foreign direct investment in indonesia
DESCRIPTION
This presentation was made as part of group assignment for my prerequisite subject EconometricTRANSCRIPT
FOREIGN DIRECT INVESTMENT
Prepared by Group 7 :
1. Luqman Hakim Handoko
2. Mahamat Al Mourtada
3. Rininta Nurrachmi
THE MODEL
FDI = Foreign Direct Investment inflows in Indonesia in billion $US
GDP growth = the growth rate of domestic product, measured in (%), is positively
related to FDI.
TO (Trade openness) is reflected as the ratio of the Export plus Import divided by GDP
(Nunes et al. 2006; and Sahoo, 2006). It measured in (%) and expected to be
positively related to FDI.
Exchange rate (The strength of a currency) is used as proxy for level of inflation and
the purchasing power of the investing firm. we expect a positive and significant
relationship between the currency value and FDI inflows. The currency value can be
proxied by the Real Exchange Rate. It measured in (Rupiah / USD)
THE RESULT Dependent Variable: LNFDI
Method: Least Squares
Date: 03/08/11 Time: 14:33
Sample (adjusted): 3 39
Included observations: 35 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LNEXCHANGE_RATE 1.106468 0.286553 3.861302 0.0005
TRADE_OPENESS -8.366246 2.922459 -2.862742 0.0075
GDPGROWTH 3.252764 1.873492 1.736204 0.0925
C 1.477872 1.323961 1.116250 0.2729
R-squared 0.347500 Mean dependent var 6.052299
Adjusted R-squared 0.284355 S.D. dependent var 1.112233
S.E. of regression 0.940902 Akaike info criterion 2.823256
Sum squared resid 27.44422 Schwarz criterion 3.001010
Log likelihood -45.40698 F-statistic 5.503192
Durbin-Watson stat 0.983693 Prob(F-statistic) 0.003778
LNFDI = 1.477872 + 1.106468 LNEXCHANGE_RATE + -8.366246 TRADE_OPENESS +
3.252764 GDPGROWTH
From the result above we can conclude that:
The exchange rate and GDP growth have positive relationship to foreign direct
investment (FDI), but trade openness is negatively related to FDI. The trade
openness is not in line with our expectation.
If we see the result of exchange rate, the P-value (0.0005) < critical value at 5%
(0.05), it means the H0 is rejected. So, we can conclude that the exchange rate has
significant relationship to FDI.
If we see the result of GDP growth, the P-value (0.0925) > critical value 5% (0.05), It
means the H0 is accepted. So, we can conclude that the GDP growth has no
significant relationship to FDI.
If we see the result of trade openness, the P-value (0.0075) < critical value at 5%
(0.05), It means the H0 is rejected. So, we can conclude that exchange rate has
significant relationship to FDI. But the relationship is negative.
Overall, we can see that the P-value (0.003778) < critical value at 5% (0.05), it
means all independent variables are significant to FDI
R2 is 34.75%. it means that there is 34.75% the changes of proportion in dependent
variable by the independent variables.
1. TEST OF SPECIFICATION ERROR
The table above is the result of Ramsey RESET test which the number of
fitted is 4. From the result above we can see that P-value (0.012179) >
only at critical value 1% (0.01).
It means the H0 is accepted at this (1%) level of significant. So the
model is correct at 1% level of significant. But since we choose the level
significant 5%, the P-value is no longer greater than critical value. Or we
can say that H0 is rejected at this level. So the model has specification
error.
Ramsey RESET Test:
F-statistic 3.930146 Probability 0.012179
Log likelihood ratio 16.05954 Probability 0.002940
2. MULTICOLLINEARITY TEST
A. Auxiliaries regression
When we run regression LNEXCHANGE RATE as dependent variable
and trade openness and GDPGROWTH as independent variables we
found that R-squared is 0.702527 (70.2527%). It is fewer than 80%. So
there is no multicolliniearity.
When we run regression TRADE OPENNESS as dependent variable
and LNEXCHANGE RATE and GDPGROWTH as independent
variables we found that R-squared is 0.707735 (70.7735%). It is fewer
than 80%. So is there no multicolliniearity.
When we run regression, GDPGROWTH as dependent variable and
LNEXCHANGE RATE and TRADE OPENNESS as independent
variables we found that R-squared is 0.427491 (42.7491%). It is fewer
than 80%. So is there no multicolliniearity.
Dependent Variable: LNEXCHANGE_RATE Method: Least Squares
Date: 03/08/11 Time: 14:43
Sample (adjusted): 2 41
Included observations: 40 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
TRADE_OPENESS 8.665719 0.963470 8.994279 0.0000
GDPGROWTH -4.728580 0.982011 -4.815201 0.0000
C 4.080670 0.498001 8.194102 0.0000
R-squared 0.702527 Mean dependent var 7.531483 Adjusted R-squared 0.686447 S.D. dependent var 1.218469
S.E. of regression 0.682291 Akaike info criterion 2.145319
Sum squared resid 17.22430 Schwarz criterion 2.271985
Log likelihood -39.90638 F-statistic 43.69049
Durbin-Watson stat 0.524201 Prob(F-statistic) 0.000000
Dependent Variable: TRADE_OPENESS Method: Least Squares
Date: 03/08/11 Time: 14:43
Sample (adjusted): 2 41
Included observations: 40 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LNEXCHANGE_RATE 0.079182 0.008804 8.994279 0.0000
GDPGROWTH 0.458274 0.093045 4.925316 0.0000
C -0.179344 0.074223 -2.416278 0.0207
R-squared 0.707735 Mean dependent var 0.515588 Adjusted R-squared 0.691937 S.D. dependent var 0.117506
S.E. of regression 0.065220 Akaike info criterion -2.550065
Sum squared resid 0.157385 Schwarz criterion -2.423400
Log likelihood 54.00131 F-statistic 44.79878
Durbin-Watson stat 0.814015 Prob(F-statistic) 0.000000
Dependent Variable: GDPGROWTH
Method: Least Squares
Date: 03/08/11 Time: 14:44
Sample (adjusted): 2 41
Included observations: 40 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
TRADE_OPENESS 0.864122 0.175445 4.925316 0.0000
LNEXCHANGE_RATE -0.081471 0.016919 -4.815201 0.0000
C 0.383166 0.089772 4.268198 0.0001
R-squared 0.427491 Mean dependent var 0.215102 Adjusted R-squared 0.396544 S.D. dependent var 0.115288
S.E. of regression 0.089558 Akaike info criterion -1.915818
Sum squared resid 0.296765 Schwarz criterion -1.789152
Log likelihood 41.31636 F-statistic 13.81388
Durbin-Watson stat 1.166810 Prob(F-statistic) 0.000033
LNFDI
LNEXCHANGE_RA
TE TRADE_OPENESS GDPGROWTH
LNFDI 1.000000
LNEXCHANGE_RATE 0.415810 1.000000
TRADE_OPENESS 0.088970 0.733673 1.000000
GDPGROWTH -0.126132 -0.228319 0.262156 1.000000
B. Pairwise correlation matrix test
If we see pairwise correlation matrix test, the value of correlation
between each variable shows fewer than 80%. It means there is no
correlation between variables.
3. AUTOCORRELATION TEST
Breusch-Godfrey Serial Correlation LM Test:
F-statistic 2.754895 Probability 0.080280
Obs*R-squared 5.588056 Probability 0.061174
A. Breush-Godfrey LM test Number of fitted is 2
We can see from the result above which the number of fitted is 2, F
statistic is 2.754895 and P-value is 0.080280. We can say that at level
of significant 5% the P-value (0.080280) > critical value 5%. H0 is
accepted. So we can conclude that the model has no autocorrelation.
Breusch-Godfrey Serial Correlation LM Test:
F-statistic 0.265812 Probability 0.849444
Obs*R-squared 0.969193 Probability 0.808706
Breush-Godfrey LM test Number of fitted is 3
We can see from the result above which the number of fitted is 3,
F statistic is 0.265812 and P-value is 0.849444. We can say that
at level of significant 5% the P-value (0.849444) > critical value
5%. H0 is accepted. So we can conclude that the model has no
autocorrelation.
4. HETEROCEDASTICITY
White Heteroskedasticity Test:
F-statistic 2.000348 Probability 0.082727
Obs*R-squared 14.65265 Probability 0.100930
A. WHITE TEST
When we test the model by White Heteroskedasticity Test, we find that
obs*R-squared is 14.65265 and P-value is 0.100930. In this model we
choose the level of significant at 5%. From this result we can say that the
model has no heterocedasticity because the P-Value > critical value at
5%, it means Ho is accepted.
B. PARK TEST
Dependent Variable: LOG(E*E)
Method: Least Squares
Date: 03/08/11 Time: 15:14
Sample (adjusted): 3 39
Included observations: 35 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
TRADE_OPENESS 3.589758 4.083052 0.879185 0.3857
C -4.019399 2.156766 -1.863623 0.0713
According to park test, we find that the value of t-statistic of
TRADE_OPENESS is 0.879185 and P-value is 0.3857. Since we choose
the level of significant at 5%, we can see the P-value > critical value 5%. It
means that Ho accepted. So there is no heterocedasticity.
Dependent Variable: LOG(E*E)
Method: Least Squares
Date: 03/08/11 Time: 15:16
Sample (adjusted): 3 39
Included observations: 35 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LNEXCHANGE_RATE 0.790101 0.393334 2.008728 0.0528
C -8.005511 2.939651 -2.723286 0.0102
According to park test, we find that the value of t-statistic of
LNEXCHANGE_RATE is 2.008728 and P-value is 0.0528. since we
choose the level of significant at 5%, we can see the P-value > critical value
at 5%. It means that Ho accepted. So there is no heterocedasticity.
According to park test, we find that the value of t-statistic of GDPGROWTH is
-1.723651 and P-value is 0.0941 since we choose the level of significant at 5%,
we can see the P-value (0.941) > critical value 5%. It means that Ho accepted.
So there is no heterocedasticity.
Dependent Variable: LOG(E*E)
Method: Least Squares
Date: 03/08/11 Time: 15:16
Sample (adjusted): 3 39
Included observations: 35 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
GDPGROWTH -6.449413 3.741717 -1.723651 0.0941
C -0.748224 0.935873 -0.799494 0.4297
WASALAM
THANK YOU