Determinants of Credit Growth in Africa

The importance of bank credit as an essential instrument that determines the efficacy of the financial system has been widely discussed in the literature. Some studies even opine a minimum threshold for it to generate the desired impact while others suggest efficient allocation rather than volume of credit. All these underlie the need for an increase in bank credit without overlooking the efficient utilization hence our choice of private sector credit as proxy. The paper conducts a panel study of the main factors that propels credit growth and covers 1970 – 2006 for thirtythree African countries. The research finds that real export is inversely related to real private sector credit while real capital inflow and real imports is positively related to real private sector credit.


INTRODUCTION
The finance literature provides support for the argument that countries with better and efficient financial systems grow faster while inefficient financial systems bear the risk of bank failure (Kasekende, 2008).In a review of finance literature, the study opined that better functioning financial systems ease the external financing constraints that impede firm and industrial expansion.Banks accept deposit from individuals and institutions thus transferring funds from the surplus sector to the deficit sector of the economy (Mishkin, 2007).Though they are subject to certain regulations by the regulatory authorities, financial intermediaries still determine the rules for allocating funds, and as such they play a significant role in determining the type of investment activities, the level of job creation and the distribution of income (Gross, 2001).
Generally, private sector credit is favoured by researchers as a proxy for financial development.The importance attached to the use increases over time thus studies use different measures of the variable overtime (see for example Beck et al 2005;Levine 2002;Odedokun 1998;King and Levine 1993).Boyreau-Debray (2003) uncovers a negative correlation between growth and banking debt due to the fact that Chinese banks were mobilizing and pouring funds into the declining parts of the Chinese State Enterprise, and hence the system has not been growth promoting.Demirguc-Kunt and Levine (2008) emphasised the importance of focusing on allocation of credit to the private sector as opposed to all bank intermediation while Beck et al, (2005) highlight the importance of private credit as a strong predictor of growth.The recent study by Crowley (2008) also supports this postulation and emphasises the importance of private sector credit.
Despite the above observations, some recent studies suggest a reducing impact for financial proxies.According to Demetriades & James (2011) in a study of eighteen Sub-Saharan African countries, they observe that while bank liabilities still makes some impact on the economy, private sector credit does not have any impact at all.Kumar (2011) also gave a similar opinion that financial development proxies do not have any significant long run effect on per worker income.He suggests developing exports and remittances market for long term sustainability of the economy while efforts are geared into searching for innovative ways to make the financial sector more integrated to economic activities.These papers however did not discuss about the concept of threshold postulated by Rioja & Valev (2003) as many African countries are known to have low ratio of private sector credit to total volume of credit.
Albeit the importance attached to bank credit in the literature, the factors that determine its growth are under-researched.This paper serves to cover the vacuum and add to the literature in that area.It also tries to highlight the current state of the continent and assist in identifying essential factors that are instrumental for growth of private sector credit.

Overview of Africa
Africa consists of fifty-four (54) countries, most of whom they classify as under-developed.Presently, the continent is the least developed in the world with about twenty-five of these countries ranking high in the list of impoverished countries in the world.One reason attributable to the level of poverty in Africa is the level of corruption.It is practically difficult to transact a genuine business without having to grease the palm of the officials.This to some extent widens the gap between the rich and the poor with some individuals even said to be wealthier than their country.Table 1 below highlights some details about some of the countries that are included in the empirical analysis.The above data shows that twenty-four of the forty-one countries included in the table are classified as lowincome countries; eleven others falls within the lower middle income countries while six are categorised as upper middle income countries.This confirms the earlier assertion that most of the countries within the continent are classified as low-income countries hence the level of poverty.All the low-income countries exhibit similar features such as very low GDP per capita.The lowest for the group was Burundi, which had 107.87 while Nigeria had the highest, which were 794.08.All these figures are very low and underscore ability for development.It is therefore not surprising to see that the GDP per capita growth for some of the countries namely Burundi, Benin, Cote d'Ivoire, Guinea-Bissau, Malawi and Togo were negative for year 2005.Even Lesotho that falls under the lower middle-income country also has a negative per capita of 0.08.A recent report by the World Bank (ADI; 2008) stated that the GDP of the Sub-Saharan African countries was $744 billion.This is just 28% of China; 69% of Brazil; 74% of Russia and 80% of India.Out of this, Nigeria and South Africa accounts for almost 60% of the sub-regional GDP.
For the financial variables, the observation is not significantly different from what we earlier discussed.Essentially, when we express liquid liabilities as a percentage of GDP, the percentage for Chad is as low as 0.07, Angola was 0.12 while Congo was 0.13.Similar with the exception stated in respect of Lesotho above, Sudan, which is a lower middle-income country, recorded the lowest figure of 0.001 for the whole series.This implies that the spate of under-development transcends the low-income countries, as some of the signs are visible with the middle-income countries.If the postulation of Rioja and Valev (2004) is anything to take into consideration, not less than fifteen (15) of the listed countries had their ratio of liquid liabilities to GDP below the estimated figure of 0.20.The basic inference is that liquid liabilities may not be contributing significantly to growth in these countries.
The ratio of Private Sector Credit when expressed as a percentage of GDP can be described as low.About twenty-five (25) countries in the sample are below the Rioja and Valev (2004) estimated threshold of 0.14 required for Private Sector Credit to exert a meaningful impact on growth.Specifically, sixteen countries are even within the range of 0.001 to 0.08.This range has the highest figure at about 50% of the estimated minimum requirement.The basic question is "How does one expect a positive impact from intermediation, if the bulk of the fund is channelled to non-growth promoting areas of the economy?Some countries like Sudan, even though classified as lower middle-income country has the ratio of private sector credit to GDP as low as 0.001 and Guinea-Bissau with 0.01.What is more pertinent is that many countries that are classified as upper middle income countries such as Gabon and Libya; and those that are classified as lower middle income countries such as Algeria, Angola, Cameroon, Congo and Lesotho, all exhibited very poor ratios.This shows that the problem with allocation of credit to the Private Sector transcends the income level, but a peculiar situation with most of the countries within the continent.The chart in figure 1 below depicts the aforementioned situation and shows the current volume of credit that is available to the Private Sector by these countries.According to Honohan and Beck (2007), there is still a long way to go for finance to have a desired impact on African countries.This they attributed to limited access by small firms and households to any formal financial services, especially in the rural areas.In accordance with the submission of Honohan and Beck (2007), the size of the financial system in Africa is relatively small.

Factors Affecting Credit Growth
Increase in credit generates growth hence GDP is a potent factor that affects credit growth.This is attributed to the role of financial institutions in allocating savings effectively to the productive sectors thereby enhancing economic efficiency and capital accumulation (Rajan & Zingales, 1998;Beck, Levine &Loayza, 2000 andLevine, Beck &Loayza, 2001).This variable is used as the main explanatory variable in this study.
The level of trade is another factor that is said to affect credit growth.This consists of exports and imports.Specifically, countries in Africa produce raw materials which they export to the advanced countries where they are processed and subsequently imported back to the continent.Trade inflows exert positive influence on banks as it increases bank liquidity thereby make more funds available for intermediation purposes.In addition to that, exports earnings are diversified internationally and are not usually affected by local disturbances.
Foreign inflow is another variable discussed in the literature that impacts credit growth but there is no consensus opinion about the effect so far.Crowley (2007) finds that foreign inflows are significant for growth of credit in Slovak Republic.Several other previous studies support this assertion (Arvai, 2005 andDuenwald et al, 2005).However, Cottarelli et al (2003) posited that domestic savings flows is the main factor responsible for the growth of credit in Eastern Europe, and as such there was no evidence that foreign inflows was significant in stimulating credit growth.

Research Question
Based on the aforementioned, it may be apt to state the research question as:-What factors are significant in determining credit growth within Africa?

Data
The data for this study is from the World Development Indicator (WDI) 2008 dataset and the International Financial Statistics (IFS).The study covers thirty -three African countries.
The analysis covers the period from 1970 to 2006.Availability of data underlies inclusion of countries in this study.

METHODOLOGY
The determinants of credit growth are a prominent discussion in the credit literature.What is very clear is that, there is no universal model for dealing with this issue.According to Rioja and Valev (2003) in their study of seventy-four countries divided into three regions of low, medium and high based on the level of their financial development.They find that what appears not to have statistical significance in one area may have a positive significant effect in other areas, even with varying degrees of significance.According to them, financial development can only exert positive influence only when it has reached a threshold, thus the situation with the low region (developing economies) is uncertain mainly because it is below the threshold.
The paper uses variables as defined by Crowley (2008) in their study of Middle East, Mediterranean North Africa and Southwest Former Soviet Union countries of Central Asia to determine the factors that are crucial in driving credit growth.The panel method of estimation is used.All essential tests to this approach were taken into consideration and are reported in the table of results.

Analytical Method and Model Formulation
The combinations of the variables used in the model are stationary at level as reported in the co-integration result reported in table 2 below.The relationship that exists between Private Sector Credit and GDP is revealed in the correlation result presented in table 3 below.

Table 3: Correlation Result between Private Sector Credit and GDP
Variables PRIVATE SECTOR CREDIT GDP 0.143 (0.000) From Table 3, private sector credit is highly correlated with GDP.The model that we test in this study is - Where:β 0 denotes Constant; Real Trade Growth is used to proxy total exports and total imports while Real Total Capital Flow is used to proxy foreign capital flow.
Data for the study is from the World Bank (WDI) database.All variables are in their real values.The study uses normal random effects and random effects GLS regression with AR (1) disturbances methods; both of which produces similar results as presented in Table 4 below.The regressions with normal random effects are 1a to 5a; while those with the GLS are named 1b to 5b respectively.The hausman test supports the random approach for the study.Each of the five regressions represents different models through the inclusion of additional variables considered to impact credit growth.

INTERPRETATION OF RESULTS
From table 4, the intercept is significant for all the regressions.This is contrary to the findings of Crowley (2008) who had all the intercept not significant for his regressions.The growth rate of GDP is significant at 1% and consistent with the findings of Crowley too.An observation in the result is that Private Sector Credit is significant only in regressions 2, 3 and 4 i.e when real capital inflow is included as one of the variables.The coefficient for lagged private sector credit is not large and negative.The trade variable included, (exports) has a negative coefficient.A continent that possesses natural endowments which are exported to other parts of the world has a negative coefficient for such an important channel of growth.The coefficient is also not large, but significant at 5% all through for regression 1 and 3.This observation is similar in all the regressions.The inclusion of real capital inflow to the regression improves the level of significance of the variables.The level of significance for the intercept changes from 5% in model 1 to 1% in model 2, while real private sector credit is also significant at 5% and 1% for the random effects and random effects with auto-regressive disturbances (AR) approach respectively.The inclusion of real capital inflow to model 3 had similar effect like model 2, thus all variables included in the regression are significant at varying levels.The R 2 also shows slight improvement although the coefficient for real capital inflow is very tiny, but positive and proves to be important in driving financial development within the continent.
Import growth included into model 4 shows positive result.The coefficient is positive and large.It is also significant at 1%.We therefore postulate that the trading activities of companies within the continent mostly those engaged in import activities has a positive and significant contribution to the development of financial development.The inclusion of both real capital inflow and real imports gave the best R 2 of about 34% effect on financial development obtained throughout the regression results.Based on this result, the combination of real capital inflow and real imports are variables that are very significant in driving financial development within African continent.The widely supported real export exhibits a negative relationship with the proxy for financial development.
Import growth is included in equation 1 and the result presented in table 5 below shows a positive result as expected.This suggests that imports exert positive influence on financial development in contrast to the result with exports growth.This is attributed to the banking system intermediating for importation activities and the fact that the proceeds are repatriated back to the respective economies.

CONCLUSION
In this paper, I examined the factors that propel credit growth.The result suggests that the financial institutions are not very relevant in intermediating for trade, mostly exports that happens within their environment.Real exports exhibit negative relationship with financial development while variables such as real capital inflow and real imports are positive and significant hence relevant for driving financial development within the continent.The basic inference from this is that financial institutions have been basically financing local businesses that are engaged in importation of goods and services while major aspect of trade (exports) is not captured within their operational horizon.This may be because most of the companies handling the domestic export trade are foreign oriented hence source for credit within their respective area of strength.Likewise, it may be that the domestic financial institutions are not strong enough to meet the financial requirements of these companies.As a result of this, these foreign companies look beyond the shores of their operational base to seek for financial assistance.Likewise, it may be that these companies divert the proceeds of exports to foreign accounts or private sources where those involved find it difficult to explain the source of such funds.Whatever may be the reason responsible for this situation, it is not beneficial to the continent and needs to change so that the continent can be on the path of sustained and beneficial growth.

Figure 1 :
Figure 1: Private Sector Credit as a ratio of GDP for African Countries in 2005Source: -The World Bank DevelopmentIndicator (2007)

Table 1 -Economic and Financial Highlights of Some African Countries in 2005
L means Low Income Country while LM and UM means Lower Medium Income Country and Upper Medium Income Country respectively.

Table 2 : Co-integration Result for the Variables used in the Models
Figures reported are the p-value for each combination.The combination includes other exogenous variables which are Exports, Capital Inflow and Imports.

Table 4 -Panel Regression Output Of Credit Growth (Rpscrgdp), 1970-2006
Note: Figures in parenthesis ( ) are the p-values of the variables.The symbols of ** and * depicts 1% and 5% level of significance for the coefficients and with the expected sign while ## and # also denotes significance at 1% and 5% level of significance but the sign of the coefficient does not tally with the literature.The symbol of * in the diagnostic section denotes significance at 5% or 10% level.Regressions numbers with a and b represents approaches using random effect and panel with AR (1) disturbances respectively.KEY: -RPSCRG is Log of Real Private Sector Credit Growth; RGDPG is Log of Real GDP Growth; RIMPG is Log of Real Import Growth; REXPG is Log of Real Total Export Growth; RCAPACG is Log of Real Total Foreign Inflow Growth; RPSCRGDP is Log of Real Private Sector Credit to GDP