JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
DOES COMPLEMENTING EXTERNAL DEBT WITH POLICY VARIABLES ENLARGE THE BORROWING SPACE? INSIGHT FROM NIGERIA MACRO DATA CALLISTAR KIDOCHUKWU OBI, Ph.D Department of Economics Delta State University, Abraka, Nigeria
[email protected] or
[email protected] INNOCENT ABANUM IFELUNINI, Ph.D Department of Economics, University of Nigeria, Nsukka, Nigeria
[email protected] RICHARD KOJO EDEME, Ph.D Department of Economics, University of Nigeria, Nsukka, Nigeria
[email protected] Abstract In this study, we empirically examined the effect of some complementary intervention variables in expanding the limit of external debt. In doing so, we proceeded by establishing the critical levels of external debt leverage for Nigeria’s sustainable economic growth within the periods 1965 to 2013. This was followed by the introduction of some policy variables. The complementary intervention variables include public investment, productivity of capital and labour, and private savings rate. Two variant optimization algorithms namely GaussNewton/BHHH, and Marquardt algorithms, respectively, for three variant estimatorsGeneralized Linear Method (GLM), Least Square (LS) and Maximum Linear-Autoregressive Conditional Heteroscedasticity (ML-ARCH) were used. The estimated result showed that at the initial period debt acquisition reduces economic growth, but at an increasing rate up to a turning point of 12.5% after which growth is infinitely unhindered by debt acquisition. We found that with guided policy intervention, there is no maximum limit to debt acquisition. Based on these findings, we concluded that debt accumulation is not detrimental to the economy; hence there is no maximum limit to debt acquisition, in as much as debt acquisition goes with complementary policy intervention. Key words: External Debt, policy variable, borrowing space JEL Classification: H63, H60
166
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
1. INTRODUCTION There is an increasing awareness that foreign debt hinders or improves the economic growth of most developing countries. Low and middle income countries contract foreign loans to bridge resource gaps. In this regard debt is seen to be inevitable (Obi, 2015). But over time, it has been observed that resources are diverted from investment and other productive uses into servicing the debt and this is not good for a growing economy. Therefore, it becomes pertinent to ascertain the particular threshold or tuning points where debt begins to have a negative or positive impact on growth, and also ascertain that level of external debt that is growth sustaining if resources are not diverted from investment into servicing, but debt accumulation is complemented with policy instruments or variables. This justifies this study as it aims at ascertaining the critical levels of external debt leverage that is consistent with Nigeria’s sustainable growth when complemented with policy instruments or variables. It covers a period of 47 years (1965-2012), the early years of borrowing in Nigeria, through periods of high debt accumulation, to debt relief period and the gradual building up of external debt after relief.
2. EMPIRICAL REVIEW ON EXTERNAL DEBT External debt with growth relationship has been considered by several researchers. Patillo, Poirson and Ricci (2002, 2003, and 2011) noted that for a nonlinear debt on growth; debt affects growth positively at low levels of accumulation up to a point where additional debt begins to impact negatively on growth. It was also observed that the policy environment affects the debt/growth relationship. This view was also shared by Obi (2015) in a study on critical levels of debt and sustainable growth. It was observed that at the early stage of debt acquisition, output growth rises at a decreasing rate until debt begins to depress growth at a given threshold. Other studies on non-linear debt-growth relationship include studies done by Maghyereh, Omet and Kalaji (2002); Cordella, Ricci, and Ruiz-Arranz (2010); Presbitero (2008); Espinoza (2014); among others.
3. METHODOLOGICAL FRAMEWORK The study adopted ex-post facto design. Quadratic hill-climbing technique was used in model building. Capital and labour productivity, public investment and savings were incorporated into the model. 3.1. THE MODEL VOLUME 8 NUMBER 2 JULY 2016
167
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
The general form of the model is given as: are coefficients.* 𝑓(𝑑𝑥) = 𝛽0 𝑑𝑥 2 + 𝛼0 𝑑𝑥 + 𝜆0
(1)
From case (1), the critical value of debt acquisition for Nigeria (dx) occurs at the point where: 𝛼
𝑑𝑥 = 𝑓(− 2𝛽0 )
(2)
0
The primary objective of estimating equation (2) is to determine if Nigeria’s external debt obligation has a critical point and at what point it occurs. That is, to establish if the country’s debt obligation has a minimum or a maximum level when policy control variables are introduced since either of these outcomes has varying implications for the economy. For example, if external debt stock has a maximum, it means that growth in debt stock will continue to augment growth up to a point where it becomes a constraint, while the minimum level will mean that debt at the early stage it retards growth up to the point where it infinitely augments growth. Next we estimate a function of debt-growth relationship by incorporating some policy intervention variables in the traditional growth model such as: (1) (2) (3) (4)
Public investment, captured with public sector capital expenditure; Productivity of capital, captured with output-capital ratio; Labour productivity, captured with output-labour ratio; and Gross private domestic savings rate, captured with total private savings-output ratio.
𝜕𝑙𝑛(𝑔) = [𝑎2 (𝜕𝑙𝑛(𝑑𝑥)) + 𝛽2 (𝜕𝑙𝑛(𝑑𝑥 2 ))] + 𝜃(𝑔𝑝𝑜) + 𝜆2 + µ2
(3)
Where (𝑔𝑝𝑜) is a vector of debt-growth policy control variables (capital expenditure, capital productivity, labour productivity, and savings rate); (𝜃) is a vector of policy parameters; (𝑎2 ) 𝑎𝑛𝑑 (𝛽2 ) are debt and non-linear debt parameters; while (µ2 ) is the random error term. 3.2. DEBT-GROWTH CONTROL VARIABLES The choice of our control variables to model the critical debt levels in Nigeria is informed by the theoretical underpinning which says that in order to achieve an optimal steady state level of individual utility, public debt should be zero when there is only public investment and even negative when there is also some preference for 168
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
public consumption; secondly because public consumption reduces total investment in the economy, it should be accompanied by compensating public investment at the capital market. Consequently, three control variables namely public investment rate defined as the ratio of public sector capital expenditure to GDP; gross fixed capital formation, a proxy for capital; and private savings rate, measured as the private domestic saving ratio to GDP; inter- temporal optimization of consumption, measured as the lead of private consumption expenditure or lag of private savings rate; and labour productivity, measured as the output per labour force. 3.3. ESTIMATION TECHNIQUE The estimation applied Gauss, Berndt, Hall, Hall and Hausman optimization algorithms Generalized Linear Model (GLM) which was cross-matched with Least square and ML-ARCH estimators was used 𝜕𝑙𝑛(𝑔) = [𝛼0 (𝜕𝑙𝑛(𝑑𝑥)) + 𝛽0 (𝜕𝑙𝑛(𝑑𝑥 2 ))] + 𝜆0 + µ0
(4)
3.4. DATA HANDLING The study spans a period of 48 years (1965-2013). Data was collected from the Central Bank of Nigeria (CBN) online database; National Bureau of Statistics (NBS).
4. PRESENTATION OF RESEARCH FINDINGS The result of the critical levels of debt-growth model (with three variant estimators); with the inclusion of four policy control variables is shown in table 4.01 below. The control variables are used to examine how some selected macroeconomic variables will help to reduce or reverse the negative effects of external borrowing since debt is almost inevitable. Three variants of estimation techniques namely Generalized Linear Method (GLM); Least Square (LS); and Maximized Linear-Autoregressive Conditional Hetrokesdaticity are used. Table 4.0 1: Summary of growth-augmenting and growth-constraint of external debt accumulation Eq Name: Method: VOLUME 8 NUMBER 2 JULY 2016
EQ01 GLM
EQ02 LS
EQ03 ML-ARCH 169
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
∆(LOG(G))
∆(LOG(G))
∆(LOG(G))
∆(LOG(DX))
-0.248695 (0.0952)** [-2.6128]**
-0.248695 (0.0952)* [-2.6128]*
-0.274690 (0.1176)* [-2.3364]*
∆(LOG(DX)^2)
0.016936 (0.0072)* [2.3430]*
0.016936 (0.0072)* [2.3430]*
0.018047 (0.0083)* [2.1625]*
∆(LOG(GKE))
0.077555 (0.0347)* [2.2330]*
0.077555 (0.0347)* [2.2330]*
0.092518 (0.0054)** [17.0450]**
∆LOG(SVNG(-3))
0.169216 (0.0527)** [3.2097]**
0.169216 (0.0527)** [3.2097]**
0.209223 (0.0049)** [42.3230]**
∆LOG(RCA)
0.073544 (0.0237)** [3.0973]**
0.073544 (0.0237)** [3.0973]**
0.061669 (0.0208)** [2.9666]**
∆LOG(LABP(-3))
0.105161 (0.0478)* [2.2011]*
0.105161 (0.0478)* [2.2011]*
0.098258 (0.0394)* [2.4932]*
Dep. Var: C
4.2 DISCUSSION OF FINDINGS The research objectives are evaluated and analysed from table 4.0.1. In table 4.0.2 we demonstrated that for debt to be growth-augmenting or acts as a binding constraint dependent to a very large extent on several policy interventions and what government does with the acquired debt, and the interaction between some macroeconomic fundamentals and debt. In this section, we consider the effects of policy control variables in facilitating or reversing the adverse effects of external debt accumulation on growth. These macroeconomic variables are carefully chosen from the determinants of aggregate 170
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
demand. Theoretically, debt accumulation facilitates immediate consumption and retard inter-generational consumption. If we assume that the current debt acquisition facilitated increase in both public and private spending, savings, and labour productivity as such, we included the following control variables, public investment in infrastructure, capital accumulation, labour productivity, and savings rate. This is done by controlling for the mentioned macroeconomic variables in the original debt-growth equations in case (1). The result is in table 4.0.1 Generally, the influence of these variables on debt-growth relationship is informative and interesting. First, the inclusion of these variables reverses the a priori signs between the linear and non-liner debt coefficients. The coefficient of the linear debt changed from positive to negative, while the non-linear alternated from negative to positive. This means that debt accumulation at the initial period reduces economic growth, but at an increasing rate. Secondly, the implication of having the parameter of non-linear debt (dx2) to be positive is that there is no maximum limit to debt acquisition as long as there are policy variables to control it as a catalyst to improve growth. Thirdly, the debt-growth relationship will have a “U-shaped” curve rather than the “inverted-U” where debt accumulation has maximum limit – as such, growth is infinitely unhindered. Fourthly, the economy will have a larger borrowing space when it is well- managed. Fifth, given countries capacity to borrow and efficiently manage debt to boost domestic output, the opportunity cost of non-acquisition of debt (the attainable infinite growth) is infinitely high. Finally, the capping is that the economy will possibly not get to a steady state growth since output growth is continuous. Corollary to it is that production function will be continuous. See figure 4.0.1 and table 4.0.2. 4.2.1
CRITICAL LEVELS INTERVENTION
OF
EXTERNAL
DEBT
WITH
Table 4.0.1 shows that given various interventions like increase in public and private investment; private savings; and labour productivity, early acquisition of debt decreases economic growth, but at an increasing rate up to a point where growth rate of external debt stock reaches about 6.3% as shown in case (5). 𝛼
−0.25
(β) = 0.02 > 0; 𝑚𝑖𝑛𝑖𝑚𝑢𝑚 𝑑𝑒𝑏𝑡 = − (2𝛽) = − (2∗(0.02)) = (+6.3%) 𝑓(6.3) = 0.02. (6.3)2 − 0.25. (6.3) = −0.8%
(5)
(6)
In case (6), we demonstrated that the minimum attainable growth rate when external debt stock composition of GDP reaches 6.3% is -0.8%. Cases (5) and (6) are
VOLUME 8 NUMBER 2 JULY 2016
171
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
at their minimum since the parameter of the non-linear debt (dx^2) is greater than zero (positive). See equation (5).
Debt-growth augmented policy 50,0
Attainable output growth rate
45,0
29,3; 45,0
40,0 35,0 30,0 25,0 20,0 15,0 10,0 5,0 0,0
-5,0
0,0
5,0
10,0
15,0
20,0
25,0
30,0
35,0
Critical and optimal levels of of external debt stock %
Figure 4. 0 1: Debt-augmented growth with policy control parameters Source: Author based on regression estimates
4.2.2 OPTIMAL EXTERNAL DEBT WITH INTERVENTION Figure 4.0.1 shows that the minimum debt-GDP composition requirement for optimal growth is about 12.5% as demonstrated in case (7). At that point economic growth is zero. Furthermore, beyond that point debt acquisition has an infinite positive impact on economic growth, since maximum growth rate is not attainable. However, the economy is not likely to get to a steady state since any addition or decline in debt acquisition will decline or increase growth to infinite level because output growth has a fairly continuous curve. 𝑓(12.5) = 0.02. (12.5)2 + 0.25. (12.5) = 0
(7)
4.2.3 PUBLIC INVESTMENT AND GROWTH-AUGMENTING DEBT (GKE)
172
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
In the face of rising debt, public investment (GKE) is shown to have a positive impact on growth. Public spending on infrastructure as shown in our result facilitates the necessary impetus required for the initial condition of debt acquisition to be growth-augmenting. Thus, a well-applied and managed debt is a necessary and sufficient condition to determine the impact of debt on growth. The foregoing is evident from the fact that in table 4.0.1 there is a positive non-linear relationship between external debt and linear government capital expenditure and real output growth rate. 4.2.4 GROSS PRIVATE SAVINGS RATE (SVNG) We defined savings rate (SVNG) as the ratio of gross domestic private savings to gross domestic output (GDP). Theoretically, past savings constitute current consumption, hence the three-period lag between growth and savings shown in table 4.0.1. Thus, if past savings are channelled into current investment in the economy, the growth benefit is that accumulation of debt will not push the economy to any steady state or growth limit. Ceteris paribus, increase in savings by 10% pushes the economy to a higher growth by about 1.2% (taking the average of the parameters of the three estimators GLM, LS, and ML-ARCH). 4.2.5 ACCUMULATION OF MATERIAL CAPITAL (RCA) Capital (RCA) on the average facilitates at least about 0.6% of the 10% increase in domestic output as shown in table 4.0.1. This contribution seemingly looks low, but given that its transmission to growth is instantaneous (contemporaneous value of capital) it will not be out of place to say it has a stronger policy effect, in terms of time, on growth than savings with three period lags. 4.2.6 LABOUR PRODUCTIVITY (LABP) This is derived as the ratio of gross domestic output to Nigeria’s active labour force or economically active population. It is also shown in table 4.0.1 that productivity facilitates output growth thus, reducing the effects of debt accumulation on growth potentials of the economy, though with lags. It is evident that labour productivity facilitates average of about 1% of the total 10% increase in domestic output in the face of debt acquisition. This contribution is also low given that it takes about three periods for such impact to manifest in the economy. Public and private investment will be said to constitute a quicker policy response in driving the economy given the short transmission between them and the VOLUME 8 NUMBER 2 JULY 2016
173
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
economy. The implication is that for debt acquisition to generate the necessary impetus of driving the economy, borrowed fund will need to be channelled into infrastructural development.
Critical debt and Economic Growth Deficit 5,0 0,0 0,0
10,0
20,0
30,0
40,0
50,0
Growth deficit%
-5,0 -10,0 -15,0 -20,0 -25,0 -30,0
Critical and optimal levels of of external debt stock %
Figure 4. 0 2: Critical debt and economic growth deficit Source: Author based on regression estimation and partial equilibrium simulation
4.3 CRITICAL DEBT AND ECONOMIC GROWTH DEFICIT Figure 4.0.2 shows that the economic growth deficit associated with debt acquisition increases at an increasing rate and reaches maximum where additional acquisition leads to negative growth deficit. The implication of this is that there is a huge gap between the attainable growth rate associated with debt acquisition and the attained growth resulting from the minimum debt acquisition. That is, there is enough borrowing space, given a policy response that is capable of pushing or extending the boundary of optimal growth as debt rises. This is also shown in table 4.0.3
174
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS Table 4.0 2: Critical and optimal debt level and attainable growth rate with intervention. External debtGDP ratio %
GDP Growth rate
Growth deficit
External debtGDP ratio %
GDP Growth rate
Growth deficit
External debtGDP ratio %
GDP Growth rate
Growth deficit
0.5
-0.1
2.0
15.5
0.9
1.0
30.5
11.0
-9.1
1.0
-0.2
2.1
16.0
1.1
0.8
31.0
11.5
-9.6
1.5
-0.3
2.2
16.5
1.3
0.6
31.5
12.0
-10.1
2.0
-0.4
2.3
17.0
1.5
0.4
32.0
12.5
-10.6
2.5
-0.5
2.4
17.5
1.8
0.2
32.5
13.0
-11.1
3.0
-0.6
2.5
18.0
2.0
-0.1
33.0
13.5
-11.6
3.5
-0.6
2.5
18.5
2.2
-0.3
33.5
14.1
-12.2
4.0
-0.7
2.6
19.0
2.5
-0.6
34.0
14.6
-12.7
4.5
-0.7
2.6
19.5
2.7
-0.8
34.5
15.2
-13.3
5.0
-0.8
2.7
20.0
3.0
-1.1
35.0
15.8
-13.9
5.5
-0.8
2.7
20.5
3.3
-1.4
35.5
16.3
-14.4
6.0
-0.8
2.7
21.0
3.6
-1.7
36.0
16.9
-15.0
6.5
-0.8
2.7
21.5
3.9
-2.0
36.5
17.5
-15.6
7.0
-0.8
2.7
22.0
4.2
-2.3
37.0
18.1
-16.2
7.5
-0.8
2.7
22.5
4.5
-2.6
37.5
18.8
-16.9
8.0
-0.7
2.6
23.0
4.8
-2.9
38.0
19.4
-17.5
8.5
-0.7
2.6
23.5
5.2
-3.3
38.5
20.0
-18.1
9.0
-0.6
2.5
24.0
5.5
-3.6
39.0
20.7
-18.8
9.5
-0.6
2.5
24.5
5.9
-4.0
39.5
21.3
-19.4
10.0
-0.5
2.4
25.0
6.3
-4.4
40.0
22.0
-20.1
10.5
-0.4
2.3
25.5
6.6
-4.7
40.5
22.7
-20.8
11.0
-0.3
2.2
26.0
7.0
-5.1
41.0
23.4
-21.5
11.5
-0.2
2.1
26.5
7.4
-5.5
41.5
24.1
-22.2
12.0
-0.1
2.0
27.0
7.8
-5.9
42.0
24.8
-22.9
12.5
0.0
1.9
27.5
8.3
-6.4
42.5
25.5
-23.6
13.0
0.1
1.8
28.0
8.7
-6.8
43.0
26.2
-24.3
13.5
0.3
1.6
28.5
9.1
-7.2
43.5
27.0
-25.1
14.0
0.4
1.5
29.0
9.6
-7.7
44.0
27.7
-25.8
14.5
0.6
1.3
29.5
10.0
-8.1
44.5
28.5
-26.6
VOLUME 8 NUMBER 2 JULY 2016
175
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
15.0 0.8 1.2 30.0 10.5 Source: Author based on regression estimates
-8.6
45.0
29.3
-27.4
5. Summary and Conclusion The study introduce some policy control variables such as public investment, productivity of capital and labour, and private savings rate in the original quadratic model (since theoretically, debt accumulation facilitates immediate consumption and retard inter-generational consumption) to see how the conclusion reached when there was no policy will be affected. This is to ascertain the effects of these policy control variables in facilitating or reversing the adverse effects of external debt acquisition on growth. The estimated result shows that at the initial period debt acquisition reduces economic growth, but at an increasing rate up to a turning point (optimal level of 12.5%) after which growth is infinitely unhindered by debt acquisition. That is, with guided policy intervention, there is no maximum limit to debt acquisition. Based on these findings the study therefore, concluded that debt accumulation is not detrimental to the economy, hence there is no maximum limit to debt acquisition, with guided policy intervention. 5.1 POLICY IMPLICATION Based on the research findings above, the following policy recommendations are made. (1) The country’s external borrowing space is capable of generating higher output growth, particularly when there is a complementary policy. Consequently, there is no binding constraint to external borrowing as long as there are complementary policy interventions. (2) Debt acquisition is potentially economic growth steady state phenomenon. Thus, various policy interventions could potentially delay countries economic steady state generated by debt accumulation. (3) Increasing country’s borrowing space is sacrosanct with country’s complementary policy instrument, which also addresses enabling macroeconomic environment.
176
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
REFERENCES Ayadi, F.S and Ayadi, F.O (2008). The Impact of External Debt on Economic Growth: A Comparative Study of Nigeria and South Africa, Journal of Sustainable Development in Africa, 10(3), 234-264. Chenery, H.B. and Strout, A.M. (1966), Foreign Assistance and Economic Development. American Economic Review, 56(4), 679-733. Clements, B., Bhattacharya, R., and Nguyen (2003). External Debt Public Involvement and Growth in Low-Income Countries. IMF Working Paper No. WP/03/249 Cordella, T. Ricci, L., and Ruiz-Arranz, M. (2010). Debt Overhang or Debt Irrelevance? Revisiting the Debt Growth Link. IMF Staff Paper. 57(1). 1-24. Eaton, J. (1993). Sovereign Debt: A Premier. The World Bank Economic Review, 7 (2): 137-72. Espinoza, R. (2014). A Model of External Debt and International Reserves. Oxford Centre for the Analysis of Resource-Rich Countries. www.uclouvain.be Imbs, J. and Ranciere, R (2005). The overhang hangover. The World Bank Policy Research Working Paper Series 3673, the World Bank. Keynes, J. M. (1936). The general theory of employment, interest, and money: The classic work of modern-day economics. Rendered into HTML on Wednesday April 16 09:46:33 CST 2003, by Steve Thomas for the University of Adelaide Library. Lucas, R. E. (1976). Econometric Policy Evaluation: A Critique. Carnegie-Rochester Conference Series on Public Policy. 1, 19-46. Maglyereh,A., Omet, G., and Kalaji, F. (2002). External Debt and Economic Growth in Jordan: The Threshold Effect.Content.csbs.utah.edu Obi, C.K (2015). External Debt and Economic Growth in Nigeria. The Nexus. Ilorin; Ilorin Journal of Management Sciences. 2(1), 11-21 Obi, C.K (2014). An Estimation of the Critical Levels of External Debt Accumulation the Nigerian Experience . Canada: OIDA International Journal of Sustainable Development. 07(12), 73-82 Pattilo, , Poirson, H. and Ricci, L, (2002). External Debt and Growth. IMF Working Paper 02/69. Washington: IMF. Pattilo, C., Poirson, H., and Ricci, L. (2003). Through What Channel Does External Debt Affect Growth. Brookings Trade Forum. Pattilo, C., Poirson, H., and Ricci, L.A. (2011). External Debt and Growth. Review of Economies and Institutions. 2(3), 1-30
VOLUME 8 NUMBER 2 JULY 2016
177
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
Presbitero, A.F. (2008). The Debt-Growth Nexus in Poor Countries: A Reassessment Economics. The Open- Access, Open- Assessment E- Journal, 2(30). 1-30 Reinhart C.; Rogoff, K. and Savastano, M. (2003). Debt Intolerance. Brookings paper in Economic Activity. 1, 1-74. Tasos, S. (2014). Direct and Economic Growth: Is There Any Causal Effect? An Empirical Analysis with Structural Breaks and Granger Causality for Greece. Theoretical and Applied Economics. 21(1), 51-62.
178
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
APPENDIX EQ01 Variable
Dep. Var: Coefficient
D(LOG(G)) Std. Error
t-Statistic
Prob.
D(LOG(DX)) D(LOG(DX)^2) DLOG(GKE) DLOG(SVNG(-3)) DLOG(RCA) DLOG(LABP(-3))
-0.248695 0.016936 0.077555 0.169216 0.073544 0.105161
0.095183 0.007228 0.034730 0.052719 0.023744 0.047777
-2.612820 2.343047 2.233048 3.209750 3.097338 2.201092
0.0090 0.0191 0.0255 0.0013 0.0020 0.0277
EQ02 Variable
Dep. Var: Coefficient
D(LOG(G)) Std. Error
t-Statistic
Prob.
D(LOG(DX)) D(LOG(DX)^2) DLOG(GKE) DLOG(SVNG(-3)) DLOG(RCA) DLOG(LABP(-3))
-0.248695 0.016936 0.077555 0.169216 0.073544 0.105161
0.095183 0.007228 0.034730 0.052719 0.023744 0.047777
-2.612820 2.343047 2.233048 3.209750 3.097338 2.201092
0.0159 0.0286 0.0360 0.0040 0.0053 0.0385
EQ03 Variable
Dep. Var: Coefficient
D(LOG(G)) Std. Error
t-Statistic
Prob.
D(LOG(DX)) D(LOG(DX)^2) DLOG(GKE) DLOG(SVNG(-3)) DLOG(RCA) DLOG(LABP(-3))
-0.274690 0.018047 0.092518 0.209223 0.061669 0.098258
0.117568 0.008345 0.005428 0.004943 0.020788 0.039410
-2.336433 2.162483 17.044999 42.322952 2.966613 2.493242
0.0195 0.0306 0.0000 0.0000 0.0030 0.0127
Dependent Variable: D(LOG(G)) Method: Generalized Linear Model (Quadratic Hill Climbing) Sample (adjusted): 1962 2013 Included observations: 49 after adjustments Family: Normal Link: Identity Dispersion computed using Pearson Chi-Square Convergence achieved after 1 iteration Coefficient covariance computed using observed Hessian Variable
VOLUME 8 NUMBER 2 JULY 2016
Coefficient
Std. Error
z-Statistic
Prob.
179
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS
D(LOG(DX)) D(LOG(DX)^2) DLOG(GKE) DLOG(SVNG(-3)) DLOG(RCA) DLOG(LABP(-3))
-0.248695 0.016936 0.077555 0.169216 0.073544 0.105161
Mean dependent var Sum squared resid Akaike info criterion Hannan-Quinn criter. Deviance statistic Pearson statistic
0.046260 0.069084 -2.711309 -2.624037 0.003140 0.003140
0.095183 0.007228 0.034730 0.052719 0.023744 0.047777
-2.612820 2.343047 2.233048 3.209750 3.097338 2.201092
S.D. dependent var Log likelihood Schwarz criterion Deviance Pearson SSR Dispersion
0.0090 0.0191 0.0255 0.0013 0.0020 0.0277 0.069385 43.95832 -2.425836 0.069084 0.069084 0.003140
Dependent Variable: D(LOG(G)) Method: Least Squares Sample (adjusted): 1962 2013 Included observations: 49 after adjustments Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(LOG(DX)) D(LOG(DX)^2) DLOG(GKE) DLOG(SVNG(-3)) DLOG(RCA) DLOG(LABP(-3))
-0.248695 0.016936 0.077555 0.169216 0.073544 0.105161
0.095183 0.007228 0.034730 0.052719 0.023744 0.047777
-2.612820 2.343047 2.233048 3.209750 3.097338 2.201092
0.0159 0.0286 0.0360 0.0040 0.0053 0.0385
R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat
0.468517 0.347725 0.056037 0.069084 44.33459 1.602691
Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.
0.046260 0.069385 -2.738185 -2.452712 -2.650913
Dependent Variable: D(LOG(G)) Method: ML - ARCH (Marquardt) - Normal distribution Sample (adjusted): 1962 2013 Included observations: 49 after adjustments Convergence achieved after 25 iterations Presample variance: backcast (parameter = 0.7)
180
VOLUME 8 NUMBER 2 JULY 2016
JOURNAL OF ACADEMIC RESEARCH IN ECONOMICS GARCH = C(7)*RESID(-1)^2 + (1 - C(7))*GARCH(-1) Variable
Coefficient
Std. Error
z-Statistic
Prob.
D(LOG(DX)) D(LOG(DX)^2) DLOG(GKE) DLOG(SVNG(-3)) DLOG(RCA) DLOG(LABP(-3))
-0.274690 0.018047 0.092518 0.209223 0.061669 0.098258
0.117568 0.008345 0.005428 0.004943 0.020788 0.039410
-2.336433 2.162483 17.04500 42.32295 2.966613 2.493242
0.0195 0.0306 0.0000 0.0000 0.0030 0.0127
-7.291140 40.81673
0.0000 0.0000
Variance Equation RESID(-1)^2 GARCH(-1) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat
VOLUME 8 NUMBER 2 JULY 2016
-0.217480 1.217480 0.434133 0.305527 0.057822 0.073554 49.30553 1.609783
0.029828 0.029828
Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.
0.046260 0.069385 -3.021824 -2.688773 -2.920007
181
