



































Contents


Indian Journal of Finance and Banking 

Vol. 1, No. 1; 2017 

Published by Centre for Research on Islamic Banking & Finance and Business 

 

8 
 

 

Bank Performance versus Recession Indicators: A Linear 

Cointegration Approach 
 

Ashamu Sikiru Oyerinde 1 

 

1 Department of Banking and Finance, Faculty of Management Sciences, Lagos State University, Ojo, Lagos. 

Correspondence: Department of Banking and Finance, Faculty of Management Sciences, Lagos State University, 

Ojo, Lagos, Tel: +2348033713969, E-mail: soyerinde2012@gmail.com 

 

Received: July 13, 2017        Accepted: July 15, 2017      Online Published: July 17, 2017 

 

 

Abstract 

In this study, the researcher provides an empirical investigation of the nexus between banks’ performance and 

recession indicators. A sample size of 35 years was selected on annual data. A linear cointegration method was 

adopted after accounting for seasonality through logarithmic transformation. The results revealed that indicators 

of recession-exchange rate, inflation and interest rate maintain long run relationship with bank performance, and 

evidence of long run influence was established. Furthermore, we discover that within the purview of short run 

dynamic situation, inflation influences banks’ performance inversely, while exchange rate and interest rate 

increase with increase in banks’ performance. We therefore conclude that banks’ performance is driven by 

indicators of recession both in the short and in long run. 

Keywords: Banks, exchange rate, inflation, interest rate and cointegration. 

 

1. Introduction 

Over the past decade, the credit quality of loan portfolios across most countries in the world  remained  

relatively  stable  until  the  financial  crises  hit  the  global  economy  in 2007-2008.  Since  

then,  average  bank  asset  quality  deteriorated  sharply  due  to  the global  economic  recession.  

The  fact  that  loan  performance  is  tightly  linked  to  the economic  cycle  is  well  known  

and  not  surprising.  Yet the deterioration of loan performance was very uneven across countries. For 

example, the Baltic countries which stand out in cross-country comparisons of GDP performance during the 

crisis had very large increases in non-performing loans (NPLs) even when controlling for the severity of the 

recession.    

Thus, the 2007-2009 global financial turmoil affected banking systems around the world, including Nigeria. 

Nevertheless, Nigerian banks have not performed relatively well. They for example, require capital injections at 

some point in time, and deposit guarantees have to be expanded. According to Kithinji and Waweru (2007), 

banking problems is back-dated as early as 1986  culminating  in  major  bank  failures  (37  failed  

banks  as  at  1998)  following  the crises  of  1986  to  1989,  1993/1994  and  1998;  they  

attributed  these  crises  to  NPLs which  is  due  to  the  interest  rate  spread.    According  to  

(McNulty  et  al  2001), controlling  NPLs  is  very  important  for  both  the  performance  of  

an  individual  bank and  the  economy’s  financial  environment. Due to the nature of their business, 



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

9 
 

commercial banks expose themselves to the risks of default from borrowers. Prudent credit risk assessment and 

creation of adequate provisions for bad and doubtful debts can cushion the banks risk. It is established that when 

the level of nonperforming loans is very high, the provisions are not adequately protected (Waweru and Kalani, 

2009). The occurrence of banking crises has often  been  associated  with  a  massive  accumulation  of  

nonperforming  assets  which can  account  for  a  sizable  share  of  total  assets  of  insolvent  

banks  and financial institutions. Therefore, the causes of loan defaults should be established so as to reduce 

the level of nonperforming loans.  

From the foregoing, it seems there is negative relationship between recession and banks’ performance.  

However, management of recession in an economy helps improve the financial performance of banks, as interest 

on loans is the biggest income of banks, and increase the liquidity position of banks, protect the bank reputation 

and reduce loan losses. Also, Golden,  Sam  and  Harry  (1993) emphasized that since one  of  the  most  

important  aspects  of  lending  is  determining  the  customer  desire  to repay the loan, information 

in the credit file will give the credit officer document on the customer’s  repayment  history. This is 

accomplished with the fact that commercial banks have measures in place to guard against loans, taking 

mortgages on landed properties, stock, bonds and other securities to fall on in time of default. This was 

debunked by Samuel (2011) who claimed that loan default   has   become   an   inevitable   part   of   

the   banking   or   lending   business in a recessive economy, there is high default risk because the 

income of borrowers has to be reduced. Accordingly leading to repression situation and consequently reducing 

the ability to honor obligations at maturities. The argument perhaps is do banks drive well in recession, when 

exchange rate and interest rate increase sporadically. Many studies have answered this question differently. This 

growing gap has inspired me to investigate the dynamic causations that exist between recession proxies and 

bank performance in Nigeria with the aim of justifying or expanding existing positions. The rest of the paper is 

organized as literature review, methodology and data, results, conclusion and recommendations 

2. Literature Review 

The Great Recession is strongly associated with the US real estate crisis and recent work has explored the link 

between real estate stresses and bank distress.  Aubuchon and Wheelock (2010) find that both during the 

2007-2010 and the 1987-92 episodes, bank failures were concentrated in regions with the largest declines in 

economic activity (as evident in declines in personal income or gross state product, and in increases in 

unemployment) and with the most severe stresses in the real estate market (as evident in declines in house prices 

and increases in delinquency rates).  Beltratti and Stulz (2012) find that banks whose returns were more 

exposed to US real estate exhibited better stock performance during the crisis, but this effect they find to only be 

present in a subsample of very large banks (assets greater than $50 billion).  

New England experience of the early 1990s has shown that even well-capitalized banks can fail, as capital 

buffers deteriorate rapidly when banks approach failure (Peek and Rosengren (1997)). In a more recent paper, 

Berger and Bouwman (2013) exploit an exogenous source of variation in the stock of capital buffers to study the 

effect of capital on two dimensions of bank performance, probability of survival and market share, and find the 

effect to vary across banking crises, market crises, and normal times. In particular, capital increases the 

probability of survival and market share of smaller banks for all three types of crises, but improves the 

performance of medium and large banks primarily during banking crises. Though these studies advance our 

understanding ECB Working Paper 1779, April 2015, 9 of bank capital’s primary function as the buffer that 

stands between bank survival and failure, they do not speak to the root causes of the losses that capital buffers 

are set to absorb. Bank failure is normally attributed to excessive risk-taking by banks, and some studies have 



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

10 
 

modeled bank risk as an outcome variable of managerial quality and/or corporate governance. Wheelock and 

Wilson (2000) show that during the 1984-1993 banking crisis banks with low managerial quality were more 

likely to fail and less likely to be acquired.  Laeven and Levine (2009) use an international sample of   banks 

to study the effect of conflicts of interest between managers and shareholders on risk-taking during the early 

2000s. They find that banks with more powerful owners tend to take greater risks and also find the impact of 

banking regulations to vary with shareholder concentration. 

Cole and White (2012) is the only study known to the author that identifies the composition of a bank’s loan 

portfolio into real estate products as a relevant predictor of bank failure during the Great Recession. Cole and 

White (2012) decompose the bank’s loan portfolio into various categories of real estate loans and show that 

exposure to commercial real estate loans, loans for construction and development projects, and multifamily 

mortgages increased the probability of failure relative to exposure to non-real estate loans. The authors do not 

find similar effects stemming from holdings of traditional mortgages or MBS, though the absence of an effect 

for MBS holdings is likely a result of their model not explicitly accounting for holdings of MBS.  

Beltratti and Stulz (2012) confirm the findings of Laeven and Levine (2009) for the pre-crisis period, but 

challenge the view that poor bank governance was a major cause of the crisis by showing that banks with more 

shareholder-friendly boards performed significantly worse during the crisis. In addition, Fahlenbrach and Stulz 

(2011) find no evidence that banks with incentive structures which better aligned the interests of shareholders to 

those of the CEO performed better during the crisis.These studies identify agency-type drivers of bank risk, but, 

with the exception of Beltratti and Stulz (2012), abstract away from the specific ways in which risk-taking by 

bank management manifests itself in the choice of business model. Related research has addressed directly the 

question of whether certain business models were more prone to distress than others during the financial crisis of 

2007-2008. Ratnovski and Huang (2009) examine bank performance on a sample of large OECD banks and 

show that higher levels of capital adequacy, deposit funding, and asset liquidity were associated with both a 

lower probability of government intervention and lower stock price declines. Altunbas, Manganelli and 

Marques-Ibanez (2012) use a sample of listed banks operating in the European Union and the United States, and 

explore several measures of realized bank risk to show that credit expansion, lower dependence on 

deposit-funding, larger size, and weaker capital buffers in the run up to the crisis accounted for higher ex-post 

levels of distress.  Fahlenbrach, Prilmeier and Stulz (2012) identify the presence of persistence in bank risk, 

showing that stock return performance during the1998 crisis can predict stock return performance and the 

probability of failure during the Great Recession.The authors also show that reliance on short-term funding, high 

leverage, and high growth rates are all associated with poor bank performance in both crises. Beltratti and Stulz 

(2012) examine the determinants of stock return performance across an international sample of banks during the 

period from July 2007 to December 2008, and find that banks which were more dependent on wholesale funding 

and held less capital performed poorly during the crisis.  

DeYoung and Torna (2013) focus on composition effects, and examine the degree to which the composition of a 

bank’s income sources affected bank distress during the recent financial crisis.They show that for distressed 

banks the probability of bank failure increased with asset-based nontraditional activities (venture capital, 

investment banking and asset securitization), but declined with pure fee-based nontraditional activities 

(securities brokerage and insurance sales). DeYoung and Torna (2013) also show that banks with a substantial 

amount of asset-based nontraditional activities tended to take more risk on their traditional banking activities. 

 

 



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

11 
 

3 Methodology and Data 

3.1 Model 

The model adopted in this study follows the approach of Ratnovski and Huang (2009) but with distinct focus on 

the link between bank asset and recessive factors (exchange rate, inflation and interest rate). With this objective, 

our specification is quite stance from Fahlenbrach, Prilmeier and Stulz (2012). Thus, our baseline VAR (P) 

relation can be represented in log form as. 

10 1 1 1 1 1

1 1 1 1

20 2 2 2 2 2

1 1 1 1

30 3 3

1 1

int inf

int inf

int int

p p p p

t i t i i t i i t i i t i t

i i i i

p p p p

t i t i i t i i t i i t i t

i i i i

p p

t i t i i t

i i

Inba a a Inba b Inex c In d In u

Inex a a Inex b Inba c In d In u

In a a In b Inex

   

   

   

   

 

 

     

     

  

   

   

  3 3 3

1 1

40 4 4 4 4 4

1 1 1 1

inf

inf inf int

p p

i i t i i t i t

i i

p p p p

t i t i i t i i t i i t i t

i i i i

c Inba d In u

In a a In b In c Inex d Inba u

 

 

   

   

  

     

 

   
    3.1                                                                   

Where: ba (bank asset), ex (exchange rate), inf (inflationary rate) and int (interest rate). The VAR (P) above is a 

four-dimension specification and it is plausible to represent it in a compacted form 

0 1 1 ...
t t p t p ty a a y a y u     

                3.2                                                                                                                 

Where y is vector of four dimension log dependent variables (ba, ex, int and inf). Then equation 3.2 with lag or 

backward shift operator can be expressed as. 

0 1 1 ...
t t p t p ty a a y a y u     

                3.3                                                                                                                   

0 1 ...
t

P

t p t ty a a Ly a L y u    
               3.4                                                                                                                    

1 0...
t

P

t p t ty a Ly a L y a u    
                3.5                                                                                                                   

1 0(1 ... )P

t p t ta Ly a L y a u    
                  3.6                                                                                                                

0( ) t tL a y a u 
                    3.7                                                                                                                                                  

 1

1 1
p

p

p

i

L a


 
                     3.8                                                                                                                                                

Equation 3.8 shows that all the roots of the VAR polynomial rest on the complex cycle, meaning that the VAR 

system is not ergordic and all the variables in their log form are I(1) and therefore not stationary. The 

nonstationary VAR (P) needs to be transformed and as usual, we propose a VEC representation as expressed 

below. 

0 1

1

...
t

p

t i t p t

i

y a y a y e  



                     3.9                                                                                                                                                                                                                  



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

12 
 

By assumption: 

( )Rank m k  
, therefore, there is existence of cointegration and  can be decompose into adjustment 

parameter   and long run coefficient
'
. Then, equation 3.9 becomes 

'

0 1

1

...
t

p

t i t p t

i

y a y a y e  



      
            3.10                                                                                                     

Like Johansen (1980), we employed trace statistic to obtain the rank of the cointegration. The statistic is 

specified as follows. 

1

( ) (1 )
p

tr i

i m

lik T P In 
 

    
             3.11                                                                                                                       

Where: 1,...,i p  
, 1 2 ... p    

 are the Eigen values and the Eigen vectors show the 

cointegration rank. 

3.2 Data Source 

The data for this study are secondary and are were sourced from the CBN annual statistical bulletin over the 

period of 1980 to 2015 

4. Results 

4.1 Descriptive Statistics Results 

The descriptive statistics results are based on values of the mean, standard deviation, skewness, kurtosis and 

Jaque-Bera. These values are presented in table 1 below. 

Table 1 Result of Descriptive Statistics on EX, INF, INT and TOT ASS (1980-2015) 

  EX INF INT TOT ASS 

 Mean  11.44389  19.39667  21.24750  6004.744 

 Minimum -21.95000  3.230000  9.500000  18.45630 

 Std. Dev.  13.88579  17.29469  5.895561  9147.837 

 Skewness  0.626376  1.617202 -0.057698  1.325447 

 Kurtosis  4.423916  4.595930  3.232494  3.167121 

 Jarque-Bera  5.395383  19.51254  0.101055  10.58275 

 Probability  0.067361  0.000058  0.950728  0.005035 

Source: Author’s Computation using E-views Window 9 

 

The mean values of exchange rate, inflation interest rate and total asset of banks in table 1 are about 11.44, 

19.40, 21.25 and 6004.74. It appears that all these values are positive. This an indication that these variables do 

not decrease throughout the period of investigation. The standard deviation of these variables are 13.89, 17.29, 

5.90 and 9147.84 respectively for EX, INF, INT and TOT ASS. Looking at these values, it shows that total asset 

has the highest standard deviation implying that it likely going to be the most volatile. Table 1 also shows that 

only interest rate is negatively skewed and as they asymmetric in nature. In the same token, we discover that the 

kurtosis values of the exchange rate and inflation are in excess of three, meaning that they all leptokurtic, while 



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

13 
 

interest rate and total asset are approximally three. Meaning they are mesokurtic. However, the probabilities of 

the Jarque-Bera statistics show that exchange rate interest rate and total asset are normally distributed while 

inflation does not follow a normal distribution. 

4.2 Unit Root Test 

Table 3: The Result of Unit Root Test on EX, INF, INT and TOT ASS 

Variable ADF Stat  5% CV P.V 

EX -4.934  -2.95  0.0003 

D(EX) -6.76  -2.95  0.0000 

INF -2.99   -2.95  0.0458 

D(INF) -5.66  -2.95  0.0000 

INT -2.96  -2.95  0.0458 

D(INT) -6.64   -2.95  0.0000 

TOT ASS 0.97  -2.95  0.9953 

D(TOT ASS) -2.10         -2.95       0.2475 

Source: Author’s computation using E-view 9 

 

From the above table, it is seen that at level exchange rate, inflation, and interest rate are stationary while the 

null holds for total asset that the series of total asset is not stationary or has a unit root. In the same token, at first 

difference exchange rate, inflation, and interest rate are stationary while total asset is not stationary.  

4.3 Test of Cointegration on the Variables EX, INF, INT and TOT ASS  

The cointegration technique introduced by Johansen was adopted in this study. The results however are reported 

in table 4  

Table 4 Cointegration Test Result on EX, INF, INT and TOT ASS  

     
     Hypothesized  Trace 0.05  

No. of CE(s) Eigenvalue Statistic Critical Value Prob.** 

     
     None *  0.597136  69.61549  63.87610  0.0152 

At most 1  0.475563  38.70421  42.91525  0.1239 

At most 2  0.232552  16.75957  25.87211  0.4331 

At most 3  0.204070  7.760295  12.51798  0.2719 

     
     Source: Author’s computation using E-view 9 

Table 4 reports the results of the trace test and maximum Eigen test in panel A and B respectively.  The highest 

Eigen value is about 0.60, meaning that each of the Eigen values is less unity. This result indicates that the 

system is stable and cointegration test can be conducted. The trace statistic under the null hypothesis of no 

cointegration is 69.62 while the corresponding critical value at 5 percent is 63.88. This suggests that the null of 

no cointegration is rejected. Also, the trace statistic under the null hypothesis of one cointegrating equation is 

38.70 associating with critical value 42.92. This means the hypothesis of one cointegrating vector cannot be 

rejected. Thus, the trace test indicates one cointegrating vector and two cointegrating ranks. Therefore, there is 

evidence of cointegration or long run relationship between exchange rate, inflation, interest rate and total asset 

of banks. 



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

14 
 

Table 5 Results of Short Run Relationship and ECM Coefficient 

Regressor                nt 
 

       Coefficient Std. Error t-Statistic Prob.   

ECM(-1) -0.001172 0.006732 -0.174072 0.8625 

DTOT ASS(-1) 0.895007 0.261394 3.423974 0.0012 

DEX(-1) 0.159291 0.331641 0.480311 0.6330 

DINF(-1) -0.462181 0.324225 -1.425497 0.1600 

DINT(-1) 0.439379 0.290755 1.511163 0.1368 

C 10.09596 35.91628 0.281097 0.7798 

Source: Author’s computation using E-view 9 

Note that * and ** imply significance at 1% and 5% respectively, the critical t- statistics @ 1% and 5% are 2.70 

and 1.67 respectively. The degree of freedom is 33. 

The ECM coefficient -0.001 or 0.1 percent as shown in table 5 has the right theoretical sign and significant at 5 

percent. This means that any disequilibrium in the long run can be corrected at the rate of 0.1 percent. Also, 

these results show that there is long run influence or causality running from exchange rate interest rate and 

inflation to total asset. The short run dynamic results reveal that has. Meanwhile exchange rate and inflation 

have significant relationship with total asset. 

5. Conclusion and Recommendation 

Conclusively, recession has a negative significant relationship with bank performance in Nigeria. And on this 

note, it is recommended that monetary authority should formulate policies that will reduce the negative impact 

of recession on banks’ performance since the Nigeria economy is bank driven.  

 

References 

Altunbas,  Yener,  Simone Manganelli,  and David Marques-Ibanez. (2012). “Bank Risk during the 

Financial Crisis:  Do business models matter?”   Bangor Business School,  Prifysgol Bangor 

University (Cymru / Wales) Working Papers 12003 

Aubuchon, Craig P., and David C Wheelock. (2010). “The geographic distribution and characteristics of U.S. 

bank failures, 2007-2010: do bank failures still reflect local economic conditions?” Federal Reserve 

Bank of St. Louis Review, , (Sep): 395–415. 

Beltratti, Andrea, and Ren M. Stulz. (2012). “The credit crisis around the globe: Why did some banks perform 

better?” Journal of Financial Economics, 105(1): 1–17. 

Berger,  Allen  N.,  and  Christa  H.S.  Bouwman. (2013). “How does capital aff ect bank performance 

during financial crises?” Journal of Financial Economics, 109(1): 146–176.. 

Cole, Rebel Allen, and Lawrence White. (2012). “D´ej`a Vu All Over Again:  The Causes of U.S. 

Commercial Bank Failures This Time Around.” Journal of Financial Services Research, 42(1): 5–29. 

DeYoung, Robert, and Gkhan Torna. (2013). “Nontraditional banking activities and bank failures during the 

financial crisis.” Journal of Financial Intermediation, 22(3): 397–421. 

Fahlenbrach, Rdiger, and Ren M. Stulz. (2011). “Bank CEO incentives and the credit crisis.” Journal of 

Financial Economics, 99(1): 11–26. 

Fahlenbrach, Rdiger, Robert Prilmeier, and Ren M. Stulz. (2012). “This Time Is the Same:  Using Bank 

Performance in 1998 to Explain Bank Performance during the Recent Financial Crisis.” Journal of 

Finance, 67(6): 2139–2185. 

Golden, Sam and Harry (1993): The Ten Commandments of Commercial Credit.  The  Cs of good and bad 



Bank Performance versus Recession Indicators: A Linear Cointegration Approach 

                                            

Ashamu Sikiru Oyerinde  

15 
 

loans.  Journal of Commerial Bank Leading 

Laeven, Luc, and Ross Levine. (2009). “Bank governance, regulation and risk taking.” Journal of Financial 

Economics, 93(2): 259–275 

Peek, Joe, and Eric Rosengren. (1997). “How well capitalized are well-capitalized banks?” New England 

Economic Review, , (Sep): 41–50. 

Ratnovski, Lev, and Rocco Huang. (2009). “Why are Canadian Banks More Resilient?”  IMF Working 

Paper, 1–19. 

Wheelock, David C, and Paul W. Wilson. (2000). “Why do Banks Disappear?  The Determinants of U.S. 

Bank Failures and Acquisitions.” The Review of Economics and Statistics, 82(1): 127–138. 

Waweru, N.M. and Kalani, V.M. (2009). Commercial Banking Crises in Kenya:  Causes  and  Remedies.   

African  Journal  of  Accounting,  Economic,  Finance and Banking Research, (N), 12-33. 

McNulty, J., Akhigbe, A. &Verbrugge, J.(2001).  Small Bank loan quality in a  Deregulated  

Environment:  The  information  advantage  hypothesis.Journal  of Economics and Business, 

53, 325-39. 

Samuel, K.S.(2011): Lending practices and causes of loan defaults. Leading practices  and causes of loan 

default among some selected banks in Ghana. 

 

 

 

 

 

Copyrights 

Copyright for this article is retained by the author(s), with first publication rights granted to the journal. 


