Reference no: EM132600566
Question 1
The Excel file namebond contains 102 monthly observations on AA railroad bond yields for the period January 1968 to June 1976.
a. Obtain a line plot for the data. Do railroad bond yields appear stationary, or non-stationary?
b. Use the ADF test to formally show that the series is nonstationary.
c. Find the first difference of the bond yield series and obtain a line plot for the first-differenced series.
d. Now use the ADF test to formally test for a unit root in the first-differenced data series.
e. What do you conclude about the order of integration for the bond yields data series?
Question 2
You are given quarterly U. S. data in an Excel file namedusdatafor the following data series from January 1947 to December 1988:
Output or real private GNP per capita (Output)
Real consumption expenditure per capita (Cons)
Real private investment per capita (Inv)
a. Obtain a line plot of the consumption series in levels.
b. Based on the line plot, do you judge aggregate consumption to be nonstationary in its levels?
c. Conduct the Augmented Dickey Fuller (ADF) test to formally examine the existence of a unit root in the consumption series in levels. Choose lags 1, 2, 3, and 4 for the ADF test.
d. Now obtain a line plot of the consumption series in first differences.
e. Based on the line plot, do you judge aggregate consumption to be nonstationary in its first differences?
f. Conduct the ADF test to formally examine the existence of a unit root in the consumption series in its first differences. Choose lags 1, 2, 3, and 4 for the ADF test.
g. Repeat steps a, c, d and f for the output and investment series.
h. Briefly summarize your findings in a table. In your table, clearly state whether each of the three variables is found to be stationary or not using the ADF test result.
You want to estimate a model of output, consumption, and investment. You decide to run the following regression: Outputt: β1 + β2const + β3Invt + εt and want to obtain accurate parameter estimates for β1 and β2.
i. Show the estimation results for the above regression.
j. Is the above estimation problematic? Discuss briefly.
k. If problematic, suggest an alternative estimation that would work and show the relevant estimation results.
Attachment:- usdata.rar