is considered. Nevertheless, the reduction in the cost-income ratio seems non-negligible (there seems to be some success in cutting costs, on average). Regional bank mergers in Austria appear to have no negative consequences for bank loan growth. However, there is huge variation in the success of mergers. Although the average consolidation does not lead to a significant improvement in profitability, there are many successful mergers. After investigating the robustness issues, we seek to uncover the factors connected to the diversity of merger outcomes in Section 4.3.

4.2 Further Analysis and Robustness of Results on Average Merger Success

Using the same procedure, we examine which strategic variables are altered after an average merger. The (non-exhaustive) list of associated indicators evaluated contains the number of bank offices, ratios from the balance sheet and income statement, market power, average interest rates in the form of interest income (expenses) as a fraction of interest-earning (interest-bearing) assets (liabilites), ratios of personnel and administrative expenses, interest income, and income from fees and commissions (all relative to total assets) and, finally, the ratio of loan revaluations (including direct write-offs). After balancing merging and non-merging banks on the same variables as before, two significant differences emerged. First, consolidated banks in our sample close more branches than the control group (up to three years following the merger); second, they reduce the administrative expense ratio more strongly. Corresponding results can be found in Table A.5 in the Appendix.

However, these activities do not seem sufficient to improve profitability significantly (see the results above). While this seems to confirm the focus on cost reductions (as in Hagendorff and Keasey 2009), post-merger enhancements in income generation seem less pronounced. In addition, we did not observe significantly stronger reductions in personnel expenses or increases in (exercised) market power compared with the control group. In summary, the consequences affecting stakeholders are ambiguous. On the one hand, the welfare-reducing effects related to personnel and market power do not seem to be strongly forced by merging banks, at least in the short and medium run. There are also no adverse effects on loan growth, but relatively more branches are closed, and administrative expenses are cut, which may potentially hurt local suppliers.

Two exercises were conducted to test the robustness of the results.^[10] First, the respective outcome variable is used to balance the treatment and control groups in the pre-merger year. Behr and Heid (2011), for example, argue that assignment to the group of merging banks is not independent of merger outcomes if the average merging bank underperforms, leading to biased estimates. While further reducing selection problems, all differences in, for example, pre-merger profitability or efficiency, are levelled out. For all variables evaluated, the results of merger success do not seem to be qualitatively affected. No significant changes in dependent variables are observable on average three years after the merger also in this setting, as also the effect on social efficiency loses the weak significance it had.

Second, we measure whether the results change when considering a five-year evaluation period for merger cases observable until 2016. Sherman and Rupert (2006), for example, argue that merger benefits may not be realized even after four years if integration and cost-saving procedures are delayed. However, the number of mergers that can be traced for such a long period reduces to 91 (out of the 99 cases observable for the truncated period, eight merged entities are taken over during the post-merger term). With this reduced sample but a longer evaluation period, the estimates of matched merger success increase, with loan growth and financial efficiency as exceptions. The average matched merger success in the form of ROA increases to 0.135, which is not significantly different from zero. Social efficiency is 0.042 higher after five years, the CIR is 2.901 points lower, and the latter effect becomes significant at the 10 % level (if the control group is not balanced on the prior CIR). Thus, certain cost savings, as expected, seem to be realized only after some time.

4.3 Factors Associated with Merger Success

Several factors, practices, and strategies may explain why some regional banks improve strongly after a merger, whereas others stagnate. The matching procedure considers the characteristics (and environmental factors) of the pre-merger year; however, these bank attributes may change with and after consolidation. Thus, we sought to explore the information from our dataset to find out which attributes are correlated (most) with merger success.^[11] Environmental data was not used at this stage because of its expected relative constancy.

Figure 1 depicts merger success in all five variables of interest (their change three years after consolidation compared to the pre-merger level) graphically. Out of 121 cases of consolidation that are still present three years later, rates of improvement range from approximately half of the banks (with loan growth) to 65 percent (with respect to the cost-income ratio).

Figure 1:

Distribution of changes in the ROA (top left), the CIR (top right), loan growth (mid left), financial efficiency (mid right), and social efficiency (bottom) from the pre-merger year to three years after the merger.

The merger success variables are then regressed (including time dummies) on a uniform set of factors consisting of the differences in the characteristics of the acquirers and targets in the pre-merger year (those not applied in the matching procedure), and for the consolidated entity, changes in these factors during the merger period. Personnel and other administrative expenses are entered in the form of their percentage ratio of total assets (%), and value adjustments are relative to total loans to customers. A merger experience dummy is also applied, which takes the value of 1 if the bank (acquirer only) had a merger before (but after 2000 only). Furthermore, there are dummies for mergers in the Raiffeisen sector, for multi-bank mergers (consolidations with two or more acquired banks) and merging banks targeted later in the examination period (more than three years after the evaluated merger).

For each performance variable evaluated, its (consolidated) pre-merger level is statistically significant. For example, the worse the CIR in the year before the merger, the more it improves during the evaluation period. An influental ex-ante difference in the dependent variable between acquirers and targets appears only for the CIR, with banks struggling to improve the CIR if targets were much more inefficient than the acquirer before the merger. This seems to contradict the result of Koetter (2008) or may suggest that there is a threshold that indicates problems with the target that are not easily resolved. The dummy for later takeover is significantly related to the change in the CIR, suggesting that those taken over are banks that are unsuccessful in decreasing the CIR in the post-merger period. Merging banks in the Raiffeisen sector experience higher post-merger loan growth than do consolidated savings banks or Volksbank credit cooperatives. Merger experience positively affects both efficiency variables.

Other ex-ante differences between acquirers and targets (combined targets, if there are two or more) are hardly significantly connected to the dependent variables’ post-merger development. For example, the higher the difference in the administrative expense ratio, the lower the success of CIR reduction, which might hint at poorer results emerging with inefficienct acquirers. However, cost-income ratios appear to be increasingly pushed downwards if the market power of the acquirer is higher than that of the targeted bank or banks. Contrary to expectations, the differences in size were never significant.

Table 4 shows how the changes in potential strategic variables and merger outcomes (both from the pre-merger year to three years after the merger) are interrelated, Table A.6 in the Appendix reports the underlying regression results. Some of these interrelations occur naturally, such as higher post-merger loan growth being associated with an increasing loan ratio in the balance sheet, or that increasing interest income ratios indicate improvements in financial efficiency (given that noninterest income is one of the outputs applied). From these results, it can be reasoned that bank mergers with improved profitability are those that succeeded in reducing administrative costs more strongly. However, personnel expenses and market power seem to be levers for improvements in the cost-income ratio. At least in the sample at hand, banks that increasingly reduced personnel expenses or obtained relatively more pricing power also experienced greater improvements in the CIR. Post-merger improvements in financial efficiency are also related to personnel expenses, and, as argued before, expense cuts and branch closures affect social efficiency outcomes. In addition, more pronounced increases in market power seem to indicate adverse social efficiency developments.

Developments in bank equity ratios have also been examined, although it is not entirely clear whether these are strategic adjustments. Nevertheless, strong drops in the equity ratio are observable for some institutions in the sample (up to 3pp), and this group of banks appears to drive the connection between equity ratio changes and post-merger profitability, CIR, and loan growth (see Table 4). The combined size of these institutions is rather low, and they start with a rather high capitalization level (so there is no need to further improve it by merging) and good profitability. They seem to pursue a credit expansion strategy (as loan growth over the merger evaluation period is much larger than that for the remaining merging banks) and are successful in cutting costs (bringing the CIR down). By contrast, profitability remains low and significantly behind pre-merger levels after the three post-merger years. Further examination of the characteristics of merging banks with falling equity ratios reveals no slump in operating income. Delayed value adjustments to the credit portfolio (with a peak in the second post-merger year) and high write-downs of other financial assets (bonds, stocks, participations, and shares in affiliated enterprises) reduce the evolution of the ROA. Thus, there seems to be a cluster of banks that potentially drives some of our main results, which are expanding the traditional loan business (while coping with charge-offs, especially in financial assets), and ambitiously downscaling administrative costs.