I. Introduction

Studies on the contribution of the financial sector to economic development focus much attention on the competing finance growth nexus. This is traced back to the seminal work of Schumpeter in (1911) who argued that a creation of credit through banking was a crucial factor of entrepreneur capabilities to drive growth. This idea sat well with many governments of Poor countries and sparked a mass of reforms in their financial sector Maduka & Onwuka, 2013). This also led to a spread of a substantial volume of studies (Otto Godly, 2012), but with much less emphasis on gaining and understanding of weather really, financial sector causes poverty reduction and whether finance growth nexus hypothesis can explain poverty reduction. Few questions appear to be resolved or largely undisputed (Odhiambo, 2010). In Nigeria, one of the salient properties of the recent economic growth performance is the reform exercise of the financial sectors. For example, the consolidation exercise in the banking sector plays a major role in the recent increase in the economic growth to the average of 7% since 2006. This apparent increase indicates that Nigerian economic growth is moving faster than the 6.5% Millennium development target goal for poverty reduction in sub-Saharan Africa (MDG, 2010). Surprisingly, despite this increased in economic growth, yet 51.6% of Nigerians were living below US$1 per day in 2004, and sharply increased to 61.2% in 2010. More so, inequality of income distribution rose from 0.429 in 2004 to 0.447 in 2010 (NBS, 2011). While the proportion of the extremely poor increased from 6.2 percent in 1980 to 29.3 percent in 1996 and then came down to 22.0 percent in 2004 before reaching 38.7% in 2010. In fact, the poor continue to be marginalized and deprived access to financial market (Yusuf, Malarvizhi & Khin, 2013). Nigeria is a veritable case for investigating to what extend does the financial sector contribute to economic growth and transforms economic growth into reduction of poverty. In fact, neglecting poverty in the studies of Nnanna (2004), Nzotta & Okerette (2009) constitute a serious measurement error of omission variable. This study contributes by adding poverty reduction in the ongoing debate on the finance-growth relationship.

This main focus of this study is to integrate poverty reduction into the ongoing debate on the direction of the relationship between the financial sector and economic growth in Nigeria by using time series data from 1970 to 2011. The study of Nnanna (2004) and Nzotta & Okerette (2009) applied ordinary least squares to show a correlation, but correlation does not mean causality. This study contributes in various ways, 1) utilizing both conventional and structural break unit root test. (2) Employing the ARDL bound testing approach to co integration for a long run relationship between the variables in the presence of structural breaks (3) Utilizing Vector autoregressive (VAR) model with augmented lag order for the estimation of the Toda & Yamomoto no causality. (4) Studying both short run and long run inputs. (5) Using VECM Granger causality approach for a causal relationship. (6) Using comprehensive measures of financial deepening.

1.1 The Financial Sector Development in Nigeria

The linkage between financial sector development, economic growth and poverty reduction can theoretically be established based on the two strands school of thought who emerged to explain the competing finance-growth and poverty reduction relationship. These are the financial regulation based versus financial deregulation based theoretical paradigm. According to the financial repression based theory, banking sector provide capital for economic growth and access to information to take up (Levine, 2002). On the other hand, the letter approach, pointed out that the market provides the efficient mediating role between financial sector development in one and economic growth in the other. Indeed, the unregulated financial sector was seen as the locomotive of economic development. This theory went further to explain that a sound and liquid financial market serves as engine economic growth and enable growth to transmit into poverty reduction (Levine, 2002). On the Nigeria scene the sector has passed through a various transformation from Independence to date. The era of 1950 to 1970 was the establishment era, when major commercial banks, including Central bank were established in the country. The structural Adjustment period, which was established in 1986, set the beginning of real financial sector transformational agenda. Before this time financial sector was highly repressed characterized by interest rate ceilings, selective credit expansion, control reserve requirement and various forms of direct monetary policy controls. Government owned, monopolized the banking industry with strict entrance regulation to the private sectors and regulated foreign exchange regime. However, with the implementation of Structural Adjustment Program there was a policy reversal toward the deregulation of the economy, by establishing the Nigerian deposit insurance corporation, putting in place corporate governance policies through tight supervisory institutions. Capital market was deregulated, and the indirect monetary policy instrument was introduced. The central bank took over the control of the distress banks in the country. Exchange rate reforms were put in place and this was achieved by establishing market based autonomous foreign exchange market. Reintroduction of exchange control in the 1994 but immediately after one year of its maturity, it was replaced back to the deregulated exchange regime. Similar goes to the interest rate regime which was capped in for a year and then reverses to the deregulated interest rate regime. Another major change was the deregulation of the banking sector; this was followed by the liberalization of the interest rate regime and the free entry of the banking business by the private individuals. This also led to the raised in the nominal interest rate in the country.

However, it is worth nothing that despite all these lofty efforts of the transformation process of reforming the financial sector in Nigeria. The sector remains disconnected with the real sector of the economy. However, in order to address the major problems of poor capitalization of banks in Nigeria the banking consolidation program was established in the country in the mid-2004. Under the banking consolidation program banks are mandated by the Central Bank of Nigeria, to increase their capital based from the N2 billion amount to N250 billion. This was the aim of enhancing the availability of the loan able funds in the economy. Banking sector consolidation was to be supported with improvement in cooperating governance in terms of banking supervisory functions. The consolidation created a shift in the banking sector performance from efficient supervision to a risk based system strategy, based on the Based-11 Accord.

Referable to the overheating crisis on the financial sector emanating from the global melt down. Certain important reform measures were introduced, which includes, improving the quality of banks in role and contribution towards a sound financial sector and transforming the sector to play role toward the real sector performance. Despite all these measures put in placed the financial sector reforms and development do not translate into meaningful economic growth and poverty reduction (Yusuf et al., 2013). Instead the poverty rate in the nation continues to soar and the percent of the entire population in poverty continues to rise (NBS 2010). The relative poverty also continues to increase moving from 54.4 percent in 2004 to 69 percent in 2010 (NBS, 2010).

2. Literature Review

Theoretically, there are two strands school of thought which emerged to explain the competing finance-growth nexus. These are the banks based versus market based theoretical paradigm and the second was based on the law and finance based. According to the bank based theory, banking sector provide capital for economic growth and access to information to take up (Levine, 2002). On the other hand, the market friendly approach, pointed out that the market provides the efficient mediating role to the economic growth, in fact, the market was examined as the locomotive of economic development. This theory gave way further to demonstrate that a sound and liquid market provide higher economic growth and enable risk management and general diversification (Levine, 2002). So the finance friendly paradigm which was brought forth from the bank based approach stressed the importance of financial service provided by the financial organization. In the end, the last paradigm emphasized on the law based approach which stressed the important role of the sound system in creating growth (Upchurch, 2012).

On the empirical ground, studies in these perspectives, is divided into two dichotomies,

Some take the view that finances follow where enterprises lead. This was set up in the work of McKinnon (1973) and Shaw (1973) in their surveys; they find out that the development of intermediaries causes economic growth. In addition, they study the negative effect of government participation in the activities of the financial sector and economic growth. They indicated that government barriers in the banking system decrease the process of financial growth and economic development. The argument that finance follows where enterprise leads are often followed in the work of Robin & Salina Martin (1992), these studies applied cross-sectional data, their studies provide pooled effects outcomes without providing details on country specific issues. Likewise, it neglected to get the dynamic effects between the two variables under investigation.

One of the outstanding studies on financial sector development and economic growth emerged from the works of Mehran (1998) who examine the progress recorded on the transformation of the financial sector of sub-saharan African countries. The findings suggest that financial sector development in Sub-Saharan African requires reforms in the area of financial supervision, financial development of the monetary operations and financial market. Marry, Chibuzo & Okelue (2012) investigate the effects as well as the relationship between financial sector development and economic development in Nigeria. The findings suggest a significant positive linkage between government consumption and trade openness. However, on the contrary the measures of financial development suggest a negative linkage with economic growth. In a related development Saibu, Nwosa & Agbeluyi (2011) investigate the effects of financial development and foreign direct investment on economic growth in Nigeria. The results indicate that there is a bidirectional causality relationship between some of the proxies of financial development and economic growth. Odeniran & Udeaja (2010) examine the relationship between financial sector development and economic growth in Nigeria, using Granger causality tests. The findings suggest bidirectional causality between financial sector development and economic growth in Nigeria.

However, despite a wide number of studies, few have focused on identifying the extent to which poverty reduction can be included into the competing debate on the financial sector, economic growth relationships. The hypothetical argument why poverty need to be included in the growth, financial sector hypothesis is based on the argument which can be demonstrated through a transmission mechanism which shows that financial sector affects poverty reduction in two channels of influence. The first is by encouraging the less privilege people in the society to have access to funds for productive investment. Financial sector development is assumed will enables the poor access to micro credit, which leads to income generation and employment opportunity for the poor (Department for International Development (FDID), 2004). The second channels of influence are through economic growth, whereby economic growth is expected to trickle down to the poor and translate it into sustainable development (World Bank, 1995; Mellor, 1999; Ravallion & Datt, 2002; Dollar & Kraay, 2002). Though, the automatic role of economic growth in poverty reduction remains contested due to the inequality of income which affects income distribution in the economy and worsens distribution.

Adam (2011) examines the impact of Ghana financial openness induced growth on poverty between 1970 and 2007. The finding suggests a positive relationship between growth and financial liberalization. On whether financial liberalization benefits the poor the findings suggest a positive relationship, but in a disproportionate manner. Therefore, on the basis of the findings they argued that credit channels are more efficient ways of addressing poverty if supported by a good policy intervention. Among the few studies supporting this argument is the study of Pradham (2010) who draw data from India and investigates the direction of the relationship between the financial sector, economic growth and poverty reduction for the period 1951-2008, through the technique of co -integration and causality test. The study finds a long run relationship running from finance to growth and poverty reduction. Odhihiambo (2009) using data from South Africa, established that economic growth granger causes financial development and, therefore, lead in the process of poverty reduction in South- Africa. Even though, the study suffers from measurement error by using granger frame work in the presence of small sample size, which consumes a lot degree of freedom, yet the study has been one of the few breakthrough studies in Africa. Ndebbio (2004) used ordinary least square regression techniques and the findings of the study provide an interesting revelation of a very weak relationship from finance to growth. This, according to him was attributed to poor and weak financial sector development in Nigeria. Alayande (2003) analyzed the patterns of inequality in Nigeria, using the regression-based approach to decompose inequality of income from sources using the Gini index. The findings of the studies suggest that geography contribute significantly in explaining poverty. The results of Alayande (2003) also show that the sector of residence alone accounted for the largest source of inequality in Nigeria. Therefore, went further to reveal that the level of education contributes significantly in explaining inequality of income. Alayande (2003) discovered that certain variables provide a little contribution such as the age and the sex of the respondents. Therefore, suggests that any policy on poverty reduction in Nigeria should give more emphasis to geographic distribution of the population. Obayelu & Awoyemi, (2012) applied regression decomposition techniques using National living standard survey data. The findings of the study attributes that household size contributes significantly to poverty.

3. Data and Methodology

3.1. Sources of Data

The data for poverty were source from the Nigeria National bureau of statistics Abuja Nigeria for various years based on the national living standard surveys. Data for the ratio of banks deposits liabilities to GDP were source from the World Bank global financial development data set for various issues. The data on the credit to private sector to nominal GDP were source from the World Bank global financial development data set for various years. The data on Real interest rate were source from the World development indicators. The data on Trade ration were obtains from the World Bank Development Indicators.

Two approaches were used to ascertain the direction of influence between finance, growth and poverty nexus. The first approach involves the newly introduced ARDL approach to co- integration and causality. The second approach applied the Vector autoregressive (VAR) model with augmented lag order for the estimation of no causality Toda and Yamamoto test. The hypothesis of interest is that poverty reduction can be predicted by finance and growth interaction with the aid of certain control variables which includes interest rate, international trade and Real interest rate. However, based on the theoretical postulates a positive and significant coefficient estimate is expected. Therefore, following the work of Odhiambo (2009) and Pradhan (2010) the model is specified as follows:

GDPt = [alpha]0 + [alpha]1FD + [alpha]3RIt + [alpha]4TRt + [alpha]5POt + Kt -----(1)

Where GDP is the proxy of economic growth, FD is the financial depth; RI is the real interest rate which is the deposits rate minus the inflation rate. TR= is the total value of export and import as share of nominal GDP. PO is proxy of poverty reduction. [alpha] 0 to [alpha] 5 is the slopes coefficient and Kt is the error terms

Following the work of (e.g. King & Levin, 1993; Odhiambo, 2009; Pradham, 2010). This study takes the log of GDP per capita as a proxy of economic growth. However, the work of Gelb (1989), World Bank (1989) and King & Levine (1993) applied broad money (M2) ratio to nominal GDP for financial depth, improvement in this ration will mean improvement in financial depth. Thus in Nigeria this ratio involves huge amount of currency increased, as such it may not be appropriate because it is measuring monetization instead of financial sector development. To address this mistake, this study identified two important variables to measure financial development which includes the ratio of bank deposits liabilities to GDP proxy as LQL. This ratio measure the extent to which the banking sector discharged it primary functions of allocating funds from mobilized saving to productive investment. This study equally considered the credit to private sector to nominal GDP, proxy as CPS in order to measure the capacity of the financial sector ability to sourced productive investment and ensure that risk is properly managed. The choice of this variable is due to the fact that there is the possibilities that credit to private sector my likely continue to be fixed even in situation where the deposits are rising. However, this study equally, considers certain control variables to growth and financial development which includes the real interest rate RI, which constitutes the deposits rate minus inflation rate and the trade ratio TR which measure the total volume of export and import as share of nominal GDP. However in order to integrate poverty reduction into the equation the PO which is proxy as poverty reduction is also considered. However data are obtained from the World development Indicators and Global Financial Data base for all variables.

Thus, in order to avoid the common mistake where series are often specified at level and non-stationary, there was a suggestion that the set of series need to be stationary or differenced. But the problem with differencing of a series it takes away long run information from the time series. To address these obvious problems quite a number of techniques have been employed which includes Eangle (1987) test, Maximum likelihood of Johensen (1988) and Jehensen & Juselius (1990) tests. Thus, it should be noted that despite, their common application they are not far free from certain weakness, in fact, these techniques are found not reliable in a situation where the sample size is very small.

This study employed the newly introduced techniques by Pesaran & Pesaran (2001) the autoregressive distributed lag model (ARDL) because it can be applied irrespective whether the variables are 1(0), 1(1) or fractionally co- integrated. Similarly, it easy for the error correction model (ECM) to be established from the ARDL. The model is specified as follow:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

Where [beta]0 is the constant and Kt is the disturbance term. However, the first part of the equation with the summation sign stands for the error correction. While the other part of the equation with l 1 to t l 5 represent long run relationship.

GDP is the proxy of economic growth, CPS represent credit to the private sector to nominal GDP, while LQL represent the ratio of bank deposit liabilities to GDP; RI is the real interest rate which is the deposit rate minus the inflation rate. TR= is the total value of export and import as share of nominal GDP. PO is proxy of poverty reduction. The next step in ARDL approach is the estimation of the Bound test through the (OLS) approach and the F- test which are used to examine whether the series have long run relationship. The null hypothesis of the equation (2) is HO: l 1=0, l 2=0, l 3=0, l 4=0, l 5=0, which implies lack of the presence of long run co-integration and the alternative hypothesis is HI: l1 [not equal to] 0, l2 [not equal to] 0, l3 [not equal to] 0, l4 [not equal to] 0, l5 [not equal to] 0

The results of the computed F- test is normally, used to determine whether there is long run co-integration by comparing the two sets of critical values extracted from the Pesaran critical tables which are based on 1(0) and 1(1). According to the decision rules if the estimated value of F- test is higher in values when compared with the upper critical values it implies co -integration. But if the F-test values fall below the lower critical values the null hypothesis of no co -integration cannot be rejected. But if it falls in between the upper and lower critical values the results is inconclusive.

Thus ARDL model is sensitive to the lag length which is selected on the basis of Scharwrtz Bayesian criteria (SBC) and Akaike information criteria (AIC). The SBC is known as parsimonious model, as based on the smallest possible lag length.

The last stage in ARDL model is the establishment of the error correction model through a simple linear transformation. This can be presented below:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Where GDP is the proxy of economic growth, CPS represent credit to private sector to nominal GDP, while LQL represent the ratio of bank deposits liabilities to GDP; RI is the real interest rate which is the deposits rate minus the inflation rate. TR= is the total value of export and import as share of nominal GDP. PO is proxy of poverty reduction. The error correction show the level of adjustment to the long run steady equilibrium after experiencing shocks in the short run.

The stability of the model is tested through the cumulative (CUSUM) and cumulative sum of squares (CUSUMSQ) if the plots of the CUSUM and CUSUMSQ statistics remain within critical bonds of 5 percent level of sign of significance the null hypothesis of stability is achieved.

3.2. Toda and Yamamoto Approach

The rationale behind this approach is based on the common understanding that in testing for causality, the usual F- test for linear restriction become in valid if the lags of the dependent variables enter the model as repressors. The point here is that, if the data are non-stationary, the Wald tests statistics does not follow its usual asymptotic chi-square distribution under null hypothesis (Toda & Yamamoto, 1995). Therefore, the way to address this problem is to estimate VAR with its true order K, but with lag order of (K+d), where d is treated as the maximum potential order of integration of the variables. But it should be noted that the estimation of the granger causality is conducted without the additional K+1-K+d.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)

Where K is the optimal lag order d is the maximum lag order of integration of the variables in the system and Kt1 and Kt2 are the error term. The remaining series are as defined earlier in equation (2).

4. Estimation of the Results

In this section the study starts with the presentation of the results and followed by the analysis of the results. The first table provides the descriptive statistics of the sample data.

The results in table 2 provide the descriptive of the sample data, which shows that the skewness is not near zero on CPS and GDP. And the kurtosis is not near three in the case of Po, LQL, RI and TR. This implies that the data is not normally distributed. Similarly, the probability of 5% is not achieved in all the series, with these we move to the Unit root test in order to ensure the series are integrates before any economic estimation can be conducted.

The estimated results of the Unit root test in table 3 above based on the ADF test, indicates that the null hypothesis of no unit root cannot be rejected at 5% level of significant with the exception of RI, but after taking first differencing all variable are found stationary at 5% level of significant. Thus the confirmatory Philips Perron test also suggests that all series become stationary after taking the first differencing at 5% level of significant.

Bound Test for the long Run Co-integration

The critical values are extracted from Narayan table which comprised both the upper critical values and the lower critical values represented as 1(0) and 1(1) respectively.

The results of a bound tests for long run relationship is reported in table 4 and 5 which includes critical values and computed F- test values of joint significant. The result shows the critical values of the bound testing of rejecting or otherwise of the null hypothesis. As such the results in table 5 above of the bound test for the long run relationship indicate that the null hypothesis of no co -integration can be rejected since the estimate of the coefficient of the lag GDP is having a value greater than the upper critical values at 5% level of significant. This implies that there is a long run association between financial development, economic growth and poverty reduction. However, having long run relationship does not mean causality therefore, to determine whether financial sector development causes economic growth and poverty reduction Unrestricted error correction models were estimated together with Toda and Yamamoto no causality tests. But before estimating the causality relationship the long run coefficient of poverty as a dependent variable is estimated which is presented in table 6 below.

Table 6 above presents the results of the Long run relationship between poverty and Economic growth, financial development, and the instrumental variables of Interest rate and trade percentage of GDP. However, the result indicates that all the explanatory variables are not significantly impacting poverty.

Diagnostic Tests

Test Statistics * LM Version * F Version

A: Serial Correlation CHSQ (1) = 18.5265[.056] F (1, 34) = 26.8344[.056]

B: Functional Form CHSQ (1) = 0.057956[.810] F (1, 34) = 0.046982[.830]

C: Normality CHSQ (2) = 1.0613[.588] Not applicable

D: Heteroscadasticity CHSQ (1) = 8.0812[.054] F (1, 40) = 9.5300[.054]

A: Lagrange multiplier test of residual serial correlation

B: Ramsey's RESET test using the square of the fitted values

C: Based on a test of skewness and kurtosis of residuals

D: Based on the regression of squared residuals on squared fitted values

The results of the diagnostic tests indicate that the null hypothesis that there is stability of the coefficient estimated cannot be rejection, since the serial correlation results, functional form, Normality and Heteroscadasticity are having the probability values above 0.05% respectively.

To ascertain whether financial development causes economic growth and poverty reduction a vector error correction model is estimated which were presented in table 7 above. However, the results from table 7 above in the poverty model with DPOt-1 as a dependent variable suggest that the coefficient of DGDPt-1, DCPSt-1 and DLQLt are not statistically significant implying that economic growth, credit to private sector and the ratio of bank deposits liabilities to GDP does not impact on poverty reduction. Thus in the credit to private sector model with DCPSt-1 as a dependent variable the results suggest that the coefficient of DGDPt-1 is statistically significant with a positive sign this implies that economic growth impacts on credit to private sector. However, in the ration of bank deposits liabilities to GDP model with DLQLt-1 as a dependent variable the results suggest that the coefficient of the DGDPt-1 is statistically significant and positive indicating that economic growth impact on the ratio of bank deposit liability to GDP. This also suggests that economic growth lead and financial sector follow in Nigeria. However, the coefficients of DTRt-1 and DRIt-1 in the credit to private sector are statistically significant implies that trade percentage of GDP and the real interest rate impacts on credit to private sector development. Similarly, the coefficient of the DRIt-1 is statistical significant in the Economic growth model indicating that real interest rate impact on economic growth in Nigeria. The policy implication is that the increase in the supply of the loan able funds through the financial sector development is not enough to trickles down financial resources to the poor other intervening policies accompanied by good governance and equitable income distribution are important.

The diagram above in figure 1 above represents the stability test. Thus the straight lines represents the critical bounds at 5% level of significant the vertical line represents the values of the critical values. However, the figure test the stability of the parameters in the model and based on the results in the figure in table 1 the stability of the model is achieved, since the parallel line do not crossed and is within the 5% level of significant.

Thus, the findings of table 8 to table 13 represents the Toda & Yamamoto no causality tests results. In facts, the result table 8 of the poverty equation with dependent variable PO, the estimated coefficients of GDP, CPS, LQL, TR, RI are not significant indicating that the null hypothesis of no causality cannot be rejected. This indicates that financial sector development and economic growth have not significantly impact on poverty reduction. The findings of this study do not support the findings of the work of Odhihiambo (2009).

The results of table 9 of the economic growth equation with dependent variable GDP, the estimated coefficients of PO, CPS, LQL, TR, and are not significant. This suggests that poverty reduction, the ratio of bank deposit liabilities to GDP, trade percentage of GDP, do not significantly impact on economic growth. But the coefficient of the RI is statistically significant implies that real interest impact on economic growth.

In the credit to private sector equation with dependent variable CPS, in table 10 the estimated coefficient of GDP is significantly positive. This indicates that economic growth is significant with positive value. This implies that economic growth impact on credit to private sector. The coefficients of TR and RI is statistically significant implies that trade percentage of GDP and the real interest rate impact on credit to private sector.

In the ratio of bank deposits liabilities to GDP equation in table 11, with dependent variable LQL, the estimated coefficient of GDP is significantly positive. This suggests that economic growth has a significantly positively impact on the ratio of bank deposits liabilities to GDP.

In the trade percentage of GDP equation as provided in table 12, with dependent variable TR, the estimated coefficient of economic growth is significantly positive. This suggests that economic growth significantly positive impact on trade as a percentage of GDP. While in the same estimated equation the coefficients of PO, CPS, LQL, RI, are not significant. Indicating that poverty, credit to private sector, the ratio of banks deposits liabilities to GDP, real interest have not significantly impact on trade percentage of GDP.

In the real interest rate equation with dependent variable RI as presented in table 13 the estimated coefficient of TR is significantly positive. This indicates that trade as a percentage of GDP in Nigeria has a significantly positive impact on real interest rate. The general testing of Toda Yamamoto results support the findings of the studies of Odhianibo (2009), Coccorese (2008), and Odhiambo (2011) which suggest growth led financial sector follow. But do not support the study of Odhihiambo (2009) that financial development and economic growth causes poverty reduction.

5. Conclusion

This study introduces poverty reduction in the competing financial sector economic growth hypothesis in order to investigate whether financial sector development and economic growth contribute towards the eradication of poverty in Nigeria. This study contributes by employing two method of analysis, which includes, the Autoregressive distributed lag model and Toda and Yamamoto approach to no causality. The results suggest that financial sector development and economic growth do not causes poverty reduction in Nigeria. This study provide important understanding that that despite increased in the availability of loan able funds through the financial sector development the poor segment of the Nigerian population does not have access to financial resources. In the case of Nigeria this indicates that financial development alone without equitable income distribution and good governance may not be enough to address the menace of poverty. The finding of this study does not support the work of Odhianibo (2009), who purported to show that financial sector development is positively correlated with poverty reduction. The policy implication here is that financial sector development requires a backup of supporting policies of resources distribution and good governance before, it can pools the mobilized capital into productive investment and equally, channels economic growth to the poor.

Muhammad Yusuf DANDUME

Umaru Musa Yaradua University, Economic Department, Katsina State of Nigeria, dandumey@yahoo.com

References

Adams, A.M. (2012). Financial Openness Induced Growth and Poverty Reduction. the International Journal of Applied Economics and Finance, 5(1), 75-86.

Adeyemo, A., & Alayande, T. (2003). The Impact of Government Poverty Alleviation on Entrepreneurship Development in Nigeria Readings in Development Policy and Capacity Building in Nigeria: A Book of Readings in Commemoration of the Tenth Anniversary of the Development Policy Centre.

Coccorese, P. (2008). An Investigation on the Causal Relationships Between Banking Concentration and Economic Growth. International Review of Financial Analysis, 17(3), 557-570.

DFID, (2004). Financial Sector Development: A Pre-requisite for Growth and Poverty Reduction? Policy Division, Department for International Development, London (June).

Dollar, D. & Kraay, A. (2002). Growth is Good for the Poor. J. Econ. Growth, 7, 195-225.

Eggoh, J.C., Bangake, C. & Rault, C. (2011). Energy Consumption and Economic Growth Revisited in African Countries. Energy Policy, 39(11), 7408-7421.

Engle, R.F., & Granger, C.W. (1987). Co-Integration and Error Correction: Representation, Estimation, and Testing. Econometrica: journal of the Econometric Society, 251-276, Retrieve from https://students.pomona.edu/2009/zs022009/Desktop/engle1987.pdf

Gelb, A.H. (1989). Financial Policies, Growth, and Efficiency, Policy Planning, and Research. Working Papers, No. 202 (World Bank). Retrieve from http://go.worldbank.org/ARQE6SFCT0

Hasan, M. (2011) Democracy and political Islam in Bangladesh, South Asia Research, 31(2) 97-117.

Jedidi, F.K., & Mensi, S. (2011). The Trade-off between Liberalization Policy and Financial Crises Dynamics. Asian Social Science, 7(3), 1911-2017.

Johansen, S. (1988). Statistical Analysis of Cointegration Vectors. Journal of Economic Dynamics and Control, 12(2), 231-254.

Johansen, S. & Juselius, K. (1990). Maximum Likelihood Estimation and Inference on Co-Integration-With Applications to the Demand for Money. Oxford Bulletin of Economics and statistics, 52(2), 169-210.

Kar, M., Nazlioglu, S. & Agir, H. (2011). Financial Development and Economic Growth Nexus in the MENA Countries: Bootstrap Panel Granger Causality Analysis. Economic Modelling, 28(1), 685-693.

King, R.G. & Levine, R. (1993). Finance, Entrepreneurship and Growth. Journal of monetary economics, 32(3), 513-542.

King, R.G. & Levine, R. (1993a). Finance and growth: Schumpeter might be right The quarterly journal of economics, 108(3), 717-737.

Levine, R. (2005). Finance and Growth: Theory and Evidence. Handbook of Economic Growth, 1, 865-934.

Lucas Jr, R.E. (1988). On the Mechanics of Economic Development. Journal of Monetary Economics, 22(1), 3-42.

Maduka, A.C. & Onwuka, K.O. (2013). Financial Market Structure and Economic Growth: Evidence from Nigeria Data. Asian Economic and Financial Review, 3(1), 75-98.

McKinnon, R.I. (1973). Money and Capital in Economic Development, Brookings Institution Press.

Mehran, H. (1998). Financial Sector Development in Sub-Saharan African Countries, International Monetary Fund.

Mellor, J.W. (1999). Pro-Poor Growth: The Relationship between Growth in Africa and Poverty Reduction. Abt Associates, Bethesda, MD.

Mery, O.I., Chibuzor, E.E. & Okule, U.D. (2012). Does Financial Sector Development Causes Economic Growth Empirical Evidence from Nigeria. International Journal of Current Research, 4(11), 343-49.

National Bureau of Statistics. (2010). Statistical News: Labor Force Statistics, No.476. Abuja: The NBS Publication.

Ndebbio, J.E.U. (2004). Financial Deepening, Economic Growth, and Development: Evidence from Selected Sub-Saharan African Countries Retrieved from http://mpra.ub.uni-muenchen.de/id/eprint/29330

Nnanna, O. (2004). Financial Sector Development and Economic Growth in Nigeria: An Empirical Investigation. CBN Economic and Financial Review, 42(3).

Nzotta, S. & Okereke, E. (2009). Financial Deepening and Economic Development of Nigeria: An Empirical Investigation. African Journal of Accounting, Economics, Finance and Banking Research, Retrieve from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1534212

Obayelu, O.A. & Awoyemi, T.T. (2012). Spatial Decomposition of Income Inequality in Rural Nigeria. Journal of Rural Development, 31(3), 271-286.

Odeniran, S.O. & Udeaja, E.A. (2010). Financial Sector Development and Economic Growth Empirical Evidence from Nigeria. Central Bank of Nigeria Economic and Financial Review, 48(3), 91-124.

Odhiambo, N.M. (2009). Finance-Growth-Poverty Nexus in South Africa: a Dynamic Causality Linkage. Journal of Socio-Economics, 38(2), 320-325.

Odhiambo, N.M. (2010). Finance-Investment-Growth Nexus in South Africa: An ARDL-Bounds Testing Procedure. Economic Change and Restructuring, 43(3), 205-219.

Odhiambo, N.M. (2011). Financial Deepening, Capital Inflows and Economic Growth Nexus in Tanzania: A Multivariate Model. Journal of Social Science, 28, 65-71.

Otto, G., Ekine, N.T. & Ukpere, W.I. (2012). Financial Sector Performance and Economic Growth in Nigeria. African Journal of Business Management 6(6), 2202-2210.

Padhan, P.C. (2007). The Nexus between Stock Market and Economic Activity: An Empirical Analysis for India. International Journal of Social Economics, 34(10), 741-753.

Pesaran, M.H., Shin, Y. & Smith, R.J. (2001). Bounds Testing Approaches to the Analysis of Level Relationships, Journal of applied econometrics, 16(3), 289-326.

Pradhan, R.P.P. (2010). The Nexus between Finance, Growth and Poverty in India: The Cointegration and Causality Approach. Asian Social Science, 6(9), 114.

Ravallion, M. & Datt, G. (2002). Why Has Economic Growth Been More ProPoor in Some States of India than Others?, J. Dev. Econ. 68, 381-400.

Roubini, N. & Sala-i-Martin, X. (1992). Financial Repression and Economic Growth, Journal of Development economics, 39(1), 5-30.

Saibu, I.M.O., Nwosa, P. & Agbeluyi, A.M. (2011). Financial Development, Foreign Direct Investment and Economic Growth in Nigeria. Journal of Emerging Trends in Economics and Management Sciences, 2(2), 146-154.

Schumpeter, J.A. (1911). The Theory of Economic Development. Reprinted 1969, Oxford: Oxford University Press.

Shan, J. & Morris, A. (2002). Does Financial Development 'Lead' Economic Growth? International Review of Applied Economics, 16(2), 153-168.

Shaw, E.S. (1973). Financial Deepening in Economic Development, Oxford: Oxford University Press New York.

Toda, H.Y. & Yamamoto, T. (1995). Statistical Inference in Vector Auto Regressions with Possibly Integrated Processes. Journal of Econometrics, 66(1), 225-250.

Upchurch, M. (2012). Persistent Economic Divergence and Institutional Dysfunction in Post-Communist Economies: An Alternative Synthesis. Competition & Change, 16(2), 112-129.

World Bank. (1989). World Development Report 1989. New York: Oxford University Press.

World Bank. (1995). Bangladesh: From Counting the Poor to Making the Poor Count. Poverty Reduction and Economic Management Network, Washington: South Asia Division. WorldBank.

Yusuf, M., Malarvizhi, C. & Khin, A.A. (2013), Trade Liberalization Economic Growth and Poverty Reduction in Nigeria, International Journal of Business and Management, 8(12), 42.

* The English title and abstract of this article has been translated into Turkish by the Editorial Board. Bu calismanin Ingilizce baslik ve ozeti, Yayin Kurulu tarafindan Turkce'ye cevrilmistir.

Table 1: Variables and Their Source of Data Symbol Variables Source GDP Economic growth World Bank Development Indicators FD Financial dept World Bank Development Indicators RI Real interest rate World Development Indicators TR Total value of export World Development Indicators and import PO Poverty National bureau of statistics of Nigeria Table 2: Descriptive Statistics of the Sample Data CPS GDP LPO Mean 13.62293 7316044 43.84286 Median 12.70096 5282712 46.3 Maximum 38.59012 51310010 69 Minimum 4.699551 1064975 16.2 Std. Dev. 7.063296 10094527 16.90854 Skewness 1.67857 3.876001 -0.17787 Kurtosis 6.307768 17.10966 1.854462 Jarque-Bera 38.87051 453.5578 2.517911 Probability 0 0 0.28395 Sum 572.1629 3.07E+08 1841.4 Sum Sq. 2045.496 4.18E+15 11721.84 Dev. Observations 42 42 42 LQL RI TR Mean 22.99911 -2.34184 58.36589 Median 22.32234 -3.43364 65.01609 Maximum 37.6967 25.13001 97.32115 Minimum 10.15575 -32.0573 19.6206 Std. Dev. 7.975687 13.37939 21.12774 Skewness 0.21803 -0.1776 -0.19541 Kurtosis 1.970667 2.814356 1.805794 Jarque-Bera 2.186928 0.281104 2.76301 Probability 0.335054 0.868878 0.2512 Sum 965.9625 -98.3571 2451.368 Sum Sq. 2608.075 7339.331 18301.63 Dev. Observations 42 42 42 Table 3: Augmented Dickey- Fuller (ADF) Test and Phillips Perron Test for a Unit Root Tests ADF test First Series At level Difference CPS 1.91972(0) -5.929831(1) * GDP 1.789547(4) -2.578563(4) * PO 1.678079(0) -6.593416(1) * LQL -1.886665(0) -5.05153(1) * TR -2.42861(0) -6.234625(3) * RI -5.930808(0) * -7.280181(4) * Phillip First Series Perron At level Difference CPS -1.920942(4) -5.951634(5) * GDP 0.813877(5) -5.989714(4) * PO -1.672906(5) -6.621889(3) * LQL -1.88665(0) -5.058509(3) * TR -2.420558(3) -6.234754(4) * RI -5.936787(1) -25.53783(5) * Note *** indicates the significance level at the 5%, 1% and 10% respectively. Numbers in bracket are the lag length Table 4: Critical Values of the Bound Test with Intercept and no Trend 10% level 5% level 1% level 1(0) 1(1) 1(0) 1(1) 1(0) 1(1) 3.247 3.773 3.993 4.533 5.754 6.483 Note *** indicates statistically significant at the 10% 5% and 1% level critical value Table 5: Estimated F- Test statistics of Joint Significant GDP= 0.46865 Table 6: Long Run Coefficient Poverty as a Dependent Variable Regression Coefficient Standard Error Prob. GDP -0.26758 15738 0.099 CPS 0.69833 1.4796 0.64 LQL -0.85808 0.9752 0.386 RI -0.58742 0.64276 0.368 TR -0.58187 0.46282 0.218 Notes * ** *** indicates the significance level at the 1%, 5% and 10% respectively. ARDL(1,0,0,0,0,0,1) selected based on Schwarz Bayesian Criterion Table 7: Short Run Dynamic Vector Error Correction Model DPOt-1 DGDPt-1 DCPSt-1 DLQLt-1 Dynamic model Dynamic model Dynamic model Model Variables Coefficient Coefficient Coefficient Coefficient DGDPt-1 0.34049 0.056878 7.23E-108 * 8.25E-10 * (-2.049) (0.25746) (2.3632) (2.4447) DPOt-1 0.03456 6.3503403 -0.137726 -0.131949 (0.7834) (0.63649) (1.0717) (1.07609) DLQLt-1 -0.16854 -8.31404348 0.597764 (0.2710) (1.01907) (2.71643) DCPSt-1 -0.24468 3.845678 0.7113716 * (0.36) (1.345) (3.37778) DTRt-1 -0.18138 9.9914797 0.296444 * 0.251389 (0.137) (1.16402) (2.55193) (2.18389) DRIt-1 -0.7625 9.361441 * 0.12236 * -0.026292 (0.402) (1.7911) (1.6182) (0.36604) ECMt-1 0.058224 * 0.18721 * 0.570495 * 0.208994 * (-0.587) (-2.51151) (-2.4182) (-2.0157) Note * ** *** indicates the significance level at the 1%, 5% and 10% respectively. ARDL(1,0,0,0,0,0,1) selected based on Schwarz Bayesian Criterion. Numbers in bracket are the t-ratio suggesting the significant of the parameters. Coefficients with ** indicates significant at 5% level. Table 8: VAR Granger Causality/Block Exogeneity Wald Test, PO as Dependent Variable Excluded Chi-sq Prob. GDP 13.99561 0.2029 CPS 0.683408 0.8771 LQL 1.509796 0.6800 TR 7.048445 0.0704 RI 1.970955 0.5785 All 14.54219 0.4849 Notes * ** *** indicates the significance level at the 1%, 5% and 10% respectively. Table 9: VAR Granger Causality/Block Exogeneity Wald Test GDP as a Dependent Variable Excluded Chi-sq Prob. PO 2.492726 0.4766 CPS 0.883118 0.8295 LQL 2.363688 0.5004 TR 5.388433 0.1455 RI 1.400474 * 0.0054 All 12.41527 0.6474 Note * ** *** indicates the significance level at the 1%, 5% and 10% respectively. Table 10: VAR Granger Causality/Block Exogeneity Wald Test CPS as a Dependent Variable Excluded Chi-sq Prob PO 0.369909 0.9464 GDP 13.63246 * 0.0035 LQL 1.094464 0.7784 TR 3.071064 * 0.0058 RI 0.270316 * 0.0065 All 14.72064 0.4717 Note * * * indicates the significance level at the 1%, 5% and 10% respectively. Table 11: VAR Granger Causality/Block Exogeneity Wald Test LQL as a Dependent Variable Excluded Chi-sq Prob. PO 4.118325 0.2490 GDP 10.23180 * 0.0167 CPS 0.188340 0.9794 TR 9.670154 * 0.0216 RI 0.526657 0.9130 All 14.21971 0.4089 Note * ** *** indicates the significance level at the 1%, 5% and 10% respectively. Table 12: VAR Granger Causality/Block Exogeneity Wald Test TR as a Dependent Variable Excluded Chi-sq Prob. PO 1.643073 0.6497 GDP 11.61878 * 0.0088 CPS 5.218340 0.1565 LQL 1.177704 0.7584 RI 3.228509 0.3577 All 14.83084 0.4637 Note * * * indicates the significance level at the 1%, 5% and 10% respectively. Table 13: VAR Granger Causality/Block Exogeneity Wald Test RI as a Dependent Variable Excluded Chi-sq Prob. LPO 6.767211 0.0797 LGDP 2.382484 0.4969 LCPS 5.846998 0.1193 LLQL 6.913989 0.0747 LTR 12.42697 * 0.0061 All 14.21872 0.5089 Note * ** *** indicates the significance level at the 1%, 5% and 10% respectively.