EuroEconomica

Issue 1(38)/2019 ISSN: 1582-8859

Banking System Stability and Economic

Growth in Nigeria: A Bounds Test to Cointegration

Gamaliel Oghenerugba Eweke1

Abstract: This research examined the impact of banking system stability on the Nigerian economy alongside
key macroeconomic variables. The study employed banking stability index, return on assets, financial depth
and interest rate, while real GDP was used to capture economic growth, using annual data from 1986 to 2016.
The Augmented Dickey Fuller (ADF) and Phillip Perron (PP) tests reveals that apart from interest rate, all other
variables were stationary at first difference. The Bounds test to cointegration confirms the existence of a long-
run relationship amongst the variables considered for the study. The ARDL results suggests that in both long
and short-run estimations that a rise in banking sector stability, financial depth and return on assets will lead to
an increase in economic growth, conversely, an increase in interest rate will result to a fall in economic growth.
Finally, we recommend that regulators improve both the micro-prudential and the macro-prudential supervision
of the banking industry, while an upward review of the current minimum capital base has become imperative
owing to the effect of inflation and fall in the country’s exchange rate.
Keywords: Banking Stability Index; Return on Asset; Financial Depth; Interest Rate
174
JEL Classification: G24

1. Introduction
A nation’s financial system which is usually dominated by its banking sector, plays a very critical and
pivotal role in the smooth functioning of her economy. Banks through the vital function of financial
intermediation have over the years helped to move idle funds from the surplus units to the deficit units
of the economy thus helping to reduce the cost of transaction and information asymmetry. Through the
transformation of small-sized, low-risk and highly liquid customers deposits (bank liabilities) into bank
loans (bank assets), which are of larger size, higher risk and illiquid banks are able to perform what is
regarded as “transforming function”. This ultimately reconciles the varied needs of depositors (lenders)
and borrowers (spenders).
Many economists have acknowledged that the financial system, with banks as its major component,
provide linkages for the different sectors of the economy and encourage high level of specialization,
expertise, economies of scale and a conducive environment for the implementation of various economic
policies of government intended to achieve non-inflationary growth, exchange rate stability, balance of
payments equilibrium and high levels of employment (Sanusi, 2011). However, the trajectory of the
development of the Nigerian banking sector has over the years been characterized by numerous

1
Federal University Otuoke, Nigeria, Corresponding author: apache664@gmail.com.

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
fluctuations and instabilities which can be traced to 1892 when the business of banking really
commenced in Nigeria (Babalola, 2011).
The history of the Nigerian banking sector has over the years witnessed the establishment and extinction
of several banking institutions in Nigeria. In 1952, the banking ordinance was promulgated and this
marked the beginning of the regulated banking era. Prior to this legislation, the banking sector in Nigeria
was, in a phase, popularly referred to as the free banking era where the industry was left with little or no
regulation. The ordinance was designed to prevent non-viable banks from mushrooming and to ensure
orderly and viable commercial banking. Although banking ordinance triggered a rapid growth in the
industry, the growth was accompanied with disappointment as only 4 out of 25 indigenous banks
established between 1952 and 1958 survived while 21 others went under. Owing to this sorry trend and
in a bid to forestall further failures, the Central Bank Act was promulgated in 1958 so as to increase the
level of regulation and supervision of Banks in Nigeria. However, between 1994 and 2003 the country
witnessed another outbreak of bank failures culminating in withdrawal of the licenses of a good number
of banks by the Central Bank of Nigeria (CBN). The subsequent liquidation by the National Deposit
Insurance Company (NDIC) swept away fourteen more banks by the 2004 banking sector reforms.
In 2009, the Nigerian banking sector experienced another rounds of reforms which saw the exposure of
a humongous sum of non-performing loans and the subsequent collapse of 9 out of the 24 banks in the
country. Sanusi (2010) highlighted 8 main interdependent factors which led to the creation of an
extremely fragile financial system, namely: macro-economic instability caused by large and sudden
capital inflows, major failures in corporate governance at banks, lack of investor and consumer
175
sophistication, inadequate disclosure and transparency about financial position of banks, critical gaps in
regulatory framework and regulations, uneven supervision and enforcement, unstructured governance
& management processes at the CBN/Weaknesses within the CBN, weaknesses in the business
environment. Hence the Central Bank of Nigeria has made concerted effort via several banks reforms
especially from the wake of last decade through effective surveillance and prudential guidelines, a more
stringent procedure for licensing and increase in the capitalization base of the banks, among others. This
was meant to ensure a sound and stable banking system capable of providing effective intermediation
that would stimulate growth, encourage medium and long term lending to the real sectors capable of
diversifying the productive base of the economy. (Iwedi & Igbanibo, 2015)
Accordingly, apart from the introductory section, this paper is organised into four sections with the
second section considering the review of literature, third section the methodology, the fourth section
focuses on the results and discussion of findings while the fifth section concludes and makes
recommendations.

2. Review of Literature
2.1. Theoretical Review
Micro-Prudential Approach
The micro prudential regulation assumes a partial-equilibrium condition and is aimed at averting the
failure of individual financial institutions. According to Sere-Ejembi, Udom, Salihu, Atoi and Yaaba

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
(2014), the paradigm of micro-prudential supervision views that risks arise from individual malfeasance.
Therefore, micro-prudential regulation focuses on the stability of the components of a financial system.
The regulation seeks to enhance the safety and soundness of individual financial institutions by
supervising and limiting the risk of distress. The principal focus is to protect the clients of the institutions
and mitigate the risk of contagion and the subsequent negative externalities in terms of confidence in
the overall financial system.
Macro-Prudential Approach
The macro prudential approach, on the other hand, adopts the general-equilibrium condition and is aimed
at safeguarding the entire financial system (Charles, 2015). Macro prudential policies aim to increase
the overall resilience of the financial system, contain the build-up of systemic risk over time. It is also
reputed to address vulnerabilities stemming from structural relationships between financial
intermediaries. (Ananthakrishnan, Heba & Pilar, 2016) The macro-prudential approach argues that
safety and soundness of the entire financial system is not necessarily guaranteed by the safety and
soundness of the individual financial institutions. In fact, there are times when individual actions of the
financial institutions aimed at keeping such institutions safe and sound may pose dangers to the stability
of the entire system. (Charles, 2015) According to Ananthakrishnan, Heba and Pilar, (2016), a macro
prudential policy framework should ideally encompass:
(i) A system of early warning indicators that signal increased vulnerabilities to financial stability;
(ii) A set of policy tools that can help contain risks ex ante and address the increased vulnerabilities
176 at an early stage, as well as help build buffers to absorb shocks ex post; and
(iii) An institutional framework that ensures the effective identification of systemic risks and
implementation of macro prudential policies.
Micro and macro-prudential supervisions are interlinked. Macro-prudential supervision cannot achieve
its objective except it has some level of impact on supervision at the micro-level.
2.2. Empirical Review
Monnin and Jokipii (2010), studied the relationship between the degree of banking sector stability and
the subsequent evolution of real output growth and inflation. Adopting a panel VAR methodology for a
sample of 18 OECD countries, they found a positive link between banking sector stability and real output
growth. This finding is predominantly driven by periods of instability rather than by very stable periods.
Laeven and Valencia (2012) presented descriptive statistics on the frequency of banking crises, their
resolution, and their real effects. They identified 147 banking crises, over the period of 1970 to 2011.
Results showed that advanced economies tend to experience larger output losses and increases in public
debt than emerging and developing countries. These larger output losses in advanced economies were
to some extent driven by deeper banking systems, which makes a banking crisis more disruptive.
Dell’Ariccia, Detragiache and Rajan (2008) studied the effects of banking crises on growth in industrial
sectors and found that while adverse shocks cause both poor economic performance and bank distress,
bank distress has an additional, adverse effect on growth, as banks must cut back their lending, and that
the differential effect is stronger in developing countries (where alternatives to bank financing are more
limited), in countries with less access to foreign finance, and where bank distress is more severe.
FINANCE, BANKING AND ACCOUNTING
 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
Demirguc-Kunt and Detragiache (1998) developed a model which identified a group of macroeconomic
variables that consist of high interest rate, inflation, output downturns, decline in asset prices, adverse
terms of trade, credit expansion, foreign exchange reserve’s losses and market pressure. These were
reported to have affected the financial system as a whole, using a multivariate logit framework and
considering both industrial and emerging market economies. It was discovered that the characteristics
of the banking sector and structural characteristics of the country were robustly correlated with the
emergence of banking sector crisis.
Sere-Ejembi (2014) constructed a Banking System Stability Index (BSSI) for Nigeria, using a
combination of financial soundness indicators and macro-fundamentals. It applied statistical and
Conference Board Methodology normalization processes on Nigeria’s banking and macroeconomic data
from the first quarter of 2007 to the second quarter of 2012. They discovered that the resulting index
traced fairly well the episodes of crisis in the system over the study period and thus concluded that the
BSSI is capable of acting as an early warning mechanism of signaling fragility and could be used as a
complimentary regulatory policy tool to detect potential threats to enable monetary authorities take
timely pre-emptive policy measures to avert crisis. Barro (2001) examined the impact of a banking crisis
on growth. They employed data from 67 industrialized and emerging countries (five-year averages) and
the panel data approach was adopted. Results showed that a banking crisis reduced GDP per capita
growth rate of GDP of 0.6% per annum and the investment rate of 0.9%.
Kupiec and Ramirez (2010) investigated the effect of bank failures on economic growth using data on
bank failures ranging from 1900 to 1930. The sample period predated active government stabilization
177
policies and included several severe banking crises. The VAR and difference-in-difference methods
were applied to estimate the impact of bank failures on economic activity. VAR results show bank
failures have negative and long-lasting effects on economic growth. While the difference-indifference
results suggest that bank failures trigger an increase in non-bank failures. The evidence showed that
bank failures reduce economic growth and provides a lower bound estimate of the cost of banking sector
systemic risk. Soundness (i.e. reserve for money bank deposits and ratio of net foreign assets to GDP)
are the factors most likely to influence its stability. Jide (2003) designed an early-warning bank failure
model that captured the dynamic process underlying the banking sector slide from soundness to closure,
by employing a transition probability matrix. The study used “Instrumental Variables-Generalized
Maximum Entropy formalism” to assess the likelihood of the banking sector experiencing distress via
the evaluation of banking crisis probabilities.
Although several studies have examined the impact of the banking sector and the financial system on
the growth of a nation’s economy and the cost of bank failures on the economy, very few have examined
the impact of fluctuations in banking sector stability indicators on economic growth. This study therefore
seeks to fill this gap by ascertaining how banking sector stability impact on the growth of the Nigerian
economy using the average figures of statistically normalized values of selected banking sector
indicators.

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
2.3. Theoretical Framework
This study draws inspiration from the works of Akpan (2017) and Sere-Ejembi, Udom, Salihu, Atoi and
Yaaba (2014) in developing a banking sector stability index for Nigeria. The index can be determined
from:
Equation 2.1
(𝑥𝑡 − 𝜇)
𝑍𝑡 =
𝛿
Where Xt represents the value of indicators X during period t; µ is the mean and δ is the standard
deviation.
Equation 2.2
𝐶𝑃𝑆𝑡 − 𝜇𝐶𝑃𝑆 𝐷𝐸𝑃𝑡 − 𝜇𝐷𝐸𝑃 𝐹𝐿𝑡 − 𝜇𝐹𝐿
𝛿𝐶𝑃𝑆 + 𝛿𝐷𝐸𝑃 + 𝛿𝐹𝐿
𝐵𝑆𝑆𝐼_3𝑡 =
3
Equation 2.3
(𝐶𝑃𝑆𝑡 − 𝜇𝐶𝑃𝑆)
𝐶𝑃𝑆𝑡 =
𝜎𝐶𝑃𝑆
Equation 2.4
178 (𝐷𝐸𝑃𝑡 − 𝜇𝐷𝐸𝑃)
𝐷𝐸𝑃𝑡 =
𝜎𝐷𝐸𝑃
Equation 2.5
(𝐹𝐿𝑡 − 𝜇𝐹𝐿)
𝐹𝐿𝑡 =
𝜎𝐹𝐿
Where:
BSSI_3 = Banking system stability index (indicator)
CPSt = Bank claims on (credit to) the domestic private sector at a point in time
DEPt = Bank deposits at a point in time
FLt = Foreign liabilities of banks
µ= Arithmetic mean
δ= Standard deviation
The BSSI_3 measures the swings in the domestic banking system. A higher index (i.e. BSSI3≥50%)
indicates a stable system and a lower index (i.e. BSSI3 ≤ 49%) indicates a fragile system.

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
3. Econometric Procedure
This study uses the Autoregressive Distributed Lag (ARDL)/Bounds Test methodology proposed by
Pesaran and Shin (1999) and Pesaran, Shin and Smith (2001) to estimate the dynamic, long and short-
run relationship among the variables. This technique has advantages over other cointegration techniques.
Whereas other cointegration test requires that all variables to be integrated of the same order, the ARDL
technique can be applied whether the variables are purely 1(0) and/or purely 1(1) or a mixture of 1(0)
and 1(1) variables. Furthermore, the bounds test approach within the ARDL framework performs better,
as it gives more robust results in small samples than the Johansen cointegration technique which requires
a large data sample to obtain a valid result (Pesaran, Shin & Smith, 2001). Likewise, endogeneity
problems are tackled in this technique. According to Pesaran and Shin (1999), they contended that
modelling the ARDL with the appropriate lags will correct for both serial correlation and endogeneity
problems. From the variables of interest, the following model has been specified;
Equation 3.1
RGDP= f (BSSI_3 ROA FIN_D INT)
Where;
RGDP is the Real Gross Domestic Product deflated by the general price level.
BSSI_3 refers to Banking system stability index
ROA refers to Return on Asset. This is used to measure the performance of the banking industry.
179
FIN_D refers to Financial Depth. This captures the financial sector relative to the economy. It is the size
of banks, other financial institutions, and financial markets in a country, taken together and compared
to a measure of economic output.
INT represents Interest Rate. This can be defined as the cost of borrowing.
To confirm linearity and also deal with heteroscedascity, a double log-linear model was specified;
Equation 3.2
logRGDPi,t = βo + β1logBSSI_3 + β2logROA + β3logFIN_D + β4logINT + εi,t
Consequently, upon applying the ARDL methodology, it becomes imperative we specify the ARDL
representations of equation 3.2 as:
Equation 3.3
∆𝑙𝑜𝑔𝑅𝐺𝐷𝑃𝑡 = 𝛼0 + 𝛽1 𝑙𝑜𝑔𝐵𝑆𝑆𝐼_3𝑡−1 + 𝛽2 𝑙𝑜𝑔𝑅𝑂𝐴𝑡−1 + 𝛽3 𝑙𝑜𝑔𝐹𝐼𝑁_𝐷𝑡−1 + 𝛽3 𝑙𝑜𝑔𝐼𝑁𝑇𝑡−1
𝑛 𝑛 𝑛

+ ∑ 𝜑ℎ ∆ 𝑙𝑜𝑔𝑅𝐺𝐷𝑃𝑡−1 + ∑ 𝜙𝑖 ∆ 𝑙𝑜𝑔𝐵𝑆𝑆𝐼_3𝑡−1 + ∑ 𝜆𝑗 ∆ 𝑙𝑜𝑔𝑅𝑂𝐴𝑡−1

𝑗=1 𝑗=1 𝑗=1
𝑛 𝑛

+ ∑ ⍹𝑘 ∆ 𝑙𝑜𝑔𝐹𝐼𝑁_𝐷 + ∑ ⍴𝑙 ∆ 𝑙𝑜𝑔𝐼𝑁𝑇𝑡−1 + 𝜈𝑡
𝑗=1 𝑗=1

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
Where Δ signifies the first difference operator, α0 is the intercept, β1 β2 β3 are the long-run multipliers. δ,
ϕ, λ, ⍹ and ⍴ are short-run parameters and νt are white noise errors. This study estimated equation (3)
with the bounds test in other to access the long-run relationship. The F-test was used to interpret the
existence of a long-run relationship amongst the variables in equation (3). The null hypothesis of no
long-run relationship in equation (3) is tested against the alternate hypotheses of a long-run relationship
as shown below;
H0: α = β1 = β2 = 0
H1: α ≠ β1 ≠ β2 ≠ 0
The bounds test provides for two asymptotic critical value for cointegration when the dependent variables
are 1(d) (where 0≤d≤1): a lower value assuming the regressors are I(0) and an upper value assuming
purely I(1) regressors. If the F-statistic is above the upper critical value, the null hypothesis of no long
run relationship can be rejected regardless the orders of integration for the time series. Inversely, if the
F-statistic falls below the lower critical value, the null hypothesis cannot be rejected. Finally, if the
statistic falls between the lower and upper critical values, the result is inconclusive. The approximate
critical values for the F-statistic test were obtained from Pesaran et al (2001).
Immediately cointegration is detected the ARDL long-run model for RGDPt can be estimated as:
Equation 3.4
𝑛 𝑛 𝑛

180 𝑙𝑜𝑔𝑅𝐺𝐷𝑃𝑡 = 𝛼0 + ∑ 𝜑ℎ ∆ 𝑙𝑜𝑔𝑅𝐺𝐷𝑃𝑡−1 + ∑ 𝜙𝑖 ∆ 𝑙𝑜𝑔𝐵𝑆𝑆𝐼_3𝑡−1 + ∑ 𝜆𝑗 ∆ 𝑙𝑜𝑔𝑅𝑂𝐴𝑡−1

𝑗=1 𝑗=1 𝑗=1
𝑛 𝑛

+ ∑ ⍹𝑘 ∆ 𝑙𝑜𝑔𝐹𝐼𝑁_𝐷𝑡−1 + ∑ ⍴𝑙 ∆ 𝑙𝑜𝑔𝐼𝑁𝑇𝑡−1 + 𝜈𝑡
𝑗=1 𝑗=1

The next step is to obtain the short-run dynamic parameters by estimating an error correction model
within the ARDL framework. Thus specified as:
Equation 3.5
𝑛 𝑛 𝑛

∆𝑙𝑜𝑔𝑅𝐺𝐷𝑃𝑡 = µ0 + ∑ 𝜑ℎ ∆ 𝑙𝑜𝑔𝑅𝐺𝐷𝑃𝑡−1 + ∑ 𝜙𝑖 ∆ 𝑙𝑜𝑔𝐵𝑆𝑆𝐼_3𝑡−1 + ∑ 𝜆𝑗 ∆ 𝑙𝑜𝑔𝑅𝑂𝐴𝑡−1

𝑗=1 𝑗=1 𝑗=1
𝑛 𝑛

+ ∑ ⍹𝑘 ∆ 𝑙𝑜𝑔𝐹𝐼𝑁_𝐷𝑡−1 + ∑ ⍴𝑙 ∆ 𝑙𝑜𝑔𝐼𝑁𝑇𝑡−1 + 𝜗𝐸𝐶𝑇𝑡−1 + 𝜈𝑡

𝑗=1 𝑗=1

Where ϑ denotes the speed of adjustment of the parameters to the long-run equilibrium following a shock
to the system and ECTt-1 represents the residuals obtained from equation (5). Furthermore, the coefficient
of the lagged error correction term ϑ is expected to be negative and statistically significant to further
confirm the existence of a cointegrating relationship.

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
4. Results and Discussion of Findings
4.1. Unit Root Test
Before estimating the Bounds test to cointegration, unit root test would be conducted to examine the
stationarity process of the variables to ensure that none of the variables are integrated of order two, 1(2)
to avoid spurious results. This is necessary because the computed F-statistics by Pesaran, Shin and Smith
(2001) are not valid in the presence of 1(2) variables. The study utilized the Augmented Dickey Fuller
(ADF) and Phillip Perron (PP) test to access the order of integration amongst the variables. From Table
1, all variables were stationary at I(1) apart from interest rate which was stationary at levels.
Table 1. Unit Root Test Results
Variables ADF Test Remarks PP Test Remarks
logRGDP -3.229346** 1(1) -3.044705** 1(1)
logBSSI_3 -11.54228* 1(1) -11.54228* 1(1)
logFIN_D -4.586771* 1(1) -4.814938* 1(1)
logINT -4.228115* I(0) -4.228115* I(0)
logROA -5.067558* I(1) -5.180230* I(1)

Critical Values of ADF Test: Critical Values of PP Test:

1% level = -3.639407 1% level = -3.639407
5% level = -2.951125 5% level = -2.951125
10% level = -2.614300 10% level = -2.614300 181
*/**/***, indicates significance at 1%, 5% & 10% respectively.
Test includes Trend and Intercept
Source: Author’s Computation Using Eviews 10+

4.2. Bounds Test

In other to examine the presence of a long-run relationship among the variables, we therefore proceed to
estimate equation (3). A maximum of one (1) lag length was selected based on the Akaike info criterion
(AIC). According to Table 2, the F-statistic for the model with a value of 15.39791 exceeds the upper
critical bound at 10% significance level. We therefore reject the null hypothesis of no cointegration. This
indicates the existence of a long-run relationship between economic growth and its explanatory variables.
Table 2. ARDL Bounds Test
Null Hypothesis: No levels
F-Bounds Test relationship
Test Statistic Value Signif. I(0) I(1)
F-statistic 15.39791 10% 2.2 3.09
K 4 5% 2.56 3.49
2.5% 2.88 3.87
1% 3.29 4.37
Source: Author’s Computation Using Eviews 10+

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
4.3. Long-Run Estimates
Since the variables are cointegrated, we therefore proceed to estimate equation (4). From Table 3, the
results obtained by normalizing the explanatory variables on economic growth in the long-run, indicates
that banking system stability, bank performance and financial depth has a positive but non-significant
effect on economic growth in Nigeria. However financial deepening has a negative but non-significant
on economic growth in Nigeria. The result further reveals that an increase in banking sector stability
index would lead to a increase in Real GDP, similarly an increase in Return on Asset of the Nigerian
Banking Industry will lead to a rise in Real GDP.
Table 3. Estimated Coefficients of the Long-Run Model
Dependent Variable: RGDP
Variable Coefficient Std. Error t-Statistic Prob.
BSSI_3 0.492419 0.686133 0.717674 0.4822
LOG(FIN_D) 4.028961 4.435821 0.908279 0.3757
LOG(INT) -0.076497 2.000522 -0.038238 0.9699
ROA 0.159930 0.250344 0.638841 0.5310
C 0.667774 13.84699 0.048225 0.9621

Source: Author’s Computation Using Eviews 10+

4.4. Short-Run Dynamics
The study further estimates the short-run relationship among the variables. According to Table 4, the
coefficient of the lagged error correction term (ECMt-1) is of the expected negative sign and significant
182 at 1% with economic growth. The ECM captures the speed of adjustment to restore equilibrium in case
of any shock to any of the exogenous variables. The coefficient of the error term, -0.024211 which is
significant at 1% level, indicates that about 2.42% of disequilibrium from previous year’s shock in
economic growth converges back to the long-run equilibrium within the current year. This suggests a
very low speed of adjustment in the model.
Finally, the results obtained from the short-run estimates buttresses the position of the long-run model
as all the coefficients of the model has the same signs attached as found in the long-run model.
Table 4. Estimated Coefficients of the Short-Run Dynamic Error Correction Model
Dependent Variable: RGDP
Variable Coefficient Std. Error t-Statistic Prob.
C 0.016167 0.353824 0.045693 0.9641
ΔLOG(RGDP) t-1 -0.024211 0.030429 -0.795637 0.4366
BSSI_3** 0.011922 0.019544 0.609990 0.5495
LOG(FIN_D) t-1 0.097544 0.033644 2.899296 0.0096
LOG(INT) t-1 -0.001852 0.049391 -0.037498 0.9705
LOG(ROA) t-1 0.003872 0.003189 1.214053 0.2404
DLOG(FIN_D) t-1 0.034592 0.032988 1.048615 0.3082
ECTt-1 -0.024211 0.002228 -10.86511 0.0000

Source: Author’s Computation Using Eviews 10+

4.5. Model Diagnostics

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
To ensure that the model is correctly specified and to avoid spurious results, it is therefore mandatory to
examine for model misspecification which may occur due to unstable parameters and afterward lead to
bias estimates. From Table 3, the test statistics with its antecedent p-values > 10% significance level
indicates that the model is free from Serial Correlation and Heteroskedasticity Likewise, the Jarque-Bera
test statistics (0.9724) indicates that the model residuals are normally distributed.
Furthermore, from Appendix 1, the R2 with a value of 0.996956 indicates that 99.70% of the variation in
economic growth is explained by banking sector stability, return on asset, interest rate, financial depth
and one-period lag of real GDP, while the standard error of 0.028987 signifies that about 2.89% of
variations in economic growth will not be explained by the independent variables. The Durbin-Watson
statistics of 2.109927 confirms the results of the ARCH test indicating the absence of serial correlation.
The Akaike Info Criterion value of -4.012449, suggests that information loss is well minimized by the
model. The F-Statistics value of 982.5652 indicates that the overall model is significant at 1% level and
is a good fit.
The CUSUM and CUSUMQ of recursive residuals test as suggested by Pesaran and Pesaran (1997) was
used to access the coefficient stability in the model. From Appendix 2, the plot of the CUSUM and
CUSUMQ of recursive residual stability test indicates that all estimated coefficients of the model are
stable over the study period since they are within the 5% critical bounds.
Table 5. Diagnostics
Diagnostic Test Test Statistics P-value
Serial Correlation (Breusch-Godfrey) 0.731923 0.4026 183
Heteroskedasticity (ARCH) 0.014570 0.9051
Normality (Jarque-Bera) 1.790171 0.9724
Source: Author’s Computation Using Eviews 10+

5. Conclusion and Recommendation

The aim of this study was to examine the intertwining relationship between banking sector stability and
economic growth amidst other macroeconomic variables in Nigeria. The study estimates both the long
and short run models using the ARDL/Bounds Test framework using data from 1986 – 2016. Both the
Augmented Dickey Fuller and Philip Perron’s test suggested that none of the variables where integrated
of order two i.e. I(2), while the bounds test indicated the presence of a long-run relationship among the
variables.
The findings of the study indicated that in both long and short-run estimations that an increase in banking
system stability index (BSSI_3), return on asset (ROA) and financial depth (FIN_D) will lead to a rise
in economic growth (RGDP) though not significant. Conversely, the impact of Interest rates on the
Nigeria Economy although negative and insignificant, suggests that a rise in the banking lending rate is
unhealthy for the Nigerian economy. This non-significance of the long-run impact of banking sector
performance and banking sector stability on the growth of the Nigerian economy this could be attributed
to the high level of instability that has plagued the banking sector, and inadequate loans and advance
from banks to the private sector owing to the fact that banks have over the years focused on raking in

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
profits rather than assisting to provide funding for small and medium scale enterprises which have the
potential to significantly stimulate economic growth in Nigeria.
To this extent the study therefore recommends that:
i) There should be an increased and concerted effort on the part of regulators to improve both the
micro-prudential and the macro-prudential supervision of the industry;
ii) There is need for an upward review of the current minimum capital base as it has become
inadequate owing to the effect of inflation and fall in the country’s exchange rate;
iii) Also there is need for a strict implementation of the recommendations of the Basel accord in
order to improve the health and international competitiveness of Nigerian Banks;
iv) Non-performing loans and other fictitious assets and revenue have over the years constituted a
large portion of the reported assets of banks in the country, banks made public information on
their operations on a highly selective basis, thus giving a misleading view of the performance of
the industry to regulators, investors and the general public at large. Hence there is a need for a
stricter enforcement of financial reporting standards which would help enhance the data quality
in banks to ensure their reports are accurate, also the time period taken to declare a loan as bad
should be contracted so at to reduce the number of non-performing loans in the industry;
v) Furthermore banks should be encouraged to increase their loans and advances to the real sector
at lower interest rates, there should be a regulatory framework that will ensure that banks channel
184 their resources to the viable sectors of the economy with potential to grow the economy;
vi) Although, all the banks in Nigeria agreed to set aside 10 percent of their profit before tax for
equity investments in small scale industries in order to stimulate economic growth and reduce
the growing rate of unemployment in the country, banks have however being are reluctant to
release the fund owing to the inability of the local entrepreneurs to provide collateral and good
feasibility study hence the collateral bottlenecks associated with the procurement credit facilities
should be reduced.

Appendix 1. Autoregressive Distributed Lag Model

Dependent Variable: LOG(RGDP)
Method: ARDL
Sample (adjusted): 1987 2016
Included observations: 29 after adjustments
Maximum dependent lags: 1 (Automatic selection)
Model selection method: Akaike info criterion (AIC)
Dynamic regressors (1 lag, automatic): BSSI_3 LOG(FIN_D) LOG(INT) ROA
Fixed regressors: C
Number of models evalulated: 16
Selected Model: ARDL(1, 0, 1, 0, 0)
Variable Coefficient Std. Error t-Statistic Prob.*
LOG(RGDP(-1)) 0.975789 0.030429 32.06725 0.0000
BSSI_3 0.011922 0.019544 0.609990 0.5495

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
LOG(FIN_D) 0.034592 0.032988 1.048615 0.3082
LOG(FIN_D(-1)) 0.062953 0.035372 1.779721 0.0920
LOG(INT) -0.001852 0.049391 -0.037498 0.9705
ROA 0.003872 0.003189 1.214053 0.2404
C 0.016167 0.353824 0.045693 0.9641
R-squared 0.996956 Mean dependent var 10.41173
Adjusted R-squared 0.995941 S.D. dependent var 0.455003
S.E. of regression 0.028987 Akaike info criterion -4.012449
Sum squared resid 0.015124 Schwarz criterion -3.671164
Log likelihood 57.15562 Hannan-Quinn criter. -3.917791
F-statistic 982.5652 Durbin-Watson stat 2.109927
Prob(F-statistic) 0.000000

Appendix 2. Plot of Cumulative Sum and Cumulative Sum of Squares of Recursive Residuals Stability Tests
12
1.6

8
1.2

4
0.8
0

0.4
-4

0.0
-8

-0.4
-12
2000 2002 2004 2006 2008 2010 2012 2014
2000 2002 2004 2006 2008 2010 2012 2014 185
CUSUM of Squares 5% Significance
CUSUM 5% Significance
Source: Author’s Computation Using Eviews 10+
Appendix 3. Descriptive Statistics
1. 2. RGDP 3. BSSI_ 4. FIN_D 5. INT 6. ROA
3
7. Mean 8. 36095 9. 0.000 10. 17.87 11. 19.04 12. 3.886
.70 000 692 462 538
13. Median 14. 30333 15. - 16. 18.55 17. 18.29 18. 4.295
.58 0.115000 000 000 000
19. Maximum 20. 69023 21. 1.120 22. 38.00 23. 29.80 24. 7.350
.93 000 000 000 000
25. Minimum 26. 19199 27. - 28. 8.600 29. 13.54 30. -
.06 0.590000 000 000 5.170000
31. Std. Dev. 32. 17039 33. 0.486 34. 6.633 35. 3.447 36. 2.650
.52 818 570 662 846
37. Skewness 38. 0.642 39. 1.016 40. 1.297 41. 1.397 42. -
425 083 523 108 1.772219
43. Kurtosis 44. 1.964 45. 3.090 46. 5.139 47. 5.215 48. 6.810
002 282 393 573 028
49. 50. 51. 52. 53. 54.
55. Jarque-Bera 56. 2.951 57. 4.482 58. 12.25 59. 13.77 60. 29.33
144 674 387 610 597
61. Probability 62. 0.228 63. 0.106 64. 0.002 65. 0.001 66. 0.000
648 316 183 020 000
67. Sum 68. 93848 69. 0.000 70. 464.8 71. 495.1 72. 101.0
8.1 000 000 600 500
73. Sum Sq. 74. 7.26E 75. 5.924 76. 1100. 77. 297.1 78. 175.6
Dev. +09 800 106 592 746
FINANCE, BANKING AND ACCOUNTING
 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
79. Observations 80. 30 81. 30 82. 30 83. 30 84. 30

Source: Author’s Computation Using Eviews 10+

Appendix 4. Breusch-Godfrey Serial Correlation LM Test:
Null hypothesis: No serial correlation at up to 1 lag
F-statistic 0.731923 Prob. F(2,15) 0.4026
Obs*R-squared 1.059730 Prob. Chi-Square(2) 0.3033
Source: Author’s Computation Using Eviews 10+

Appendix 5 Heteroskedasticity Test: ARCH

F-statistic 0.014570 Prob. F(1,21) 0.9051

Obs*R-squared 0.015946 Prob. Chi-Square(1) 0.8995

Source: Author’s Computation Using Eviews 10+

Appendix 6. Residual Normality Tests
6
Series: Residuals
186 5
Sample 1992 2015
Observations 24

4 Mean 5.78e-15
Median 0.000381
3 Maximum 0.058844
Minimum -0.034433
Std. Dev. 0.021746
2
Skewness 0.600244
Kurtosis 3.590763
1
Jarque-Bera 1.790171
0 Probability 0.408573
-0.04 -0.02 0.00 0.02 0.04 0.06

Source: Author’s Computation Using Eviews 10+

Bibliography
Akpan, E.S. (2017). Banking System Stability and Return on Assets Relationship: Evidence from Nigeria. The Macrotheme
Review. 6(1), pp. 46-55.
Barro, J. (2001). Economic Growth in East Asia before and after Financial Crisis. National Bureau of Economic Research
Working paper, 8330, pp. 1-42.
Dell’Ariccia, G.; Detragiache, E. & Rajan, R. (2008). The real effect of banking crises. Journal of Financial Intermediation,
17, pp. 89–112.

FINANCE, BANKING AND ACCOUNTING

 EuroEconomica
Issue 1(38)/2019 ISSN: 1582-8859
Demirguc-Kunt, A. & Detragiache, E. (1998). The determinants of banking crises in developed and developing countries. IMF
Staff Papers, no. 45, pp. 81-109.
Iwedi, M. & Igbanibo D.S. (2015). Modeling Financial Intermediation Functions of Banks: Theory and Empirical Evidence
from Nigeria. Research Journal of Finance and Accounting, 6(18), pp. 159-174.
Jide, L. (2003). An Early Warning Model of Bank Failure in Jamaica: An Information Theoretic Approach. Financial Stability
Department, Bank of Jamaica.
Jokipii & Monnin. (2010). The Impact of Banking Sector Stability on the Real Economy. The Swiss National Bank Working
Papers, no. 5.
Kupiec, P. (2013). Bank failures and the cost of systemic risk: Evidence from 1900 to 1930. Journal of Financial
Intermediation, 22, pp. 285-307.
Laeven, L. & Valencia, F. (2012). Systemic Banking Crises Database: An Update. IMF Working Paper 12/163, pp. 1-33.
Pesaran, H.; Shin, Y. & Smith, R. (2001). Bound Testing Approaches to the Analysis of Level Relationships. Journal of Applied
Econometrics, 16(3), pp. 289–326.
Pesaran, M.H. & Shin, Y. (1999). An Autoregressive Distributed Lag Modelling Approach to Cointegration Analysis. In
Econometrics and Economic Theory in the 20th Century. The Ragnar Frisch Centennial Symposium. Cambridge: Cambridge
University Press.
Sanusi, S.L. (2011). Banks in Nigeria and national economic development – a critical review. Becoming an economic driver
while applying banking regulations, pp. 1-6. Lagos: CIBN.
Sere-Ejembi, A.U. (2014). Developing Banking System Stability Index for Nigeria. CBN Journal of Applied Statistics, 5(1),
pp. 49-77.
187

FINANCE, BANKING AND ACCOUNTING

 © 2019. This article is published under
http://creativecommons.org/licenses/by/4.0/(the “License”). Notwithstanding
the ProQuest Terms and Conditions, you may use this content in accordance
with the terms of the License.