THE IMPACT OF MACROECONOMIC VARIABLES ON BANK PERFORMANCE IN NIGERIA

1. INTRODUCTION

Commercial banks play a vital role in the growth and development process of an economy. They also act as financial intermediaries as well as being a key source of financing for businesses. Banks also ensures the efficient allocation of resources in that they mobilize funds for various productive activities. In practice, they transfer funds from those with surplus funds to those that need it for productive activities which in turn stimulate investments and improve economic growth and development. On the contrary, when banks fail in their financial intermediation function, we expect declining economic growth and development. Poor banking performance will also result in bank failures and financial crisis like those witnessed during the global financial crisis that began in 2007 (Marshall 2009). This explains why banks are one of the most regulated sectors in the world today. Saunders (1994) notes that the role banks play in an economy is so significant that the collapse of a large bank would be worse than that of any other institution in an economy.

However, we must point out that the performance of banks is contingent on several factors. In specific terms, how well the banking system performs can be affected by both internal and external factors. The internal factors majorly entail the competition among banks while the external factors include the financial and macroeconomic conditions surrounding a country. Overall, how well banks can compete are mainly driven by their capacity to innovate. Thus, we expect the competitiveness of each bank to vary according to the features and advantages they possess. On the contrary, the financial and macroeconomics conditions facing all banks are the same. On one hand, favorable macroeconomic conditions will stimulate the development of the banking system while on the other hand; unfavorable financial and macroeconomic conditions might spell doom for banking performance. To put it simply, an unstable and uncertain macroeconomic climate can affect the credit and market risk of banks which in turn results in a dismal banking performance. In addition, spillover effects from financial crisis in most especially, the developed financial markets in other countries of the world can also hinder the domestic performance of banks.

Beginning from the late 1980s, many developing countries began adopting the Structural Adjustment Programs (SAP). This policy led to major transformations in the banking sector. Countries eased controls on interest rates and reduced government involvement in the financial sector thereby opening doors to international banks (Ismi 2004). In Nigeria, the Structural Adjustment Programme (SAP) was adopted in 1986. Another reform measure worthy of mention is the National Economic Empowerment and Development Strategy (NEEDS) programme which sought to re-capitalize the financial system as well as address incidences of systemic distress in the financial sector. This programme led to the increase in minimum capital requirement for banks from 2 billion to 25 billion naira in 2004. Before this time, the banking system was highly concentrated with about four banks controlling about 50 percent of the entire industry (Asogwa 2004; Nwokoma 2006). The banking consolidation policy also resulted in the reduction of banks operating in the country from 89 in 2004 to 25 in 2006, a process that resulted from mergers and acquisitions.

Despite these reforms, the sector continues to face numerous challenges. Not only is the banking sector dominated by few banks, five banks dominate the shareholders’ funds accounting for about 48.52% of capitalization of the entire banking industry. According to the CBN (2015), 46.34% of the total branch network in the country is accounted for by the top five banks. The KPMG (2013) report also states that only about 20% of Nigeria’s population is banked. As at 2014, the number of commercial banks branches stood at 5,526 from 5,639 in previous years. Also, the ratio of bank deposit to GDP declined to 19.97% in 2016 from 20.7% in 2015. In terms of GDP contribution, the banking sector only contributes 2.63% to GDP compared to telecommunications (9.28%), real estate (7.72%) and crude petroleum and natural gas (13.70%) (Central Bank of Nigeria 2015). Turning to domestic credit to the private sector as a ratio of GDP, it is also observed that the banking sector with about 14.5% is behind the world average which is 84.5% and Sub-Saharan Africa, Low Middle Income countries, South Africa, Kenya and Cameroon with about 28.2%, 40.6%,67.2%, 34.1% and 15.5% respectively.

Banking sector profitability also declined from N601.02 billion in 2014 to N113.827 billion in 2015. Turning to the macroeconomic environment, we also observed that the overall economic climate has not been favorable. In the second quarter of 2016, the Nigerian economy was declared to be in recession having posted three consecutive quarters of declining growth. Since the last quarter of 2013, GDP growth has consecutively declined reaching 2.11% in the fourth quarter of 2015 and further reporting negative growth in the first, second, third and fourth quarters of 2016 (-0.36, -2.06, -2.24 and -1.3) (CBN 2017). Recent figures also show the situation has not improved. A recent release by NBS (2018) reported that the Nigerian economy only grew by 1.5% in the second quarter of 2018.

One key factor responsible for the dismal economic growth figures since 2014 is the decline in crude oil prices which has adversely affected foreign exchange earnings. Consequently, the naira has depreciated greatly against the dollar and other major currencies. In addition, the prices of imported inputs and products are higher which explains the high inflation in recent years. In October 2016, inflation stood at 18.3% which is the highest rate since 2005. The index for business confidence has also reported negative values since the first quarter of 2016, from 10.3% to -24.1% in the third quarter. Based on casual observations, one would observe that the fortunes of the economy are correlated with banking performance. To put it simply, the banking sector performed better during periods of sustained macroeconomic stability.

The global financial crisis that began in 2007 also resulted in several banks becoming distressed. Consequently, the CBN had to bail out some of these banks. The recapitalization policy of the CBN greatly stretched the financial capacity of banks. They were therefore able to finance big projects in the oil and gas industry. However, the fall in the international price of crude oil made Nigerian banks vulnerable to the vagaries of the international crude oil market. On average, around 23 percent of total banks’ loan exposure between the period, 2010 and 2014 was from the oil and gas sector (NDIC 2014). It is therefore expected that the fall in crude oil prices would present added risks to banks with its attendant impact on bank profitability and balance sheet. In Nigeria, the broader economic environment has been characterized by declining GDP growth rates, falling oil prices, rising inflation, uncoordinated fiscal and monetary policies and a depreciating currency. All these are factors that are external to the banks’ operations. They therefore have little or no control over them with its attendant impact on bank performance (Osamwonyi and Micheal 2014). At this point, an important question to ask is whether macroeconomic performance affects bank performance in Nigeria. Though casual observation answers this question, that is not however enough to establish that macroeconomic performance affects banking industry performance in Nigeria, hence, this study.

In the literature, scholars have put forward arguments on the need to curtail the adverse impact of macroeconomic shocks on bank performance using appropriate policy measures. While some argue on the need to ensure proper corporate governance, others are of the opinion that attention should be paid to issues surrounding proper regulation by the appropriate authorities. This study is motivated by the apparent dearth of empirical studies on the impact of macroeconomic variables on bank performance in Nigeria. The literature is replete with empirical studies focused on two categories: those that are country specific (see Berger, Hanweck, and Humphrey 1987; Berger 1995; Barajas, Steiner, and Salazar 1999; Naceur and Goaied 2001; Naceur 2003; Athanasoglou, Sophocles, and Matthaios 2005; Siregar, Maulana, and Hasanah 2015; Sheefeeni 2015) and those based on panel studies (see Bourke 1989; Molyneux and Thornton 1992; Demirgüç and Huizinga 1999, 2001; Abreu and Mendes 2002; Flamini et al. 2009). Studies on Nigeria are apparently scant, and the few available ones majorly focused on bank-specific determinants, leaving out the influence of macroeconomic factors on bank performance (see Baba and Nasieku 2016; Obamuyi 2013; Osuagwu 2014; Olaoye and Olarewaju 2015; Akani, Nwanna, and Mbachu 2016). This study therefore takes a different approach by exploring the influence of macroeconomic variables in view of the macroeconomic factors that has been hindering bank performance in recent times in Nigeria. We therefore expect that this study and the novel approaches adopted will greatly enrich the extant literature.

Our general findings are that a long-run relationship exists among macroeconomic variables and bank performance when ROA and CAD serve as the dependent variables. We also report that GDP growth rate, interest rate and trade are the most important predictors of bank performance in Nigeria. In specific terms, we found that GDP growth rate and trade promotes bank performance while higher interest rate impedes bank performance. The main take away from this study is that GDP growth rate, trade and interest rate are the most important determinants of bank performance in Nigeria.

The rest of this paper is organized as follows. Section 2 presents the literature review while section 3 presents the empirical methodology. In section 4, we present the empirical results while section 5 concludes with pointers for further research.

2. LITERATURE REVIEW

The empirical literature on the determinants of bank performance and the influence of macroeconomic variables on bank performance has attracted several studies, but so far, the empirical evidences have yielded mixed and inconclusive results. In the empirical contribution by Gizycki (2001), based on fixed effect modeling and impulse response function with a focus on Australia, it is reported that macroeconomic variables exert a strong influence on bank’s risk and profitability while Gerlach, Peng, and Shu (2005) in their study based on a confidential supervisory bank-level data set for Hong Kong, reported that a decline in economic growth resulted in an increase in the risk of non-performing loans of banks. A similar study for Singapore by Clair (2004) also reported that banking performance is explained to a large extent by changes in exchange rate, interest rate, aggregate demand and unemployment. In specific terms, it is reported that these variables explain around two thirds of the changes in bank performance.

In a related study, Siregar, Maulana, and Hasanah (2015) documented the effect of macroeconomic indicators such as inflation, production index, exchange rate, crude oil price, Jakarta stock index and Bank Indonesia rate on the performance of state-owned banks. The study employed a Vector Error Correction Model with empirical results reporting that the shock of Bank Indonesia’s rate provided the largest response of most of the bank performance indicators. Simiyu (2015) also investigated the impact of macroeconomic variables on the profitability of commercial banks listed on the Nairobi Securities Exchange. In this study, panel data and fixed effect analysis was employed with empirical results revealing that exchange rates, GDP and interest rates do not significantly influence bank performance. On the contrary, Sheefeeni (2015) reported dissimilar results in that macroeconomic variables were found to significantly influence bank performance in Namibia. In the empirical study by Evans and Kiganda (2014) based on correlation analysis and the OLS technique, it is reported that inflation, GDP and exchange rate did not significantly influence bank profitability in Kenya. This is also in line with the empirical results reported by Kanwal and Nadeem (2013) in that macroeconomic variables had an insignificant impact on the profitability of commercial banks in Pakistan.

In a study based on Kenya, Ongore and Kusa (2013) employed panel data analysis with empirical results reporting that macroeconomic variables did not find significantly impact bank performance. Whilst on the other hand, Combey and Togbenou (2017), reported that inflation, GDP growth and real effective exchange rate significantly impacted bank profitability in Togo. Patrama (2015) also employed the Vector Error Correction Model in a study based on the influence of macroeconomic variables on the performance of Indonesian Islamic banks. Empirical results revealed short term shock of banking performance to fluctuations in macroeconomic variables. In a related study by Carvallo and Pagliacci (2014) based on the Venezuela banking industry, it was found that rising interest rate and domestic currency appreciation contributed to banking sector instability. The empirical study by Ongeri (2014) also found that macroeconomic variables significantly impacted bank performance in Kenya. While Festic and Beko (2008) reported that GDP growth influences bank performance positively in Togo.

Saeed (2015) also made an empirical contribution to the literature on the impact of industry-specific, bank-specific and macroeconomic variables on bank profitability in the UK. The study was based on data from 73 UK banks. Empirical results based on correlation and regression analyses revealed that inflation and GDP growth had a negative impact on bank profitability while loan, deposits, bank size, capital ratio, interest rate and liquidity had a positive impact on ROA and ROE. Gikombo and Mbugua (2018) also examined the impact of macroeconomic variables on the performance of 44 listed commercial banks in Kenya. From the empirical results, it is concluded that the profitability of commercial banks is affected most by the GDP while interest rate significantly influenced the return on assets and return measures of profitability.

Several studies have also documented the impact of macroeconomic variables on bank performance in Nigeria. However, the evidences presented thus far are largely mixed, inconsistent and inconclusive. Olaoye and Olarewaju (2015) in a study based on panel data analysis related macroeconomic variables and bank specific factors to the profitability of Nigerian commercial banks. Empirical results did not reveal any significant relationship among the variables. In a similar study based on Granger Causality tests, Johansen Cointegration tests, multiple regression model and Vector Error Correction Modeling, Akani, Nwanna, and Mbachu (2016) finds that macroeconomic variables did not significantly impact bank profitability. Osamwonyi and Micheal (2014) also documented the impact of macroeconomic variables on the profitability of Nigerian listed commercial banks. Results revealed that inflation and interest rate were the key variables influencing banking performance. A similar study was also carried out by Abusomwa (2018). This study employed the Generalized Method of Moments technique based on based on data from 120 bank branches and 2400 bank customers in Nigeria. Empirical results revealed that macroeconomic performance positively influenced the performance of Nigerian banks. In addition, employment status and gender of managers also influenced bank performance.

The empirical study by Baba and Nasieku (2016) also provided empirical evidence on the subject matter. Empirical results revealed that exchange rate, interest rate and unemployment rate influences bank performance negatively in Nigeria. Obamuyi (2013) reported that improved bank capital and interest income and favorable economic condition resulted in an improvement in bank performance. Osuagwu (2014) also provided evidence on the determinants of bank profitability in Nigeria using a panel of selected banks. Results revealed that bank profitability is largely influenced by credit risk as well as other factors related to the internal organization of banks. Although, the literature is replete with empirical studies on Nigeria, results so far are conflicting and the controversies are far from settled. The raging issues and mixed results reported by previous studies necessitates the need to revisit the subject area. We therefore expect that our approach will greatly enrich the literature as well as uncover what macroeconomic variables portend for bank performance in Nigeria.

3. METHODOLOGY

3.1 MODEL SPECIFICATION AND ESTIMATION TECHNIQUES

In this paper, statistical and econometric models were estimated to establish the effect of macroeconomic variables on the performance of Nigerian banks. This is to verify the linkages between Nigerian banks’ performance and changes in the economy and the magnitude of the banks’ reactions to such changes. The model for this paper assumes an underlying relationship between macroeconomic variables that can affect banks’ performance variables. Given the macroeconomic nature of gross domestic product (GDP), other macroeconomic variables can be brought in. To examine the underlying relationships, the paper employs an autoregressive-distributive lag (ARDL) test approach to co-integration analysis.

To estimate the impact of macroeconomics variables on bank performance, we present a model that relates bank performance variables (BPV) to some macroeconomic variables (MV):

$$\hspace{23mm} BPV_{t} = f(MV_{t}) \hspace{23mm} (1)$$

where:

$BPV$ is a vector of $LQR, ROA$ and $CAD$

$MV$ is a vector of $GDPG, INF, FKF, INT, EXR$ and $TRD$

The relationships between the components of BPV and the different independent variables can be rewritten implicitly as follows:

$$\begin{align} \hspace{7mm} LQR_{t} = f( &GDPG_{t}, INF_{t}, FKF_{t}, \\ &INT_{t}, EXR_{t}, TRD_{t}, \upsilon_{t}) \hspace{10mm} (2) \end{align}$$

$$\begin{align} \hspace{7mm} ROA_{t} = f( &GDPG_{t}, INF_{t}, FKF_{t}, \\ &INT_{t}, EXR_{t}, TRD_{t}, \mu_{t}) \hspace{10mm} (3) \end{align}$$

$$\begin{align} \hspace{7mm} CAD_{t} = f( &GDPG_{t}, INF_{t}, FKF_{t}, \\ &INT_{t}, EXR_{t}, TRD_{t}, v_{t}) \hspace{10mm} (4) \end{align}$$

LQR, ROA, CAD, GDPG, INF, FKF, INT, EXR and TRD represents liquidity ratio, return on asset (profitability proxy), capital adequacy, growth of real gross domestic product, inflation, foreign capital flows, interest rate, exchange rate and trade respectively. In addition, $t$ represents time while $\upsilon, \mu$ and $v$ are stochastic error terms.

The explicit form of equations 2 – 4 is represented as follows:

$$\begin{align} LQR_{t} = &\varphi_{o} + \varphi_{1}GDPG_{t} + \varphi_{2}INF_{t} + \varphi_{3}FKF_{t} \\ &+ \varphi_{4}INT_{t} + \varphi_{5}EXR_{t} + \varphi_{6}TRD_{t} + \upsilon_{t} \hspace{3mm} (5) \end{align}$$

$$\begin{align} ROA_{t} = &\varphi_{o} + \varphi_{1}GDPG_{t} + \varphi_{2}INF_{t} + \varphi_{3}FKF_{t} \\ &+ \varphi_{4}INT_{t} + \varphi_{5}EXR_{t} + \varphi_{6}TRD_{t} + \mu_{t} \hspace{3mm} (6) \end{align}$$

$$\begin{align} CAD_{t} = &\varphi_{o} + \varphi_{1}GDPG_{t} + \varphi_{2}INF_{t} + \varphi_{3}FKF_{t} \\ &+ \varphi_{4}INT_{t} + \varphi_{5}EXR_{t} + \varphi_{6}TRD_{t} + v_{t} \hspace{3mm} (7) \end{align}$$

The a priori expectation is such that $\varphi_1, \varphi_2, \varphi_6 > 0$ and $\varphi_2, \varphi_4, \varphi_5 < 0$

Setting: $y_t = BPV_t$ and $x_t = MV_t$

We proceed to formulate an Error Correction Model (ECM) under the ARDL model. This is expressed as follows:

$$\begin{align} \hspace{6mm} \Delta y_t = &\varphi_o + \varphi_j y_{t-1} + \varphi_k x_{t-1} \\ &+ \sum_{i=1}^m \varphi_j \Delta y_{t-1} + \sum_{i=0}^m \varphi_k \Delta x_{t-1} + \varepsilon_t \hspace{7mm} (8) \end{align}$$

The first part of equation 8 shows the long-run dynamics of the model while the second part represents the short-run relationship. $\Delta$ is the first difference operator, $\varepsilon_t$ is a white noise disturbance term as the equation try to show that bank performance variables tend to be influenced by its previous level, hence it involves other shocks $j = 1 - 3$ and $k = 1 - 6$.

The choice of the ARDL modeling method is based on several reasons. First, it can be applied regardless of the stationary properties of the variables (series) in any given sample (Olokoyo, Osabuohien, and Salami 2009). The ARDL model also takes enough numbers of lags to capture the data generating process in a general-to-specific modeling structure (Pesaran, Shin, and Smith 2000). Furthermore, the technique allows us to obtain a dynamic Error Correction Model (ECM) through a simple linear transformation, which allows for inferences on long-run estimates (Frimpong and Oteng-Abaiye 2008). The ARDL model also yields consistent estimates of the long-run parameters that are asymptotically normal irrespective of the order of integration, i.e. whether variables are I(0), I(1) or mutually integrated since there is no need for unit root pretesting. However, it is still essential to complement the estimation process with a unit root test in order to be sure that none of the variables are integrated of higher order like I(2) (Hall and Milne 1994; Luintel and Khan 1999). The ARDL method can also distinguish between dependent and explanatory variables. Therefore, when using the ARDL method, it is possible to estimate even when the explanatory variables are endogenous (Alam and Quazi 2003). Results provided by this technique are also robust even when we are dealing with small sample size such as less than 80 observations (Narayan 2005).

In practice, the ARDL technique involves two major procedures. The first step involves testing for the existence of long run relationship while the second stage involves estimating the coefficients of the long run relationship. The technique employs the bound test to test for long run relationships. When the F-statistic is greater than the upper bound, a significant long run relationship exists while on the other hand, if the F-statistic falls within the respective bounds, the inference would be inconclusive. In this study, the appropriate lag for our model is determined using the Akaike Information Criterion (AIC). This criterion is superior to other criteria in cases where we have 60 observations or below in that they minimize the chance of underestimation while maximizing the chance of recovering the true lag length (Liew 2006). A key step in time series analysis is checking for the stationarity properties of variables. In this study, we employ the Augmented Dickey Fuller (ADF) test and Phillip-Perron (PP) test. The ADF test is robust even with the presence of serial correlation while the PP test uses a non-parametric correction to deal with any correlation in the error terms.

3.2 DATA DESCRIPTION

BANK PERFORMANCE VARIABLES

Liquidity Ratio (LQR): This is measured by the ratio of banks’ liquid assets to total assets and serves as a useful guide to assess the vulnerability of the banking sector to loss of access to market risk. It serves as an indicator of the extent to which banks can meet the short-term withdrawal of funds on demand without liquidity problems. Liquidity is expected to be positively related to economic growth.

Profitability: Return on Asset (ROA) ratio is used as a proxy for profitability i.e. earning efficiency of the banks. It is measured by the ratio of profit to total assets of the banks. It measures the extent of banks’ management of resources by minimizing risk and maximizing returns with strict adherence to international best corporate governance practices. Profitability is expected to be positively related to growth in the economy.

Capital Adequacy: This is measured by the ratio of tier 1 capital to total loans of the banks and it might signal either a positive phase of the business cycle if it is led by demand factors (suggesting a negative sign) or an aggressive supply policy of the banks that might expose the banks to excessive risks and higher future capitalization (suggesting a positive sign). BIS standard imposed a risk-based capital ratio of 8% and any capital in excess of this ratio will be a buffer. Hence, banks with a big buffer will be less sensitive to macroeconomic changes and vice versa.

MACRO-ECONOMIC VARIABLES

Gross Domestic Product Growth (GDPG): The growth in the economy is measured by growth in the real gross domestic product (GDPG) which is a measure of actual output in the economy scaled by level of inflation. This measures the volume of economic activity in real terms at the specified time period. GDPG is expected to show a positive association with banks’ earnings. Several studies have established a positive relationship between growth of real output and most of the bank performance variables.

Inflation (INF): Using inflation as a macroeconomic variable is based on the hypothesis that the growth of the economy improves bank credit portfolio quality and an inflationary trend tend to increase market rate which in turn leads to a decrease in credit extended by banks. It is thus expected that an increase in inflation would have a negative impact on the net worth of banks.

Foreign Capital Flows (FKF): This is measured by the ratio of foreign direct investment to gross domestic product. An increase in foreign capital flows would result in real exchange rate appreciation with the subsequent problem of potential overvaluation. Foreign capital flow is expected to be positively related to growth in the economy.

Interest Rate (INT): In this study, the prime lending rate is used as a proxy for interest rate (INT). The choice of this monetary policy rate is hinged on the fact that the rate is the fulcrum upon which other rates revolve in a mixed economy and a very good proxy for short term interest which constitutes the bulk of maturity profile of lending activities of Nigerian banks. An increase in interest rate is expected to have negative impact on bank asset quality because lending becomes expensive to the borrowers and consequently reduce their debt servicing ability. However, there is an expectation that interest rate will positively affect bank earnings.

Exchange Rate (EXR): For this study, the exchange rate (EXR) used is the real effective exchange rate measured by the product of the nominal effective exchange rate and the effective relative price indices. According to the theory of effective exchange rate, an increase in real effective exchange rate value is expected to cause a deterioration of bank asset quality with subsequent decline in inflation. The choice of real effective exchange rate is justified because of the nature of the Nigerian economy as an import dependent economy.

Trade (TRD): Trade is measured by the ratio of trade to gross domestic product. The fluctuations in trade balances are usually affected by economic growth and the exchange rates which may result in either a deficit or a surplus. The impact of trade on bank performance is highly dependent on the exposure of bank intermediation to external economy through the lending components of their activities.

4. RESULTS AND DISCUSSION

4.1 DATA

This paper examined the interactive influence that exists between bank performance measures such as liquidity ratio (LQR), profitability (ROA) and capital adequacy (CAD) and some macroeconomic variables such as the Gross Domestic Product Growth (GDPG) which is a measurement of economic size, inflation (INF), foreign capital flows (FKF), interest rate (INT), exchange rate (EXR) and trade (TRD). Data used in the study span the period from 1981 to 2014 and were obtained from the World Development Indicators for Nigeria and Central Bank of Nigeria Statistical Bulletin for various years. Table 1 presents the descriptive statistics of the variables considered in this study.

4.2 EMPIRICAL RESULTS

Before carrying out our empirical analysis, it is important to first ascertain the stationary properties of the variables. In this study, we employ the Augmented Dickey Fuller (ADF) and Phillip-Perron test (PP) unit root tests. The result of the unit root test reveals that we have a mixture of I(0) and I(1) variables, that is, while some variables are stationary at level, others became stationary at their first difference, thus making the ARDL technique the appropriate analytical tool. The result of the unit root tests is presented in Table 2.

In this study, we employ three measures of bank performance, liquidity ratio (LQR), profitability (ROA) and capital adequacy (CAD). In view of this, three different measures of bank performance will each serve as dependent variables in this study. As we stated earlier, the optimal lag length used is based on the AIC criterion. Having established that we have a mixture of I(0) and I(1) variables and consequently adopting the ARDL technique, we proceed to carry out the bound test. From the result of the bound test presented in Table 3, it is evident that there is only evidence of a long run relationship when profitability (ROA) and capital adequacy (CAD) are the dependent variables. For the ROA and CAD measures of bank performance, the bound test result reveals that the F-statistic (6.27 and 4.40) exceed the upper critical bound at 1% and 5% significance levels respectively. Based on this result, we conclude that two bank performance measures, ROA and CAD are cointegrated with the macroeconomic variables of interest. To put it simply, we conclude that a long run relationship exists among two bank performance measures, ROA and CAD and the macroeconomic variables of interest. Thus, our long run and short run analysis will only be based on the ROA and CAD measures of bank performance.

Next, we looked at the results of the long run and short run analysis. When ROA serves as the measure of bank performance, empirical results report that only trade (TRA) has a significant positive impact on bank performance. While the impact of GDP growth rate (GDPG) and inflation (INF) are positive but insignificant. Foreign capital flows (FKF), interest rate (INT) and exchange rate (EXR) had insignificant negative impacts. In specific terms, we expect a 1% increase in trade (TRA) to boost bank performance by 0.04. In the short run, GDP growth rate (GDPG), at current year value, had a significant negative impact on bank performance at the 10% significance level while exchange rate and trade (at lag 1) has a significant positive impact. We however report a 10% significance level for trade (TRA). Also, the coefficient of the error correction term (ECM) reports that approximately 68% deviation from the long-run equilibrium level of bank performance is corrected for annually. We further report results for the ARDL diagnostic tests. The results of the diagnostic tests reveal that the model passes the diagnostic tests for serial correlation, functional form, normality and heteroscedasticity. Our study also assessed the stability of the estimated ARDL model with tests of CUSUM and CUSUMsq. The plots of the CUSUM and CUSUMSQ reported in Figure 1 reveals that the models are stable as the graph lies within critical bounds at a 5% level of significance.

When capital adequacy (CAD) serves as the dependent variable, empirical results reveal that only GDP growth rate (GDPG) and trade (TRA) has a significant positive impact on bank performance in Nigeria. Inflation (INF) also reported a positive impact, the effect is however insignificant while the impact of interest rate (INT) is negative and significant. In addition, foreign capital flows and exchange rate also reported negative impacts, their impacts are however insignificant. In specific terms, a 1% increase in GDP growth rate (GDPG) and trade (TRA) will boost bank performance by 0.97 and 0.47 respectively while a unit change in interest rate will reduce bank performance by 1.18. Short run results reveal that the previous year value of capital adequacy (CAD) and GDP growth rate (GDPG) (at lag1) impacts bank performance negatively, although their impacts are only significant at the 10% level. The coefficient of the error correction term (ECM) reports that approximately 64% deviation from the long-run equilibrium level of bank performance is corrected for annually. The results of the diagnostic tests also reveal that the model passes the diagnostic tests for serial correlation, functional form, normality and heteroscedasticity.

Figure 1:CUSUM and CUSUM of squares

Figure 2:CUSUM and CUSUM of squares

From the results presented, there is ample evidence to conclude that bank performance is largely determined by the prevailing macroeconomic conditions and realities. Even though the empirical impacts of some determinants are somewhat inconclusive due to their insignificant impacts, nonetheless, we have been able to establish key macroeconomic variables that are largely beneficial for bank performance in Nigeria. In specific terms, our results reveal that trade, GDP growth rate and interest rate are the most important determinants of bank performance in Nigeria. While trade and GDP growth rate promotes bank performance, rising interest rate is on the other hand associated with a decline in bank performance. For the other variables, their impacts are either negative or positive though insignificant. We can therefore not make any definite conclusions.

What has clearly emerged from our empirical findings is that GDP growth rate and trade matter the most in how well banks perform in Nigeria. This is not far-fetched because the export of primary products especially, crude oil, accounts for a larger share of government revenue as well as being an important source of foreign exchange. In addition, a lot of banking activities revolves around trade and the financing of exports and imports. It is therefore expected that rising trade levels would improve bank profits and other areas related to bank operations in Nigeria. The positive impact of GDP growth rate also implies that as the productive capacity of a nation improves, real income would also rise, and banks would therefore be more optimistic to lend to customers and do other banking related activities. To put it simply, an economy that is growing represents an improvement in economic conditions. This means as production is rising, incomes are higher and there is higher demand for goods and services. The banking system is part of the economy, therefore, an improvement in key sectors of the economy would mean that the productive activities the banks engage in and finance would also yield greater returns. Finally, we also report that higher interest rate impedes bank performance. A higher interest rate discourages people with productive ideas from seeking loans. Banks will therefore finance fewer investments with its attendant impact on bank performance. Our results are largely in line with those reported by Festic and Beko (2008), Akani, Nwanna, and Mbachu (2016) and Gikombo and Mbugua (2018).

5. CONCLUSIONS

This study empirically examined the impact of macroeconomic variables on bank performance in Nigeria with a view to establishing the variables that are the most important predictors of bank performance. In this study, we considered the following macroeconomic variables, GDP growth rate, inflation, foreign capital flows, interest rate, exchange rate and trade as the key explanatory variables of interest while liquidity ratio (LQR), return on assets (ROA) and capital adequacy (CAD) served as the measures of bank performance. For our empirical analysis, we employed the ARDL econometric technique.

Empirical results revealed the existence of a long run relationship only when ROA and CAD are the dependent variables. Results for the long run and short run relationship also revealed that GDP growth rate, interest rate and trade are the most important predictors of bank performance in Nigeria. To put it simply, we found that GDP growth rate and trade promotes bank performance while higher interest rate impedes bank performance. Another interesting result is that foreign capital flows did not have a significant impact on bank performance. This is however not surprising given the fact that very little FDI goes to the banking sector. In addition, capital flows are fickle and volatile.

The main take away from this study is that GDP growth rate, trade and interest rate are the most important determinants of bank performance in Nigeria. Given the key role trade plays in the Nigerian economy and the robust GDP growth rate that was recorded up till 2015, the result reported is not surprising. The policy implications of these results cannot be overemphasized. Nigeria has been enjoying a robust GDP growth and trade volumes up until 2015, the situation has however been different in subsequent years. We therefore recommend that government initiate policies to improve GDP growth rate and trade to ensure these variables continue to impact bank performance positively. Another important finding is that higher interest rate impedes bank performance. Government should therefore adopt a lesser interest rate regime to encourage people with productive ideas. The study also suggests that for further research, future studies should account for more macroeconomic variables that can influence bank performance in Nigeria.