|Home Instructors Journals ContactUs|
MONETARY POLICY AND INVESTMENT BEHAVIOUR IN NIGERIA
Lawrence Ayemere Ibadin
Keywords: Monetary Policy, Investment Behaviour
However, investment, for the purpose of this study, refers to investment in physical assets, that is, in machinery and equipment by the private sector of the economy. As such, a genuine and adequate expenditure on business fixed assets will increase the level of economic activities (through increased production of goods and their services). This is capable of generating employment opportunities in the industrial sector, reducing or stabilizing the price levels, raising the Real Gross Domestic Products (GDP) values, and if properly directed, reducing the pressure on the country's balance of payments position.
The various macro-economic problems that beset the economy today, ranging from escalating price levels, growing unemployment, dwindling import earnings, to rising interest rates are economic consequences of stilled private sector investment spending (especially, investment in machinery and equipment).These are known to constitute the machinery of transformation of economic inputs into intermediate and final goods arid services and, so determine the level of economic activities.
A situation where, for example, expenditure on machinery and equipment accounts for only 26.9% of gross fixed capital formation (GFCF) in 1981, 31.8% in 1986 and 49% in 1989 (see table 1) shows the relative scantiness of aggregate investment in the economy. The consequence of this is a depressed level of the national economic activity.
because of evasive tendencies of corporate bodies; low performance of industries in terms of profits and capacity utilization and poor tax management system, amongst others. Beside, effectiveness of export policies is often dampened by dearth of exports (only crude oil accounts for about 80% of the country's total export, earnings, organized foreign competition and protectionist policies, low value of the naira value vis-a-vis those of other currencies, as well as the high cost of local manufactures.
Without reasonable doubt, government's spending has not been effective in adjusting and stabilizing the economy because they have not been utilized as "pure1' control mechanism but have been made anytime the government felt like spending. Moreso, some of such expenditures have not been on productive ventures (e.g NOA,NDE, etc) and therefore, have only succeeded in causing excess liquidity in the economy and fueling the inflation rate (given as 13.3% as September,2006).
Objective of the Study
We randomly sampled individual and institutions (private and public) as well as the government that invest. More essentially, we concentrated on both individuals and institutions that invest why government was mainly seen as a regulator of investment environment via its fiscal and monetary policies.
Dickey-Fuller (DF) and the Augmented Dickey-Fuller (ADF) tests were used. The choice of these statistics enable us to ascertain the long run relationship between the (Cross Private Domestic investment in plant and machinery and the explanation variables. Beside, the Engle-Granger two-step technique was used to test the existence of co integration of the variables. The choice of the technique was particularly interesting against the background of its ability to eliminate co integration from the explanatory variables.
R = minimum rediscount rate stipulated by CBN (dl/dK < Q)2 Ms= money supply (i.e M2) (di/dMs > 0)
Introducing appropriate lag function to allow for possible adjustment, equation (1) was re-written as:
This model (equation 3) is a modified version of Chenery's Stock Adjustment model of investment. However, for estimation purposes, we used the sample regression specification of equation (3), that is:
Model Estimation Techniques
(1) and the explanatory variables. Besides, it provides the relevant clues as to the order of integration of variables, that is, as to whether they are stationary or non-stationary.
This is closely followed by the test of co integration. To do this, a number of techniques are available, such as (Stock and Watson, 1988), (Engle and Granger, 1987), (Johnsen and Juseiius, 1990). However, for the purpose of this study, we shall adopt the tingle-Granger two-step technique to test the existence of co integrating relationship, because of its intuitively appealing qualities, one of which is that, it ignores the possibility of co integration among the explanatory variables. To this end, the first step entails the conduction of unit root test (DF, ADii, etc) on each individual series.
The next stage involves finding out the order of integration on the residual generated from static model using, principally, the DF and ADF. The results obtained from these tests would enable us to ascertain whether or not co integration does exist.
Finally, using the criterion of maximum R-bar squared to select the best fitting equation of the error correction model (ECM) which, ipso facto, captures the long-run relationships between the relevant variables, will be attempted where co integration is confirmed to exist.
Table 1 DF/ADF Unit Root Tests Results
Where * indicate AD test statistics
The results above show that all the variables used in this study are stationary at first difference except Aggregate Credit to private sector. Thus, the use of this variable (Aggregate credit to private sector) in regression analysis that follows would make the regression results spurious as shown by Bngie and Granger (1987). Hence, it was not further utilized in the study. Overall, what this therefore implies is that we can go ahead to check whether the model specified is co integrated.
Accordingly, the results from the test of non-stationarity of the residual (ECM) in equation (3.1) are presented below:
Table 3: Dickey - Fuller/Augmented Dickey - Fuller Unit Root Test on Residual
Source: Dicky Fuller Root Test
The test on the residual from the sate regression of gross private domestic investment on the variables that affect it was significant at the 5% level as shown in table 2 above. We thus rejects the null hypotheses of no co integration and conclude that the variables are indeed co integrated.
The normalized co integrating coefficient gives the long-run relationship and the result is presented below:
I = - 14987 + 0.028Y,-i + 0.357Kt-i + 2362.6R
S.E. of Regression - 26437.0
E (i) =104860.5
The long-run relationship (results) shows that only one period lagged nominal (JDP (Y t-i |) is significant at the traditional level of 1% significance level. Hence, one period lagged nominal GDP is the long-run determinant of gross private domestic investment.
Using the criterion of maximum' R - bar squared to selected the best -lilting equation which, ipso facto, yields the required parsimonious representation of the error correction model, the following estimated equation was obtained.
Dl - -5225.8 + 0.023Ay, + 0.446AK
The results above were obtained from the selected ARDL (Autoregressive Distributed lag) model.
An examination of the econometric results shows that the overall goodness - of fit is satisfactory. The value of R2 = 90, indicates that above 90% of the systematic variations in the dependent variable is explained by the error correction model over the period of analysis. The value of the F-statistic (0.008) indicates that at 1% significance level, there exists a significant linear relationship between the dependent and the independent variables. The coefficient of ECM is significant at the 1% significance level and it is negative. Thus, it will rightly act to correct any deviations from long-run equilibrium. There is the absence of the problem of autocorrelation. As can be seen above, the entire variable have coefficient that are of expected signs. All the relevant explanatory variables pass the t-test at 5% significance level except one-period lagged income, additions to investment and change in money supply.
Conclusion and Policy Recommendations
Tests conducted on the order of integration showed that they are stationary. However, the statistical tests on co integration between the dependent and explanatory variables revealed that only the one-period lagged nominal GDP is the only long-run determinant of gross private domestic investment, using the traditional 1% level of significance, while, in the short-run, Minimum Rediscount Rate (MRR), change in MR and one-period lagged money supply affect investment significantly.
Thus, based on our belief in the efficacy of sustained private sector investment in redressing some of the economic ills facing the country today, and having regard to the results of this study, we recommend that the use of these monetary policies should be made much more effective in terms of doses and frequencies. Particularly, sharp and wide swings in policy should be avoided.
In addition, the current effort to attract foreign private investments in the various sectors of the economy should be improved upon to complement the sound monetary policies envisaged above. To this extent, policies on income repatriation, taxation, rural development, etc as they effect private foreign investments may have to be revisited.
Brainard, W. (1967), “UNcertaintly and the Effectiveness of Policy” American Economic Review, Vol. 57, pp 411-425.
Challen, D. W. and Hagger, A. J. (1993), Macro-econometric Systems- Construction, and Applications, The MacMillan Press: Hong Kong.
Chenery, H. B. (1953), “Overcapacity and the Acceleration Principle”, Econometrica, Vol 20, pp. 1-28.
Chow, G. C. (1983), Econometrics, Singapore: McGraw-Hill,
Christ, C. (1975), “Judging the Performance of Econometric Models of the U. S Economy”, International Economic Review, Vol. 16.
Engle, R. F. and Granger, C. W. J. (1987), “Contegration and Error Correction: Representation, Estimated and Testing”, Econometrica, Vol. 55, pp. 251-276.
Friedman, M. and Mieselman, D. (1964), The Relative Stability of Monetary Velocity and the Interest Multiplier in the United State, NJ: Prentice Hall.
Fredman, M. and Schwartz, A. J. (1964), “Money and Business Cycles”, Review of Economics and Statistics, Vol. 45, pp. 61.
Goodhart, C. A. E. (1982), Money, Information and Uncertainty, London: Macmillan
IMF Staff Papers (1988), World Economic Outlook.
Johansen, S. and Juselius, C.A.E. (1990), Maximum Likelihood Estimation and Inference on Cointegration- with Application to the Demand for Money, (5th Ed.) Oxford Bulletin of Economics and Statistics, Vol. 52, pp. 269-270.
Meyer, L. H. (1980), Macroeconomics-A Model Building Approach, (Ohio: South- Western Publishing Co.
Okeya, I. O. (1991), “Monetary Policy in Nigeria”, Business Concord, pp. 3 February.
Stock, J. H. and Watson, M. W. (1988), Testing for Common Trends Journal of the American Statistical Association pp.1077-1107. December.
Teriba, O. (1976), “Instruments of Monetary Control: The Nigeria Experience”, Ibadan:” Ibadan University Press.
TABLE 3: GFCF, GDP AND ME AT CURRENT PURCHASERS VALUE (N’M)