Econometrics Paper - Thailand

Only available on StudyMode
  • Topic: Regression analysis, Variance inflation factor, National accounts
  • Pages : 7 (1434 words )
  • Download(s) : 13
  • Published : December 7, 2012
Open Document
Text Preview
ECONMET C32 : Final Paper

For this final paper, I was tasked to look at the Thailand’s time series data to examine the relationship between the Current Account Balance in USD as the dependent variable, and the average exchange rate, USD to 1 Thai Baht and Gross Fixed Capital Formation in Constant 2000 USD as the independent variables.

i.) Data of the Current Account Balance in USD (Y), and the average exchange rate, USD to 1 Thai Baht (X1) and Gross Fixed Capital Formation in Constant 2000 USD. See Appendix A.

ii.) Standard Regression Function – See Appendix B.
← The standard regression function for the current account balance will be: current account balance = 33.15782 – 727.4281 (average exchange rate) – 2.49 x 10-10. (gross fixed capital formation). This is a Lin-Lin model where units for each variable are concerned. Also, current account balance is the dependent variable and average exchange rates and gross fixed capital formation are the independent variables. ← The β1, or the intercept of the regression is 33.15782. The slope coefficients of both independent variables both resulted to a negative value. Therefore, there is an indirect relationship that exists between the current account balance and average exchange rate, and the current account balance and the gross fixed capital formation. β2, or the slope coefficient of X2, is –727.4281, which means that a unit increase in the average exchange rate would decrease the current account balance by –727.4281. β3, or the slope coefficient of X3, is –2.49 x 10-10, which means that a unit increase in the gross fixed capital formation would decrease the current account balance by –2.49 x 10-10.

Tests:
← Individual Test of Significance for Average Exchange Rate: Ho: β2 = 0
H1: β2 ≠ 0
CR: 2.052 < T < -2.052
TS: [pic][pic]
Decision: T falls under the CR; therefore, the average exchange rate is individually significant.

← Individual Test of Significance for Gross Fixed Capital Formation: Ho: β3 = 0
H1: β3 ≠ 0
CR: 2.052 < T < -2.052
TS: [pic]
Decision: T falls under the CR; therefore, the gross fixed capital formation is individually significant.

← Overall Test of Significance
Ho: β1 = β2 = β3 = 0
H1: β1 ≠ β2 ≠ β3 ≠ 0
CR: F > 3.35
TS: [pic]
Decision: F falls under the CR; therefore both the independent variables together statistically influence the current account balance.

iii.) Test for Multicollinearity
← Since the coefficient of determination for the auxiliary regression is small, .0665, then there is no multicollinearity in the given model. See Appendix C. ← VIF; Variance Inflation Factor – See Appendix D.

All the Variance Inflation Factors – both independent variables and the mean VIF – are less than ten; therefore there is no multicollinearity. ← Therefore, based on the two tests for multicollinearity, multicollinearity, or the statistical correlation between two or more independent variables, does not exist in this model.

iv.) Test for Heteroscedasticity
← Via the Breusch-Pagan Test, there is no heteroscedasticity present in the model since θ < [pic]. See Appendix E. ← White’s Test: See Appendix F.
Ho: homoscedasticity
H1: heteroscedasticity
CR: W > 11.070
TS: [pic]
Decision: W statistic falls under AR; therefore, the null hypothesis should be accepted, and there exists homoscedasticity.

v.) Test for Autocorrelation
← Via the Durbin-Watson Test, the Durbin Watson Statistic is 1.236917, which is in the zone of indecision since the value in between 1 and 2, but leaning towards the negative side. See Appendix G. ← Runs Test: See Appendix H.

Ho: no autocorrelation
H1: with autocorrelation
CR: 21.2740 < N < 10.7260
TS: N = 9
Decision: N falls under the AR;...
tracking img