Effect of Financial Liberalization on Savings and Investment in Nigeria

In Nigeria, measures were taking by government between the period 1986 to 2005 to liberalise the financial sector as part of the overall structural adjustment programme (SAP) with the aim to promote economic growth and alleviate poverty. The financial reform assumes that deregulating the financial market will eliminate the distortion and fragmentation of the financial market. It will also improve the mobilisation of savings as well as efficiency in allocation of funds to viable investment projects in the economy. This in turn will help in accelerating economic growth. Literature in the area posits that liberalising interest rate and deepening of the financial market is growth enhancing. This hypothesis is tested by identifying two basics channel through which financial liberalisation is operational – savings and investment, and allocative efficiency. The analysis was carried out using macro data from Nigeria by employing Two-Stage-Least Square (2SLS) instrumental variable technique to analyse the data. Major findings pointed out that interest rate as determinant of savings is not statistically significant, and rising interest rate negatively affect investment growth. Secondly, investment follows the standard accelerator theory. Furthermore, growth in the stock of money is positively related with investment but statistically insignificant. Lastly, there is a considerable deepening in the banking sector, however, the capital market does not


Introduction
Up to the middle of the 1980s, Nigeria's economic policy was more of direct control. Interest rate, exchange rate and credit allocation to different sectors in the economy were not on the basis of market determination, but based on government directives (Ayogu, Emenuga, & Soludo, 1998). The prevailing situation was challenged based on the following reasons: passive competition; poor customer service; significant government ownership in the banking sector; sectoral credit allocation; limited expansion of domestic credit to private enterprises; interest rate regulation; restriction to flows of foreign direct and portfolio investments due to capital account control; and fiat monetary policy (Ayogu et al., 1998).
As from 1986, the economy gradually started moving away from direct control to a more liberal or deregulated economy. As a result of that, several financial reforms were introduced, even though it encountered several policy reversals over the time as well. Broadly speaking, these reforms were referred to as financial liberalisation reform. and improving the linkages between formal and informal financial sectors (CBN, 2007).
Thus, the philosophy behind the financial liberalisation is in twofold: first, quantity effect through generating higher savings and investment in the economy, and second, quality effect by efficiently allocating capital to profitable investment. If these set of objectives were achieved, it is expected that the economic growth will be enhanced. The downward trend in the key macroeconomic indicators will be reversed especially unemployment, poverty and inflation. Although, financial repression discourages savings, but Nwaobi (2003) argued that in the case of Nigeria it is not as serious as the financial liberalisation proponents hypothesised. However, Akande and Oluyomi (2010) when examining the relevance and application of the theoretical prescription of the Two-Gap model in Nigeria noted that inadequate savings and foreign exchange are responsible for low investment and growth of the 104 economy. Similarly, statistics from World Bank Development indicators in 2016 with respect to domestic credit to private sectors shows that Nigeria's domestic credit to private sector as a percent of GDP was 15.68 percent, and as such ranked at 146 th position in the world. This posed a question as to whether the dwindling growth of the Nigeria's economy was due to savings or credit deficiency.
Despite the strategic reforms, Nigeria's growth performance has been dwindling for the past three decades, making it difficult to address the macroeconomic issues facing the country. This called for the study to assess the policy and examine it to see whether the poor performance was as a result the policy or other economic issues are behind the stagnation. The question now is why did the performance of the economy has been dwindling for the past three decades with regards to savings and investment growth? Or, in other words, why is there no steady growth in savings and investment in the economy? Can this poor performance be associated with the financial liberalisation reform? Is there evidence that the reform has led to efficient allocation of capital in the economy? Previous studies focused on the quantity effect of the reform, but less attention has been given to the issue of allocative efficiency. Thus, the objective of the study is to find out whether financial liberalisation has improved on the growth of savings and investment in Nigeria, and or test whether the reform has also improved the efficient allocation of credit in the economy.
The paper is structured as follows; section II reviewed the performance of the economy before and after the reform. While, section III reviewed the related literature pertaining to the impact of financial liberalisation on savings and investment. Section IV developed the econometrics 105 methodology used in analysing the impact. Three models were developed and Two-Stage Least Square instrumental variable regression technique was employed in analysing the data due to presence of endogeneity in the models. Section V present and discussed the results of the findings. Section VI concludes and offered recommendations.
The study is of significant as it shed more light on the controversies surrounding the effect of financial liberalisation on savings and investment growth which has long been debated within the academic circle. In the Nigerian context, the study pointed out that interest rate does not significantly and positively contributes to savings and investment growth in Nigeria. The study is also of significant as it shows how the Nigerian economy is credit constrained contrary to the assertion that the economy was savings constraint. Another significant contribution of the study is the way it observed investment in the capital market crowd out real sector investment and call for government active policy rather than passive policies to drive the economy further. Table 1 summarises the performance of the economy before and after the liberalisation period looking at some key macroeconomic variables. The average growth of GDP in the 1980s was -2.04 percent. This slightly improved to 2.62 percent in the first decade after the reform.

106
Although, the economy has undergone structural transformation towards the late 1980s, but it can also be argued here that the changes in the GDP growth between the 1980s to 1990s was not as a result of improvement in the sectoral performances rather it was as a result of improvement in the oil price (Ibrahim, 2018). Findings by Ibrahim (2018) shows that oil price exerted positive influence on aggregate national output, however, with respect to manufacturing and agricultural sector the influence is negative. In the early 1980s world oil price 107 crashed sharply, this hit the economy negatively. The study further argued that fluctuations in oil price induces uncertainty in the real sector and also weakens the effectiveness of government fiscal management.
Although, the economy continued to grow at an average growth rate of Similarly, the average growth of manufacturing GDP in 1980s was 18.37 percent. However, this continually decreasing to as low as 8.43 percent in 2010 to 2017 period. Employment in the sector also declined persistently for example, as at 1980s, the number of labour force in the textile industry stood at 700,000. However, by the end of 1995, this fell to 40,000 (Newman et al., 2016).
In 1980s, the proportion of service sector in the GDP was 2.69 percent, this significantly rise to 3.75 and 11.01 percent in the first and second decades after the reform respectively. The statistics above shows that Nigeria's service sector, even though performed poorly, but gained steady growth as compared to agricultural and manufacturing sectors, In fact, the performance of the real sector based on the statistics discussed above shows that the economy has experienced a dwindling performance during the period of the financial liberalisation in Nigeria.

Savings and Investment under Financial Liberalisation
According to McKinnon (1973) and Shaw (1973) high interest rate that evolve in economy after liberalisation will reduces inflation rate, induces savings to grow and this in turn induces higher rate of return and consequently increases investment and leads to economic growth.
Previous studies attempt to investigate the link between savings and investment and many more on financial liberalisation dwelled on whether it has positive effect on savings and investment. Attanasio, Picci, and Scorcu (2000) for example study the long run and short run correlation between savings, investment and growth for 123 countries over a period of 1961to 1994. The research concluded that there exist positive and significant relationship between lagged savings and investment rates. Secondly, it also found a negative causality between investment and economic growth. And lastly, positive causality running from growth to investment. Since the study is cross country, it is important to have a country specific study to critically examine the validity of this conclusion and particularly re-examine the assertion of negative causality between investment and growth which contradicts most of the results obtained in the literature like that of (Barro, 1991) and (Barro & Lee, 1993).
Change in reforms may account for this negative or mixed results obtained in the previous studies. Abiad, Oomes, and Ueda (2008) noted this, and argued that basically to two main reasons might account for the mixed or ambiguous quantity effect of financial liberalisation reform for instance. Firstly, increase in rate of return due to improve in risk sharing and or removal of interest rate may depend on whether income or substitution effect prevails. If income effect prevails, quantity effect will be negative. But if substitution effect holds, the quantity effect will be positive. Secondly, even if financial liberalisation provides better insurance against risk, this may possibly lower the incentive to save for future (Devereux & Smith, 1994). Therefore, financial liberalisation can improve the functioning of financial market without necessarily increasing the volume of savings and investment.
De Melo and Tybout (1986)  There was also a belief that, the impact created by the reform in boosting credit in the economy was the inducement of borrowing on the part of firms that wanted to conclude their projects before the commencement of the devaluation as well as those that were trying to fund their working capital requirements. Credit expansion continued through the 1990s. The expansion of credit in the 1990s can be linked to the directives of the government for banks to grant loans to general public without much restrictions for the purchase of share in other to accomplish the privatization exercise (Ayogu et al., 1998 (Okpanachi, 2012) but after lifting the ban regarding dividend and profit repatriation, the stock market begun to experience rapid growth that attract some global speculators into the Nigerian economy as noted by (Okpanachi, 2012).
Therefore, as from 2008 the total credit to private sector started declining which mark the beginning of the country's financial crises. In summary, according to McKinnon (1973) and Shaw (1973), financial liberalisation affects the economy through two channels; savings and allocative efficiency. Firstly, through savings channel, financial development affects the real economy by enhancing the volume of savings in the economy. This is because savings levels are sensitive to real interest rates, nominal interest rate and inflation reduces the amount of national income allocated to capital formation. This implies that in a financially repressed economy, the financial market after liberalisation will stimulate higher returns on savings instrument like the savings account via high interest rate. This will help to attract savings from the informal sector into the formal sector.
The second channel is through allocative efficiency. As the financial institutions generate more savings, they will be able to extend the accumulated savings into productive investment thereby increasing the overall economic performance (Mamoon and Nicholas, 2017). One medium through which the financial institutions can foster allocative efficiency is by extending credit to private sector which is considered to be more efficient than to public sector. Because of high efficiency in the private sector, they were expected to utilized the funds efficiently by investing in a more profitable project. Hence, the country's productive efficiency improves while economy is set at a positive growth track.

Model Specification
This section presents some models based on the work of De Melo and Tybout (1986) to account for the relationship between domestic savings and some financial liberalisation indicators. However, the specification here differs slightly from their own work as this study account for the role of stock market development in the liberalisation processes. The model is specified as follows; = 0 + 1 + 2 + 3 + 3 −1 + 3 + 4 + … … … … … … … (1) Equation (1) above states that domestic savings, , depends on: real income, , real deposit rate , foreign saving , lagged savings −1 , stock market liquidity and stock market capitalisation .
Real income growth is included in the model to account for the deviation from permanent income as suggested by the standard theory.
Changes in income influences both savings and consumption as emphasised by (Keynes, 1936). According to him, rise in income often accompanied by a rise in savings. Therefore, its coefficient is expected to be positive. Lagged saving is included in the model to capture the adjustment process which is assumed to be spread over multiple periods and its coefficient is assumed to be positive. Foreign savings is included in the model to capture the degree of substitutability between foreign savings and private savings. If foreign savings crowd out private savings, the resultant effect on the coefficient would be positive.
considering individuals may consume more at any given rate of capital  (1) should be handle with caution. Secondly, foreign savings may not be necessarily determined exogenously. This is because as the country's capital account is opened following liberalisation, economic agents may opt to optimised their present and future consumption of both domestic and foreign goods at any given interest rate, exchange rate, income and expectation. Lastly, financial liberalisation in Nigeria follows some policy reversal. Therefore, it is important to account for some 117 institutional changes that took place within the period of study such as interest rate control reversal.
To address the issues raised above, equation (1) is re-specified and estimated with private savings instead of total savings and a dummy variable is included as well to capture the structural break. Thus, equation (2) is specifies as follows; = 0 + 1 + 2 + 3 + 4 −1 + 5 + 6 + 7 Furthermore, in order to test whether factors associated with financial liberalisation explained efficient allocation of investment in the Nigerian economy, another model was developed. It is important to recall that when the economy is financially repressed, firms' investment is depended largely on their accumulated incomes. However, in a liberalised economy, real price of capital and the expected output demand play significant role in this aspect. Hence a model in the form of equation (3) is necessary that relates firms' investment with income as follows; = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 −1 + 9 −1 + 10 −1 + … … … … … … (3) Equation (3) states that investment depends on its lagged value −1 , income and its lagged value −1 , interest rate , exchange rate , real money growth and its lagged value −1 , market capitalisation , stock market liquidity and a dummy D that captures that effect of institutional changes, if any, to see if there is any significant break within sample period.

118
The complication in estimating these three equations is the possibility of inconsistent parameter estimation arising due to endogenous regressors.
In these models, interest rate and income or both may be endogenous with respect to savings in equation (l), as well as with exchange rate or both in equation (2)    From the regression results, interest rate appeared to be significant and positive in all the cases except in column 4. In addition, the correlation between the interest rate and savings tends to be moderate ranging from 15 percent to 30 percent. This provide a significant support for the claim that financial liberalisation stimulate savings controlling for other factors. If interest rate raised by 1 percent for instance, savings will grow by 15 to 30 percent. Furthermore, foreign savings also shows strong positive correlation with savings. In line with a priory expectation that financial liberalisation would cause foreign savings to crowd-out domestic savings. The coefficient in all the regressions is greater than one, meaning that foreign savings crowd out domestic savings in more than one-to-one fashion. Market capitalisation also appeared to crowd out domestic savings, albeit weak correlation. This is because of the assumption that stocks and bonds provide savings alternative to household and investors. As such, the available savings in bank will decline with a slight improvement in stock market due to the financial reforms. As stated above, the specification of model (1) slightly contradicted what has been posit by the financial liberalisation tenet, it is important re-estimate the model with private savings and observe the results. Table 3 reported the regression results based on the five different hypotheses about the issue of endogeneity of the parameters estimated.

Results and Discussion
In the first column it assumed endogeneity. In the column 2-5, it tested the presence of endogeneity using income growth, interest rate, foreign savings and both income and interest rate.
The discussion is as follows.

Conclusion
Previous research on financial liberalisation in developing countries identified number of channels through which financial market reforms affect allocation of resources. Firstly, increase in interest rate may lead to growth in savings. Secondly, relaxation of financial constraints may lead to growth in investment. The empirical relevance of these two effects to Nigerian economy was tested in this study and the major findings are presented here below.
In equation (1), total domestic savings was used and the results shows that financial liberalisation enhances savings growth in Nigeria and some key variables assumed to determine savings following the literature were included in the models. Virtually all the determinants are statistically significant and in line with the expected signs. However, based on the assumption that financial liberalisation affects private savings not total domestic savings, the model was re-estimated using private savings as dependent variable instead of domestic savings. The second regression significantly differs with the first even though, similar in some aspect. In the second estimation, interest rate, which is the key element, plays no significant role in determining savings in Nigeria.
While financial depth measured by stock market capitalisation reduces both domestic and private savings. Both domestic savings and private savings falls with a rise in income. But fall is larger in domestic income than in private savings 129 Concerning investment model, the result shows that accelerator theory exerts significant impact in explaining the investment behaviour of the Nigerian economy throughout the sample period. The standard accelerator theory posits that investment depends on expectation. And from all the four different investment models that were estimated, the results show that the lagged investment is significant and strongly correlated with investment. The conclusion is that expectation is more relevant in explaining investment behaviour in Nigeria within the study period than savings constraints. In addition, the downward structural shift, suggested that the financial liberalisation reform does not provide concrete evidence about improvement in investment in the Nigerian economy. Interest rate which is key in the reform exercise appeared negative, and insignificant in some cases same with money growth and measures of stock market performances.
The findings of this study show that Nigeria is not savings constrained, but rather credit constrained. Since interest rate plays no significant role in determining savings in Nigeria, and a unit increase in the interest rate may also lead to significant decline in domestic investment. This can be clearly seen in Table 1 where investment in fixed capital formation has been continually declining while savings remains relatively stable. The study therefore recommends that policies that favours liberalising the interest rate in the country is of little importance. What matters much is addressing other macroeconomic issues that impede the allocation of credit to productive projects such as interest regulation, income and consumption policies as well as reducing macroeconomic uncertainties.
The study also recommends that future research should focus on addressing the micro channels through which financial liberalisation affects savings and investment in Nigeria.