University of Huddersfield Repository The impact of M&A on the Nigerian financial market: a pre-post analysis

This paper examines the impact of mergers and acquisitions (M&A) on the financial performance of the Nigerian market after consolidation. The authors use data from all Nigerian banks that survived the consolidation between 2001 and 2009. Logistic regression models are structured to determine the influence of M&A activities on the financial performance of the Nigerian market. Also, the authors critically evaluate the findings by shedding the light on the lessons other developing nations can learn from the Nigerian market. The results show that M&A have a positive influence on the financial performance of the Nigerian market. Still, M&A are not enough to achieve the wider objectives of banking sector reform. Towards this end, corporate governance reform must take place vis-à-vis consolidation exercises especially when these M&A are regulatory based rather than market based. The investigation uses a novel approach by comparing pre- and post- M&A results performance of merged banks as well as comparing these results with non-merged banks. Finally, the paper puts the results in context of the wider reform context and considers the effectiveness of the M&A as a tool for banking sector reform in developing countries. The investigation offers insights into the policy of banking consolidation which can be useful for policy makers in Nigeria and other similar economies.


Introduction ©
Banking sectors play a crucial role in economic development by mobilising savings into investment activities (Abdullahi, 2002;Mordi, 2004) and in the creation of wealth by facilitating capital formation, enhancing economic growth and development, reducing information costs and offering risk management services (Dogarawa, 2011). However, their ability to undertake these functions is influenced by the soundness and stability of the system within which they operate. The need for a strong, reliable and viable banking system, capable of meeting the expectations of its stakeholders cannot be overstated. Banking system reforms may be initiated by government in developing, as well as developed countries, to remedy any deficiencies undermining the banking system (Dogarawa, 2011;Ebimobowei and Sophia, 2011).
The history of the Nigerian banking system is one of regular periods of change and adjustment as the sector evolves in response to changes in the domestic and global economies. The foundation of the Nigerian banking industry in the late nineteenth century is described by Ezeoha (2007) as a system without any legal or regulatory framework. Initial banking operations were set up to meet the needs of the expatriate community with the establishment of the African Banking Corporation based in South Africa and subsequently absorbed into the British Bank for West Africa, now First Bank of Nigeria Plc (Danjuma, 1993). Industrial and Commercial Bank was the first indigenous bank in Nigeria, established in 1929, a time when banking was effectively unregulated and entry unrestricted (Brownbridge, 2005). This bank, and a number of subsequent banks failed, as a result of a number of factors including the lack of a firm regulatory framework; inadequate levels of capitalization and poor quality management (Agbaje, 2008;Nwankwo, 1980). Despite the introduction of banking legislation, these problems continued into the 21 st century.
In recent decades Nigerian banking has shown significant weaknesses which have resulted in a loss of confidence in the system. Soludo (2004) suggests that the Central Bank of Nigeria (CBN) has identified the need for adequate capitalization of the banks as key to build a strong, competent and globally competitive banking sector. Between 1952 and 2005, there were 9 different recapitalisation requirements imposed by the CBN. The most recent, in 2005 increased the minimum capital base for all banks from 2 billion Nigerian Naira to 25 billion Nigerian Naira (Somoye, 2008). The CBN considers that mergers and acquisitions (M&A) enhance bank soundness and efficiency, and give greater scope for development of the economy.
The purpose of this paper is twofold, firstly, to identify whether there is any difference in the financial performance of all Nigerian banks pre-post the consolidation in 2005 and secondly, to investigate whether the financial performance of all the merged banks improved after the consolidation. Compared with previous investigation in this area, particularly in the Nigerian market, our fresh contribution is twofold: firstly our investigation covers the whole financial market in Nigeria and secondly we use logistic regression to distinguish the performance of the financial market pre-post-M&A. The rest of this paper is organized as follows: section 1 reviews the related studies; section 2 addresses data sources and methodology; section 3 reports our results; and final section comprises conclusion and recommendations.

Review of relevant literature
In the last couple of decades, a lack of confidence in and under-capitalization of the Nigerian banking system has resulted in instability of the economy and subsequently in runs on the banks. Issues such as weak corporate governance, opaqueness, gross insider abuses, insolvency, weak capital base and over-dependency on the public sector deposits are identified in the Nigerian banking sector (Soludo, 2004;Sanni, 2010). Agu et al. (2011, p. 23) add that the Nigerian Banking system in mid-2004 suffered from a number of challenges including "periodic distress, weak credit regulation, poor management, macroeconomic and political instability, maturity mismatches, insider abuses, fraud and conflict of interest, general insecurity and corruption". To tackle the situation and allow the banks to play their role as a catalyst for economic development, banking system reforms were introduced by the CBN on the 6 th of July, 2004.
According to CBN, consolidation can strengthen the role of the banks within the Nigerian economy and generate improved returns for shareholders. The rationale for the consolidation strategy is to allow the Nigerian banking system to reap the benefits seen around the world from M&A activities such as "cost-savings due to economies of scale as well as more efficient allocation of resources; enhanced efficiency in resource allocation; and risk reduction arising from improved management" (Soludo, 2004, p. 3 Whilst acknowledging that there are many other factors which impact on the success of the banking sector, Joshua (2010) argues that issues, such as the maintenance of price and exchange rate stability, protection of investors, and provision of development capital could not be resolved without adequate capitalization of the sector. Banks have employed a variety of financial strategies to comply with CBN's minimum capital directives including: the injection of fresh capital through initial public offers, private placings and right issues; the capitalization of reserves; mergers and or a combination of two or more of the above strategies (Otanngaran, 2004). The impact of the reforms was a rationalization of the Nigerian banking sector, and a reduction in the number of banks from 89 to 24. The aim was to create a globally competitive banking system, by allowing the remaining banks to benefit from accelerated growth, enhanced profitability, economies of scale improved risk management and greater market power (Andrade et al., 2001;Goddard, 2007;DeYoung et al., 2009;Ebimobowei and Sophia, 2011).
The nature of the market could be a reason behind the M&A activities in the Nigerian Banking System as these were not motivated entirely by market dynamics, but were initiated and incentivized by the CBN as a tool for reform (Soludo, 2004;Alao, 2010;Ebimobowei and Sophia, 2011;Agu et al., 2011). The CBN offered technical assistance, securities and exchange commission fee waivers and finally, "allowed for transition time for operations merger and regularization of employee for merged banks beyond the consolidation deadline" (Agu et al., 2011: p. 23). This would seem to make the Nigerian bank consolidation different from the conventional market based consolidations cited above in the industrialized countries.
The literature relating to the benefits arising from M&A is complex and at times contradictory. Rhoades (1998) reports efficiency and profitability improvements in most cases studied (9 selected merger cases) with no significant issues impeding the achievement of their objectives. Similarly, Altunbas and Ibanez (2008) Beccalli and Frantz (2009), in a study of 714 deals involving EU acquirers and targets located throughout the world during the period 1991-2005, found that M&A activity is associated with slight deterioration in financial performance of banks postmergers if the transaction was a cross-border deal. They concluded that institutional and regulatory factors have an impact on post-merger financial performance.
Fewer researchers have examined the relationship between M&A and financial performance in this area. Adbayo and Olalekan (2012) use correlation co-efficient and t-test and conclude that there was a significant relationship between pre and post mergers capital base and profitability, and a significant difference between pre and post-mergers earning per share. Adegbaju and Olokoyo (2008) test the relationship between recapitalization and bank performance using mean, standard deviation, test of equality of means and t-test and found that yield on earning asset, return on equity and return on assets show significant difference before and after the previous recapitalization in 2001. Joshua (2011) in a relatively limited study of 3 banks over the period 2002-2008 finds mixed results. Whilst the study concludes that there were no statistically significant overall improvements in financial efficiency post consolidation, it does identify improved performance in gross earnings, profit after tax and net assets. Sanni (2010) also identifies variations in profitability between banks post consolidation. However, Somoye (2008) examining Nigerian banks' performance post 2004 consolidation concludes that consolidation exercise has not improved the overall performance of banks significantly. This study questions whether the system would benefit from further consolidation exercises, and believes that improvements would only follow if other aspects were also improved, in particular; a reform of corporate governance and action to strengthen balance sheets.
In conclusion, although the consolidation program of Nigerian banks was initiated to enhance efficiency, none of the previous research addresses this issue using statistical techniques such as logistic regression to distinguish the performance of Nigerian banks pre-post 2005 consolidation. To the best of our knowledge, financial performance differences pre-and post-M&A in the Nigerian market has not been addressed in this way by any other researchers.

Research methodology
Our overall research question is as follows: whether there is any significant difference between the financial performance of merged and non-merged  Note We have provided in Table 1 Table 1.
We use different financial ratios to investigate whether there are any differences in the Nigerian banks' financial performance pre-and post-the 2005 consolidation. These ratios cover four different categories namely asset quality, capital adequacy, profitability and liquidity. We started the analysis with 29 financial ratios and after excluding those with missing data; and those showing high correlations between different ratios, the final sample consists of 15 financial ratios, as shown in Table 2.

Logistic regression. Logistic regression (LR)
which is also known as logit model is a technique where independent variables are used to determine an outcome of a dependent variable on the basis of continuous or categorical independents to determine the percent of variance in the dependent variable. The outcome is measured with a dichotomous variable which tests the significance of the individual independent variable to find the best fitting model to describe the relationship between the dichotomous characteristic of interest (dependent variable) and a set of independent predictor/explanatory variables.
What distinguishes a logistic regression model from the linear regression model is that the outcome variable in logistic regression is binary or dichotomous. On theoretical grounds, it might be supposed that logistic regression is a more appropriate statistical tool than linear regression, given that two discrete classes "1" and "0" have been defined (Hand & Henley, 1997;Abdou, 2009).
LR is a widely used statistical modelling technique, in which the probability of a binary outcome (zero or one) is related to a set of potential predictor variables in the form: where p is the probability of the dichotomous outcome of interest, α is the intercept term, and δ i represents the respective coefficient in the linear combination of explanatory variables, V i , for i = 1 to n. The dependent variable is the logarithm of the odds ratio, It should be emphasized that we run correlation between our explanatory variables, and results show that all variables had a correlation within an acceptable range (i.e. < 0.50). However, there was an exception with four variables as follows: there were high correlation between ROAA and both ROAE and cost to income ratios at values of 0.0767 and -0.748, respectively; and between net loans to total assets and net loan to deposit and short-term funding at a value of 0.841. Due to the importance of these variables, it was decided to keep them and to run an Orthogonalisation test to avoid the high correlation. After running the test, correlation between ROAA and both ROAE and cost to income ratios become 0.072 and 0.052, respectively; and correlation between net loans to total assets and net loans to deposit and short-term funding become 0.098.

Empirical results
In this section we exhibit our detailed results. We use data collected from fifteen Nigerian banks out of which four non-merging banks are used as a benchmark. In order to critically assess whether there is improvement in the financial performance of the Nigerian banks after M&A, the data are analyzed using financial ratios and a t-test for equality of means is used to capture any significant differences. Subsequently, three logistic regression models are structured to describe the relationship between the dependent variable and the 15 explanatory financial ratios to determine the significant changes in the financial performance of the banking sector four years before and after the merger took place.

Descriptive statistics. Asset quality ratios:
Asset quality is used to measure the quality of Nigerian banks' earning assets. This is measured by four financial ratios as shown in Table 2. Asset quality of the Nigerian market measured by impaired loans to equity suggests an improvement post-M&A with a mean value of 31.47 compared with a value of 55.21 pre-M&A. The pre-and the post-average figures corresponded to a 4 years period each (2001-2004 and 2006-2009) respectively. This result is also confirmed by the ttest for equality of means as there is a statistically significant difference between the pre-and post-M&A at the 10% level, as shown in Table 2. Capital adequacy ratios: Capital adequacy is used to determine how Nigeria banks could cope with shocks relating to their balance sheet. This category is measured by equity to total assets and equity to net loans ratios. The average means indicates that all banks experienced a great improvement in their capital level after the merger exercise as both ratios means increased after the consolidation. This is also confirmed by the t-test results which reveal that there is a significant difference between the two periods at the 1% level with a p-value of 0.000, as shown in Table 2. Therefore, this is strongly implies that M&A have improved the financial performance of Nigeria market. Liquidity ratios: Liquidity ratios are used to determine how the Nigerian banks are able to meet their financial obligations to the stakeholders. Liquidity as the lifeblood of any organisation determines the survival of banks and their inability to meet the demand of their customers exposed them to liquidity risk. This category is measured by three financial ratios namely; net loans to total assets, net loans to deposit & short-term funding and liquid assets to deposit & short-term funding. Our result for two liquidity ratios indicates that M&A have improved the performance of the Nigerian market by potentially increasing the loan activities. This is evidenced by the higher average mean of net loans to deposits & short-term funding; and the lower average means of liquid assets to deposits & short-term funding. Our t-test results confirm this and show that there are statistical significant differences between the two periods for both ratios at the 10% and the 5% levels, respectively, as shown in Table 2. These three financial ratio categories show a positive impact of the M&A on the Nigerian market.
By contrast, operations (profitability) ratios suggest that M&A in the short-term has a slight adverse effect on the Nigerian market financial performance as measured by operation ratios. Operations ratios are very significant in exhibiting the ability of bank to generate profits from its assets or equities. This category is measured by 6 financial ratios, and the average mean of the four significant ratios namely net interest margin, other operating income to average assets, non-interest expenses to average assets and return on average equity, is reduced post-M&A, as shown in Table 2. This is also confirmed by the t-test results which indicate significant differences between the two periods at the 1% level. This is considered as a downside of the M&A as the Nigerian market may need more time to capture the benefits of economies of scale.

Logistic regression models.
Results for the first model (LR 1 ): This model is designed to analyze the overall financial performance of the Nigerian market i.e. all banks four years before the financial period of 2005 in which the reform took place and comparing it with the performance four years after the M&A exercise to ascertain the influence of the M&A activities on the efficiency and performance of the whole market. The results of our logistic regression LR 1 model indicate that the model is statistically significant at the 99% confidence level with a P-value of 0.000, with R 2 value of 94.09% (R 2 Adj. = 66.71%). The model has a significantly low mean square error of 0.21% and a 15.17% mean absolute error, as shown in Table 3. This result implies that there are considerable differences between the two periods. This also implies that there are some improvements in the financial performance of the Nigerian banking industry after the reformation exercise.
The P-values for the likelihood ratio test also show significant differences in the capital ratios namely equity to total assets and equity to net loans at the 99% and 90% levels of confidence, respectively. This result strongly supports our previous findings that the banks have increased their equity and therefore they experienced a great improvement in their capital level after the consolidation. Asset quality ratios, namely loan loss provision to net interest revenue, and impaired loans to equity are both statistically significant at the 99% and the 90% levels of confidence, respectively. This result implies that the cost of running the banks has been reduced after M&A activities and thereby increases bank efficiency and profitability and the banks' assets have to some extent been used efficiently to generate income due to the effect of M&A.
Operations ratios namely non-interest expense to average asset and return on average equity are also significant at the 90% and 99% levels of confidence respectively. This result signifies that the M&A exercise has an influence on the financial performance of the Nigerian market's profitability. Finally, liquid assets to deposits and short term funding ratio is the only significant liquidity ratio at the 99% level of confidence, as shown in Table 3. This result indicates that M&A contributed to the improvement of banks liquidity in the Nigerian financial market measured by the banking industry.
As shown in Table 3, the most important explanatory variable as measured by Chi 2 value is 'loan loss provision to net interest revenue' ratio with a value of 110.92. This followed by three ratios namely return on average equity, liquid assets to deposit and short tern funding and equity to total assets with Chi 2 values of 38.758, 24.421 and 21.482, respectively.  Our LR 1 Stepwise model results show similar findings as per the LR 1 model. The overall model is statistically significant at the 99% confidence level with R 2 value of 76.64% (R 2 Adj. = 64.45%) and 1.08% and 37.36% mean square error and mean absolute error, respectively. In terms of significant explanatory variables, the model has a slight change as other operating income to average assets ratio become significant at the 99% level of confidence; and equity to total assets is no longer significant. All other variables are statistically significant at the 99% level of confidence a part form impaired loans to equity ratio which is significant at the 95% level of confidence, as shown in Table 3. Our graphical analysis shows the prediction capability for our dependent variable (pre-post M&A) describes the relationship between different cut-off points and the per cent correctly classified. As shown in Figure 1, the middle blue line refers to the overall correctly classified. The highest orange line at the lower cutoff rates is the post-M&A correctly classified set, while the lowest red line at the lower cut-off rates refers to the Pre-M&A classified set, in both LR 1 (on the left-hand side) and LR 1 Stepwise (on the right-hand side), and vice-a-versa at the higher cutoff rates. before and after the financial period of 2005 in which the M&A activities took place. Second logistic regression (LR 2 ) model results reveal that a p-value of 0.000 for the analysis of deviance is found and the model is statistically significant at the 99% level of confidence. The model R 2 is 35.56% (R 2 Adj. = 2.64%) with mean square error of 2.65% and mean absolute error of 35.65%. This to some extent indicates that there are differences between the financial performance of the 11 merged banks and the other 4 unmerged banks after the introduction of consolidation exercise, as shown in Table 4. The p-values for the likelihood ratio test show that none of the capital and the liquidity ratios is statistically significant. This implies that M&A did not have a positive influence on the performance of the merged banks due to intense completion after the exercise. By contrast, five operations ratios are statistically significant at different levels of confidence, and one asset quality ratio namely loan loss reserves to impaired loans is statistically significant at the 95% level of confidence, as shown in Table 4. As per the importance of the explanatory variables, Table 4 shows that return on average assets is the most important variable with a Chi 2 value of 10.473. This followed by four ratios namely return on average equity, loan loss reserves to impaired loans, net interest margin and other operating income to average assets with Chi 2 values of 4.9509, 4.9346, 4.5359 and 4.1053, respectively.  The LR 2 stepwise model results show slightly different results. The overall model is statistically significant at the 99% confidence level with R 2 value of 24.76% (R 2 Adj. = 13.29%) and 2.40% and 33.87% mean square error and mean absolute error, respectively. In terms of significant explanatory variables, the model shows that all the 5 significant variables are statistically significant at 95% level of confidence at least. For the capital category only one ratio namely equity to net loans is statistically significant at the 95% level of confidence confirming the LR 1 model results. This indicates that the increase in the capital base of the Nigerian market signifies some improvement in the market financial performance. In line with LR 1 model findings, one asset quality ratio namely loan loss reserve to impaired loans is statistically significant at the 95% level of confidence. In addition, both return on average assets and return on average equity are statistically significant at the 99% and 95% levels of confidence, respectively. Finally, one liquidity financial ratio namely liquid assets to deposit and short term funding is statistically significant at the 95% level of confidence, as shown in Table 4. A number of variables become insignificant while both equity to net loans and liquid assets to deposit and short term funding become significant at the 95% level of confidence; as shown in Table 3. The prediction capability for our dependent variable (pre-post M&A) describes the relationship between different cut-off points and the per cent correctly classified, as shown in our graphical analysis in Figure 2. The middle blue line refers to the overall correctly classified. The highest orange line at the lower cut-off rates is the post-M&A correctly classified set, while the lowest red line at the lower cut-off rates refers to the pre-M&A classified set, in both LR 2 (on the left-hand side) and LR 2 stepwise (on the right-hand side), and vice-a-versa at the higher cut-off rates. Clearly the distribution of the three lines is different compared to the previous model i.e. LR 1 , and leans to the right hand side or higher cut-off scores which confirms our numerical results. Generally speaking, it may be argued that our results based on this model are not strong enough as per the significantly low R 2 Adj. and therefore logistic regression (LR 3 ) model is suggested here. This may be due to the un-balanced sample used in building the LR 2 model i.e. 11 merged banks versus 4 non-merged banks.
Result for the third model (LR 3 ): This model is designed to access the effect of M&A activities on the financial performance of the Nigerian market by comparing the 4 merged banks with the other 4 unmerged banks based on their similarity in total assets (i.e. ln total asset -see shaded banks in Table 1), this is to steer clear of any bias comparing 11 banks with 4 banks, which is proposed in LR 2 model. These 8 banks are examined in order to test whether there are differences in their performance four years before and after year 2005 of the reform exercise.
Third logistic regression (LR 3 ) model results show that the model is statistically significant at the 99% level of confidence with a P-value of 0.000. The model has R 2 value of 94.12% (R 2 Adj. = 44.94). The model has a significantly low mean square error of 0.16% and 11.89% mean absolute error, as shown in Table 3. This shows that M&A have a great influence on the Nigerian market when comparing two sets of banks which are equivalent in size, as shown in Table 5.  This is also applicable to the p-value of the likelihood ratio tests which reveals very strong significant differences of 12 out of 15 financial explanatory variables at the 99% level of confidence used in building this model. Capital ratio category shows that equity to net loans is statistically significant at the 99% level of confidence. This result is in line with our t-test findings which indicate that these banks experienced a great improvement in their capital level after the merger exercise as per the positive association for the estimate value (i.e. 0.8511) which imply that equity has increased after the consolidation. All asset quality ratios, except impaired loans to equity, are statistically significant at the 99% level of confidence. Similarly, all operations ratios, except return on average equity, are statistically significant at the 99% level of confidence. These results are in line with our t-test results previously explained. Finally, all liquidity ratios are statistically significant at the 99% level of confidence which proves that the market has potentially increasing the loan activities. Our results imply that the Nigerian market asset quality, capital and liquidity have been enhanced by M&A activities even though the banks' profitability has not been efficiently improved as the Nigerian market may need more time to capture the benefits of economies of scale. As shown in Table 5, the most important explanatory variable as measured by Chi 2 value is 'cost to income ratios' ratio with a value of 27.304. This is followed by six ratios all with a very similar Chi 2 value, as shown in Table 5.
Our LR 3 Stepwise model results show similar findings as per the LR 3 model. The overall model is statistically significant at the 99% confidence level with R 2 value of 87.99% (R 2 Adj. = 55.92%) and 0.62% and 21.33% mean square error and mean absolute error, respectively. In terms of significant explanatory variables, the model includes 9 significant variables at the 99% level of confidence; which means three financial ratios are no longer significant, as shown in Table 5. Expectedly, this model has considerably improved the previous model (i.e. LR 2 ) results as the sample includes 4 merged and 4 non-merged banks with similar total assets. cut-off rates refers to the pre-M&A classified set, in both LR 3 (on the left-hand side) and LR 3 stepwise (on the right-hand side), and vice-a-versa at the higher cut-off rates.
Clearly our investigation provides an answer to the main research question and based on our results, it can be concluded that there are significant differences between the financial performance of merged versus non-merged banks in the Nigerian market. Evidently, as per our results for the three financial categories namely asset quality, capital and liquidity, further consolidation can help increase the soundness of the Nigerian financial market which can help in achieving the CBN objectives.

Conclusion and areas for future research
This paper's main aim is to measure the effect of M&A on the Nigerian market's financial performance by comparing it 4 years pre-and 4 years post the 2005 consolidation. Our main findings based on t-test show that the overall market asset quality, capital and liquidity have improved whilst the market profitability has not. This is considered as a downside of the M&A as the Nigerian market may need more time to capture the benefits of economies of scale. There is evidence that the financial performance of the market is different between the two periods. This indicates that M&A has significant impact on the financial performance of the Nigerian market regardless the fact that their profitability is not yet improved. This in fact disagrees with other researchers' findings (see for example, Kithinji and Waweru, 2007).
All logistic regression models' results show that the P-values in the analysis of deviance are less than 0.01 which denotes that these models are all statistically significant at the 99% level of confidence, indicating that M&A have a great influence on the efficiency and financial performance of the Nigerian market as measured by the banking industry. Our logistic regression models' results show that there are significant differences between the pre-and the post-M&A financial performance of the overall market, as evidenced by LR 1 model results. We also have evidence that banks which merged are significantly different from those which are not, as evidenced by LR 3 model results.
Future research should consider including those banks for which financial information is not currently available due to the new identity issues. More financial and non-financial variables could be used. Various statistical techniques should be used as it is expected that more accurate results could be achieved if more sophisticated modelling techniques such as neural networks are used. It can be argued that the lack of improvement in profitability in the sector is a result of time needed to benefit from economies of scale, a longer time frame post-M&A could be considered to capture a wider picture of the consolidation effect of the market on profitability. An extension of the time frame would perhaps also give an indication of whether there is a point at which the amount of M&A activity is optimized, and beyond which the benefits reduce or are eliminated entirely. These findings could have wider implications to other nations in which the financial systems have been in a state of instability for some time. The high degree of significance in our results suggests that other countries with developing banking systems may benefit from a period of consolidation and M&A activity, leading to greater strength in the institutions themselves and the underlying system.