Question

Use the EViews Output in the appendix. (36 points) a) Determine the estimated consumption function from the EViews output. [Also provide R'. t-ratios and the F-statistic and the corresponding p-values, DW statistic] Do the results concur with the a priori expectations? 0.S300) 34 32526) Liraha 0.5854) p0.3 160 F-st-Co.Doo0 Perform ALL the relevant ADF tests! b) Is there a unit root in the data (Y & X series)? c) Do you suspect that the regression equation stated in (a) is spurious? Why? Why not? Date: 05/11/18 Time: 14:18 Series: RCONNA RGDPNA Sample: 1991 2014 Included observations: 24 Null hypothesis: Series are not cointegrated Cointegrating equation deterministics: C matic lags specification based on Schwarz criterion (maxlag-4) ependent RCONNA tau-statistic Prob. 3.274188 0.0982 MacKinnon (1996) p-values. Dependent Variable: RCONNA Method: Least Squares Date: 05/11/18 Time: 14:23 Sample: 1991 2014 Included observations: 24 Coefficient Std. Error -Statistic Prob. Variable 0.766010 0.022316 34.32526 0.0000 864.9599 1562.146 0.553700 0.5854 RGDPNA 0.981670 Mean dependent var 0.980837S.D, dependent var 3651.982 Akaike info criteriorn 2.93E+08 Schwarz criterion -229.8830 Hannan-Quinn criter. 47986.84 R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) 26381.27 19.32358 19.42175 19.34963 0.603838 1178.224 Durbin-Watson stat 0.000000 Null Hypothesis: RGDPNA has a unit root s: Constant Linear Trend Length: 0 (Automatic - based on SIC. maxiag 5) t-Statistic Prob. Augmented Dickey-Fuller test statistic Test critical values 1.0000 1.832221 1% level -4.394309 Null Hypothesis: RCONNA has a unit root Exogenous: Constant. Liea rend Lag Length: o (Automatic -based on SIC. maxlag-5) t-Statistic Prob. 0.9274 ed Dickey-Fuller test statistic 1%level -0.986053 -4.394309 Test critical values:

Answer 1

Using the given E-views output

(a) The estimated consumption function is

Y = 864.9599+0.766010X

Where Y is the dependent variable Consumption C : RCONNA

X is the independent variable RGDPNA

The slope coefficient(independent variable regressor coefficient) 0.766010 which is highly significant since it’s t ratio value is 34.32526 with p value is 0.

The intercept coefficient 864.9599 is insignificant since it’s t ratio value is 0.5537 with p value is 0.5854 which is greater than the levels of significance 0.01, 0.05 and even 0.1.

R2 = 0.98167 and Adjusted R2 = 0.980837 which nearly 98% of variation in Y occurs due to X, that is there exists a very strong relationship between Y and X.

F statistic value = 1178.224 with p value as 0 Which indicates that the above linear regression relationship between Y and X is highly significant.

Then it indicates that there is a significant influence of independent variable X on the dependent variable Y.

Durbin Watson statistic is 0.603838 which is less than 2 and indicates that there is a positive autocorrelation in the series.

(b) The null hypothesis: RGDPNA has a unit root test.

To test the above null hypothesis, the ADF (Augmented Dicky-Fuller ) test statistic has got p value 1 which is greater than the level of significance 0.01, then the null hypothesis is accepted. That is the independent variable X: RGDPNA has a unit root test.

The null hypothesis: RCONNA has a unit root test.

To test the above null hypothesis, the ADF (Augmented Dicky-Fuller ) test statistic has got p value 0.9274 which is greater than the level of significance 0.01, then the null hypothesis is accepted. That is the dependent variable Y: RCONNA has a unit root test.

Therefore both Y and X series have unit root in the data.

(c) We may suspect that the above regression equation in (a) is spurious because

The null hypothesis: Series are not cointegrated

To test the above null hypothesis with the dependent variable Y:C: RCONNA has the p value 0.0982 for the test statistic is greater than the level of significance 0.01, then we accept the null hypothesis and conclude that the Series are not cointegrated.

But we may consider the regression equation is spurious since the p value 0.0982 is just above 0.05 but below 0.1 level of significance. That is at more than or equal to 0.1 levels of significance, the above null hypothesis is rejected and we say that he Series are cointegrated. Also the independent variable X is significantly influencing the dependent variable Y as the t value of regression coefficient of X is significant and R2 value adjusted R2 value are very high and are greater than Durbin-Watson statstic which can be obseved in (a). Moreover X and Y have unit root, that is the series in non-stationery which can be observed in (b).