FINANCIAL SECTOR REFORM AND SAVING MOBILISATION IN NIGERIA: AN ECONOMETRIC APPROACH
Ilesanmi, Alfred o.*1
Bowale,Kayode,E.*2
Abstract
This study was conducted to determine empirically, the effects of the financial sector reform on savings mobilization in Nigeria during the period under review. Econometric analyses involving unit root test, cointegration test, error correction and Granger causality test were performed on the secondary data sourced from the Central Bank of Nigeria Statistical Bulletin Vol.22, December 2011. The error correction results show that all the determinants of savings like interest rate, inflation rate, monetization variable (MTVt) and Gross Domestic Product, except exchange rate used in the study were properly signed. Both the GDPt and MTVt were statistically significant while all the rates (infalion, interest, exchange) were not. The R2, F-statistic and Durbin Watson statistic were 0.6881, 11.7656 and 1.9038 respectively. The coefficient of the ECM-1 variable was 0.4730 which means Saving was above its equilibrium value and so must be falling down to its equilibrium value in the next period with the speed of adjustment of 0.4730 or 47.3% per period. Therefore, the low saving deposit (interest) rate, high exchange rate and high inflation rate are therefore fingered as the thorn in the flesh as far as savings mobilization is concerned in Nigeria. This conclusion is corroborated by the results of our Granger causality test which indicate independence causality between amount of savings and saving deposit rate, meaning low saving deposit rate did no Granger cause savings. The results also indicate three unidirectional causality between GDPt and SAVt, INRt and EXRt, INRt and GDPt as well as five bilateral or feedback causality between SAVt and EXRt, SAVt and INFt, SAVt and MTVt, INFt and JNRt, MTVt and INRt. It is therefore recommended that the interest rate spread (irs) which is the gap between savings deposit rate and the maximum lending rate should be narrowed down to 5% and all rates (interest, exchange, inflation) should not exceed their one digit threshold levels.
INTRODUCTION
Financial sector reform was one of the most important components of the general structural adjustment programmes (SAP) embarked upon by the then Military administration led by General Ibrahim Babangida in 1986 (Yusuf, 2006). Other components of the structural adjustment programmes include deregulation /liberalization of other services sectors, commercialization and privatization of state owned enterprises (SOEs), elimination of trade barriers with the outside world and so on. It was expected that Nigeria would enjoy economies of scale, enhanced industrial specialization, increased employment of labour, increased income and high standard of living.
The banking sector especially was liberalized to provide finance for small and medium enterprises as agent of development of the economy. The need for economic deregulation arises from the fact that state owned enterprises were characterized by unnecessary bureaucratic bottlenecks which led to inefficiency in operations, production and delivery of output and services.
Economic deregulation is simply the process of removing or withdrawing government control from economic activities involving production, distribution, marketing of goods and services of enterprises or corporations that were hitherto state-owned. In fact , economic deregulation is the process of removal or withdrawal of the exclusiveness of the right to produce, distribute and market goods and services in a particular sector of the economy that is being regulated or liberalized. This definition is equally applicable to the financial sector reform in Nigeria. Financial sector liberalization was therefore carried out in other to circumvent the problem of unnecessary delay , bottleneck and inefficiency in the delivery of banking services in Nigeria. Financial services linearization will allow banking services to be competitive and efficient. To this end, both interest rate and exchange rate were deregulated in the Nigerian financial sector. Interest rate is the price paid by a borrower to a lender of money for parting with his fund for a period of time. Interest rate is the transaction price between the deficit spending unit and the surplus spending unit (Yusuf, 2006).
Before the deregulation of the financial sector, in 1986, saving deposit rate was low while prime lending rate was high, leading to high interest rate spread (irs), a situation that actually discouraged savings and investment in the economy. Financial sector reform, which tended to liberalize interest rate was justified as banks are now free to set the saving rate and the lending rate at level that will reduce interest rate spread to barest minimum so as to close the gap between saving and investment that hitherto served as impediment to financial sector growth and hence economic growth.
The exchange rate on the other hand, is the price of foreign currency in terms of a unit of local currency, which means the unit of Nigerian naira per US dollar or per pound Sterling or per any other foreign currency. Appropriate exchange rate has implication for international traders dealing with import and exports, balance of payments, terms of trade, resource allocation, income distribution, standard of living and general economic performances.
The financial sector reform of 1986 led to liberalization of exchange rate in the Second-Tier- Foreign Exchange Market (SFEM) which later metamorphosed to AFEM and IFEM such that Naira was heavily depreciated from N4.00 per US dollar in 1986/1987 to N129.36 per US dollar at the Dutch Auction System (DAS) and N137.79 per US dollar in Bureau de change (BDC) segment of the market (CBN, 2003). As at the time of compiling this report in January 2012, the depreciation of the Naira has continued to the extent that the exchange rate now stand at N150.00 per US dollar and N250.00 per Pound Sterling.
The objective of this study therefore is to empirically determine the impact of the reforms on saving mobilization in Nigeria . Specifically, to determine the causal relationship between saving and its determinants
LITERATURE REVIEW
So far in this study, we have used the terms deregulation and liberalization simultaneously in the process of financial sector reform as if they are one and the same thing. This is because the element of deregulation and liberalization overlapped (Ozughalu, 2006). But the term liberalization has more to do with financial sector reform than the term deregulation. According to Obadan(2001), liberalization refers to the removal of official restraint to trade, capital flows, investment, production and of course, financial activities by the government. As far as financial sector is concerned liberalization is used to refer to the official dismantling of the existing regulatory structures in the financial sector, and the most important two ways of financial sector liberalization therefore, is to liberalize interest rate and exchange rate. Deregulation on the other hand, is the process of removing official barriers, control and regulation of competition in business and other economic activities to pave ways for increase in private sector participation in production, investment and economic or and commercial activities ( Sloman, 2002). Still on the term deregulation, Oshionebo (2004), states that deregulation refers to substantial pull-out of government from overbearing participation in the nation’s economic activities. Hence forth the term financial liberalization will feature more in this study to refer to financial sector reform in Nigeria.Onwioduokit and Adamu (2006), defined financial liberalization as a set of reforms and policy measure designed to deregulate and transform the financial system and its structures with a view to achieving a liberalized market-oriented system within an appropriate regulatory framework.
On the purposes of financial liberalization, Onwioduokit and Adamu (2006), stated the following among others: introduction of market based procedures for monetary control, promotion of competition in the financial sector, relaxation of restriction on capital flows, greater flexibilities in interest rate, enhancement of the role of markets in credit and foreign exchange allocation, increased autonomy for commercial banks greater depth of money, securities and foreign exchange markets and significant increase in cross border capital flows. On the advantages of financial reform, writers who believe that financial liberalization leads to economic growth include McKinnon(1973), Shaw (1973), Levine and Zervos (1998 b) and Pagano (1993), Bossone (1998). But to Stiglitz (1994, 2000), financial repression leads to economic growth rather than financial reform if properly handled. This argument has been put forward earlier by Diaz-Alejandro (1985) and Demirguc-Kunt and Detriagiache (1998) from their experiences of financial reforms observed in some Latin American countries which have led to more of financial crises, bankruptcies and low saving mobilization.
Bandiera, Caprio and Schiatarelli (2000), in their study on whether financial liberalization raise or reduce saving, listed interest rates, ownership of banks, capital account liberalization, reserve requirements, deregulation of securities markets, direct credit among the components of financial liberalization but concluded that there is no positive effects of real interest rate on saving rather, negative relationship exists in most cases. Loayza, Schmidt-Herbbel and Serven (2000) also argued that financial liberalization have detrimental effects with private saving rates, which means there is a negative relationship between the real interest rate and the private saving rate.
Reinhart and Tokatlidis (2001), in their study of financial liberalization and African experience, found mixed results on the effects of financial liberalization on saving mobilsation. They concluded that financial liberalization enhanced savings in some African regions and declined in some other African regions. This paper is in alliance with the observed position by Akyuz and Kotte (1991) that financial sector reform carried out in many of the emerging economies of the world had led to mixed results of positive and negative benefits. Since Nigeria embraced financial sector reform in 1986, we have witnessed few advantages in terms of increased capitalization, more efficient or improved modern banking services in the area of introduction of the Automated Cash Dispensing Machine (ATM), which has reduced drastically the pressure on the bank workers to discharge their duties and the waiting time on the part of their customers for business transaction. But there are still problems in the area of saving deposit rate, which remain low and the prime lending rate which is on the increase leading to higher interest rate spread (irs) even more than it was before financial reform was embarked upon in 1986.
The wide gap created between saving deposit rates (sdr) and the maximum lending rates (mlr) as a result of the financial reform in Nigeria did not encourage saving mobilization as expected by the proponents of financial liberalization. Adeoye and Adewuyi (2006) clearly stated that in fact the complete liberalization of interest rates that took place in Nigeria in 1996 widen the spread between saving and lending rates instead of creating a competitive market structure and that there is no indication that efficiency has improved in the banking industry from the outset of the reforms
Still on the dis-benefit of financial reform, most bank workers have either lost their jobs or are now experiencing disguised unemployment in some banks. This was the fall-out of the increased capitalization that was raised to N25bn per bank which led to mergers of banks that could not source for such funds within the speculated time period.This recapitalization process has reduced proliferation of banks in the country and phased out distressed and unhealthy banks out of the economy. This exercise has reduced the total number of banks to only 25 healthy ones in 2006. The reform has also changed the nomenclature of community banks to micro-finance banks to serve the interest of the rural communities across the country thereby bringing banking services nearer to the rural people for saving mobilization.
Concerning the relationship between financial sector liberalization and economic growth, King and Levine(1993), Gelb (1989), Gregorio and Guidotti (1992), Ndebbio (2004), observed strong positive relationship between them. Oyaromade (2006), using the techniques of cointegration and error correction mechanism in his study of financial sector reform and saving mobilization in Nigeria found financial sector reform had resulted in higher financial deepening and saving has responded positively to changes in financial variables used in his model. The main policies adopted to achieve the objectives of the reform include liberalization of interest rate, exchange rate and access into banking business , establishment of National Deposit Insurance Corporation (NDIC), strengthening the regulatory and supervisory institutions, upward review of capital adequacy standard, capital market deregulation and introduction of indirect monetary policy instrument, liquidation of some distressed banks and taking over the management of other banks by the Central Bank of Nigeria, and selling of government share holdings in some banks to private sector.
The argument put forward by the proponents of financial sector reform is that financial sector liberalization would bring about effective domestic saving mobilization, promote financial deepening, improved access to credit facilities, reduce interest rate spread and efficient allocation of resources and rapid rate of economic growth.
METHODOLOGY
This study seeks to analyse the effects of the financial sector reform on saving mobilization in Nigeria. Saving mobilization stands for the dependent variable of the model of this study while the explanatory variables include interest rate, exchange rate, inflation rate, monetization variable (MTVt)proxy by M2/GDP and GDP. The Saving mobilization model is therefore, expressed as:
SAVt = α0 + α1INFLt + α2INRt + α3EXRt + α4MTVt + α5lnGDPt + µt ……………………………………….(1)
Where:
SAVt = savings
INTt = interest
EXRt = exchange rate
INFt =inflation rate
MTVt = monetization variable proxy by the ratio of M2t to GDPt (M2t/GDPt)
GDPt = Gross Domestic Product
µt = error term
αi = parameters to be estimated
The coefficients of the regressands of model 1 above can be interpreted as the long run behavior of variable SAVt of the model.
A priori: INFLt < 0, INRt > 0, EXRt < 0, MTVt > 0, GDPt > 0
In logarithm form the above model can be restated as follows:
lnSAVt = β0 + β1INFLt + β2INRt + β3EXRt + β4MTVt + β5lnGDPt +µt...... (2)
Where:
lnSAVt = logarithm of Saving in time t
lnGDPt = logarithm of Gross Domestic Product in time t.
INFLt, INRt, EXRt and MTVt are as defined in model 1 above. There is no need for logarithm because the first three variables are already in rates and MTVt is in ratio or percentage.
βi = parameters to be estimated.
The error correction model is specified as:
∆lnSAVt = λ0 + λ1∆INFLt + λ2∆INRt + λ3∆EXRt + λ4∆MTVt + λ5∆lnGDPt + λ6ECM-1 …………………(3).
Where: ∆ is the first difference operator
λ1….λ6 are the parameters to be estimated
ECM-1 is the residual of the result of equation 2 lagged one period.
To avoid running spurious regression, the unit root test will be carried out using both the augmented Dickey-Fuller (ADF) test and the Phillips-Peron (P-P) test. This step is very germane to this study because time series data obtained from the secondary sources of data collection used in this study are normally fraught with problem unit root. In addition cointegration test will be carried out to establish if long run relationship exist between the variables. It is also important to run the regression of the error correction model in order to tie the short run behavior of the SAVt to its long run value. Finally, the Granger causality analysis will be carried out to establish which of the variable causes the other and the direction of causality.
FINDINGS AND DISCUSSION
Table 1: UNIT ROOT TEST RESULTS
VARIABLES / ADF / P-P / DECISIONlnSAVt / 1.2543 / 1.1151 / I(1)
∆lnSAVt / -4.3173 / -4.1709 / I(0)
INRt / -1.8445 / -1.8738 / I(1)
∆INTt / -7.0101 / -6.9874 / I(0)
EXRt / -6.4998 / -6.5271 / I(0)
INFLt / -3.4589 / -3.3028 / I(1)
∆INFLt / -6.4661 / -11.5403 / I(0)
MTVt / -1.2785 / -1.5594 / I(1)
∆MTVt / -5.3797 / -5.3784 / I(0)
lnGDPt / -0.1642 / -0.1823 / I(1)
∆ lnGDPt / -5.3123 / -5.3061 / I(0)
From the above table, using both the ADF and P-P tests, it is clear that apart from EXRt which is stationary at level all the other variables were integrated of order one, or non-stationary , an indication of unit root problem. But after first differencing they were all integrated of order zero and became stationary. They are now suitable regression analysis.
Table 2: JOHANSEN COINTEGRATION TEST RESULTS: (TRACE TEST)
Hypothesisedno of CE(s) / Eigenvalue / Trace Statistics / 0.05%
critical value / Probability
None * / 0.718011 / 113.7036 / 95.75366 / 0.0017
At most 1 / 0.499531 / 65.59996 / 69.81889 / 0.1036
At most 2 / 0.366105 / 39.29600 / 47.85613 / 0.2487
At most 3 / 0.277639 / 21.97286 / 29.79707 / 0.3000
At most 4 / 0.201900 / 9.614129 / 15.49471 / 0.3116
At most 5 / 0.027108 / 1.044318 / 3.841466 / 0.3068
Trace test indicates 1 cointegrating equation at the 0.05 level
*Denotes rejection ot yhe hypothesis at the 0.05 level
** MacKinnon-Haug-Michelis (1999) p-values
Variables included in the cointegrating vector include: lnSAVt, INRt, EXRt, INFLt, MTVt, and lnGDPt
Table 3: JOHANSEN COINTEGRATION RESULTS: (MAX-EIGEN VALUETEST)
HypothesisedNo of CE(s) / Eigevalue / Ma-Eigen Statistic / 0.05 critica value / Probability**
None* / 0.718011 / 48.10365 / 40.07757 / 0.0051
At most 1 / 0.499531 / 26.30397 / 33.87687 / 0.3025
At most 2 / 0.366105 / 17.32314 / 27.58434 / 0.5523
At most 3 / 0.277639 / 12.35873 / 21.13162 / 0.5127
At most 4 / 0.201900 / 8.569810 / 14.26460 / 0.3237
At most 5 / 0.027108 / 1.044318 / 3.841466 / 0.3068
Max-Eigenvalue test indicates 1 cointegrating equation
*Denotes rejection of the hypothesis at the 0.05 level of significance
**MacKinnon-Haug-Michelis (1991) p-value
From tables 2 and 3 above, which show the cointegration results, both the trace test and the Max-Eigen statistic indicate 1 cointegrating equation. The one star (*) denotes the rejection of the null hypothesis at the 0.05 significance level and the two stars (**) denote the MacKinnon-Haug-Michelis (1999) p-values of the result of the test from no cointegrating equations to at most 5 cointegrating equations.
Table 4: REGRESSION RESULTS OF MODEL 1:(LONG RUN BEHAVIOUR OF SAVt)
REGRESSOR / COEFFICIENT / STANDARD ERROR / t-VALUEIntercept / -3.2843 / 0.1090 / -30.0231
INFLt / -0.0013 / 0.00113 / -1.1344
INRt / 0.0089 / 0.0040 / 2.2494
EXRt / 0.0000133 / 0.0000134 / 0.9931
MTVt / 0.04422 / 0.0022 / 20.0799
GDPt / 0.9848 / 0.0077 / 128.5069
R2 = 0.9985 Adjusted R2 = 0.9983 F = 4693.90 DW = 0.9343