International Journal of Economics and Financial Issues

Figurer 3:
Trend analysis of debt conversion from 1985 to 2016.
Source: Author’s estimation using e-views 9, 2017
Figure 4:
Trend analysis of debt forgiveness from 1985 to 2016
Source: Author’s estimation using e-views 9, 2017
Figure 2:
Trend analysis of debt refinancing from 1985 to 2016
Table 1: The descriptive statistics
Statistics
TD
DRF
DFG
DCV
Mean
3226.286
855.5556
448000000 41.46389
Median
1699.660 0.000000
0.000000
0.000000
Maximum
14537.12 21000.00 9670000000 257.0000
Minimum
13.52380 0.000000
0.000000
0.000000
Standard deviation
3589.830 3562.860 1830000000 66.92180
Skewness
1.307163 5.271420
4.297117
1.788781
Kurtosis
4.261850 30.08923
20.56786
5.421277
Source: Author’s estimation, 2018, DRF: Debt refinancing, DFG: Debt forgiveness, 
DCV: Debt conversion

---

Rafindadi and Musa: An Empirical Analysis of the Impact of Public Debt Management Strategies on Nigeria’s Debt Profile
International Journal of Economics and Financial Issues | 
Vol 9 • Issue 2 • 2019
131
Table 2 contains the result on the relationship between the total 
debts, DF, DCV and DRF. The values diagonally placed on the 
matrix in a constant stream of one all through, indicating that each 
variables is perfectly correlated with itself. While some of the 
variables are positively correlated with TD, others are negatively 
correlated. Negative correlation of variables implies that as the 
value of one increases, the other decreases. The decision rule 
according to Evans (1996) as the author suggested for the absolute 
value of r. *0.00-0.19: “Very weak” *0.20-0.39: “Weak” 0.40-0.59: 
“Moderate” *0.60-0.79;” Strong” and *0.80-1.0 “very strong.” 
DF and total debt has a very weak correlation as indicated in the 
table
with value of −0.02. Also, the result shows a weak significant 
negative relationship between the total debt profile and DCV in 
Nigeria with the significant value of 0.361, which implies that 
there is decrease in total debt as DCV increases. A very weak 
negative relationship is observed between DF and total debt with 
value of 0.207. This implies that there is reduction in DF as there 
is increase in total debt.
Table
3 shows the unit root test for DCV, DF, Total debt and DRF 
using the ADF and Phillips-Perron test. The table indicated that, 
Total debt was insignificant at a level but became significant at 
first difference, this is an indication that DCV was not stationary 
at level as confirmed by the p-values of ADF and PP with 0.6615 
and 0.1227 respectively. While after first difference the variables 
were found to be stationary and were said not to contain unit root 
with ADF and PP P = 0.0031 and 0.000 respectively. DRF was 
significant at 5% and 1% level of significant at both level and first 
difference respectively. This is evident from the p-values in both 
methods (ADF and PP) of 0.013 and 0.000 respectively. Total debt 
on the other hand exhibit stationary nature at first difference but 
non-stationery at a level in both models. Lastly and surprisingly, 
DRF the unit root test for DRF was tested using Phillips Peron 
method only, this is because ADF test was unable to test the unit 
root due to singular matrix (content of the data with zero values 
from 1990 to 2016 while PP shows that DRF contains no unit 
root both level and after differencing with P = 0.000 and 0.0001 
respectively.
Since all the series were at the same order, the data set was 
appropriate for further analysis. The dependent and independent 
variables are stationary at the first difference. The results from the 
Johansen Co-integration analysis were present in Table 4 where the 
Eigen value and trace statistics examine the null hypothesis of no 
co-integration against the alternative of co-integration. Therefore, 
analysis of annual data from 1985 to 2016 appears to support 
the proposition that there exists a stable long run relationship 
among the dependent variable. The values of the trace statistic 
were greater than relevant critical values which showed that the 
existence of 2 co-integration equation (s) at 5% is statistically 
significant level.