*Effects of Some Monetary Variables on Fixed Investment in Selected Sub-Saharan African... DOI: http://dx.doi.org/10.5772/intechopen.93656*

**Table 5** shows the summary of panel ARDL long-run and short-run results. As depicted in **Table 5**, lending rates, money supply and exchange rates all have a strong long-run relationship significant at 1% level with investment. Lending rates, as economic theory suggests, have been found to have a negative relationship with investment in this study [25, 44, 51]. The results are found to be in line with those of Malawi and Bader [44] and Ashraf et al. [50] where an increase in the real interest rate by 1% reduces the investment. It has been found that interest rate plays an important role in investment decision making.

It turns out that the money supply is positively related to investment for our selected panel (**Table 5**). According to the results, when the money supply is increased, a relative increase in investment follows. Many scholars established that money supply has a positive long-run relationship with investment [38, 42, 62]. On the contrary, it has been discovered that there may exist a negative relationship between money supply and investment [31, 40, 43]. Li and Yang [40] further add that money supply is a weak instrument to be used to influence real estate investment in an inflation targeting environment.

The exchange rate also shows a significant and positive long-run relationship with investment in **Table 5**. It has been argued in the literature review section that a country's investment level can benefit from the exchange rate, provided exchange rate is stable [25, 30, 34, 36]. The argument is based on the fact that a depreciating exchange rate is associated with a stable environment and strong market power [36].


#### **Table 4.**

*Linear and Non-Linear Financial Econometrics - Theory and Practice*

robust estimation technique like ARDL.

tion, respectively.

In **Table 1** gross fixed capital formation (GFCF) and money supply (MS) are generally shown to be integrated at levels I(0), while exchange rates (ER) and lending rates (LR) are integrated of order one I(1). Therefore, the variables used in the study are a mixture of I(0) and I(1) and none of them is I(2) which paves a way to run the panel ARDL [52, 60]. It is stated in Nkoro and Uko [60] that variables that show different orders of integration can be estimated best with ARDL. Moreover, cointegration results indicate the existence of a long-run relationship but do not give estimates, hence in addition to the cointegration analysis, there is a need for a

**Tables 2**–**4** provide results of panel cointegration tests as estimated for the model specified in Eq. 1 under the Pedroni, Kao and Fisher-ADF tests for cointegra-

**Panel T-statistics P-value** v-Statistic 1.316356\* 0.0940 rho-Statistic 0.863098 0.8060 PP-Statistic −0.312544 0.3773 ADF Statistic −0.132706 0.4472 **Group T-statistics P-value** rho-Statistic 0.350217 0.6369 PP-Statistic −2.365533\*\*\* 0.0090 ADF-Statistic −2.938605\*\*\* 0.0006

 *and \*\*\* indicate that the p-values are significant at 10 and 1% level of significance, respectively.*

**Variable T-statistics P-value** ADF −2.77887\*\*\* 0.0027

The Pedroni test results presented in **Table 2** confirm cointegration in three out of seven statistics. One out of four within dimensions accept the alternative hypothesis of cointegration at 10% significance levels (Panel v-Statistics) whereas two out of three between dimensions accept the alternative hypothesis of cointegration at 1% significance level (Group PP- statistics and Group ADF statistics). The Kao panel cointegration tests results, as shown in **Table 3** also confirm cointegration by rejecting a null hypothesis of no cointegration at 1% level of significance. **Table 4** illustrates a strong cointegration between the variables in the Fisher-ADF test. This is displayed by both the trace and the max Eigenvalues which both detect at least two cointegrated relationships between investment and the selected independent variables. All three cointegration tests reveal that a long-run relationship exists between the variables for the selected panel. This implies that investment has a long-run relationship with the selected monetary variables in the chosen panel of five Sub-Saharan countries. **Table 5** provides estimates of the model specified in Eq. 1, where investments are regressed against monetary variables such as lending rates, money supply and exchange rate.

**242**

**Table 3.**

*\**

**Table 2.**

*Summary of Pedroni cointegration test results.*

*Summary of Kao panel cointegration test results.*

Residual variance 26.11567 HAC variance 21.89175 *\*\*\* indicates that the p-values are significant at 1% level of significance.* *Summary of Johansen-Fisher panel cointegration test results.*


#### **Table 5.**

*Summary of long-run and short-run panel ARDL estimates.*

Market power effects tend to offset the volatility nature of exchange rate, hence it can positively affect investments.

The panel ARDL results in **Table 5** confirm that lending rates are positively related to investment in the short run at a 1% level of significance. Money supply and exchange rate, on the other hand, showed no significant short-run relationship with investment (**Table 5**). Most importantly, the error correction term met the requirement of being negative and is very high at 83% and significant at 5% level. This implies that investment will be very fast to go back to equilibrium following a change in the selected monetary variables. These results are valid and reliable as mentioned in Nkoro and Uko [60] that panel ADRL has Gaussian error terms implying normal distribution, no autocorrelation and no heteroscedasticity in error terms.
