TheExcelfile“money.xlsx“(availablefromSurreyLearn/Coursework)providesadatasettoanalyse the demand for moneyin the US. The data represent a sample observed over the period 1950.Q1–2000.Q4. The description of the variables (all seasonally adjusted) in the file is givenbelow:M1(m)=real money supply.INF(inf)=inflation rate.REALGDP(y)=real GDP.TBILRATE(tbilrate)=treasury bills (a proxy for the interestrate).
QuestionsUse the data provided to answerALLthe questionsbelow.
Q1. Plot the time series of all the variables. Describe and discuss dynamics of the data. How does money supply dynamics compare to the other three variables? [5 Marks]
Q2. Estimate thestatic demand for money equation below:๐๐๐ก=๐ฝ0+๐ฝ1๐๐๐๐ก+๐ฝ2๐๐ฆ๐ก+๐ฝ3๐ก๐๐๐๐๐๐ก๐๐ก+๐ข๐ก(1)where:๐๐๐กisnatural log of real money supply, ๐๐ฆ๐กis the natural logarithm of real gdp, ๐ข๐กis the error termand tis a subscript referring to time.
a)Estimate model(1)above using OLS. Report the estimation output.[4marks]
b)Carefully interpret the estimated coefficients including the intercept.Are the signs of the coefficients consistent with what you expect? Explain your reasoning.[4 marks]
c)Perform tests for the statistical significance of the coefficientsof the independent variables inflation, real GDP and interest rateusing critical values corresponding tothet–distribution and the test p–values. Carefully state the null and alternative hypotheses. Interpret your results.[4marks][12Marks]
Q3.a)Perform a joint significance test for the independent variables of the modelin (1)using the critical values corresponding to the F–distribution and the test p–values.Carefully state the null and alternative hypotheses.Interpret your results.[3 marks]
3b)Comment on the goodness of fit of the modelin(1).[2 marks][5Marks]
Q4. a)Using at least two different plots,provide a graphical analysis of the residualsinmodel(1)to detect the presence of autocorrelation. Do you find evidence of autocorrelation? Explain your reasoning.[3marks]
b)What are the consequences of autocorrelation on the OLS estimator?[3 marks]c)Test for autocorrelation in the residuals using an appropriate procedure.[4 marks][10Marks]
Q5. Suppose money supply in equation (1) is not directly observed and we have the long run demand for money. The desired level is defined as follows:๐๐๐๐กโ๐๐๐๐กโ1=๐ฟ(๐๐๐๐กโโ๐๐๐๐กโ1)
(2)a)How does the model in equation (1) change?[3marks]
b)State and estimate the new specification. Interpret the estimation output. [2marks]c)Based on part
(b) above, what is the speed of adjustment to the long run level?[2marks]d)Using the new specificationin (b), derive the long run static model. How does it compare to the estimated specification based on equation (1).[3marks][10Marks]
Q6. Test for autocorrelation in the residuals (from the regression in Q5) using an appropriate procedure.[5Marks]
Q7. The presence of real GDP, inflation, and lagged dependent variable in the specification in Q5 may lead to an endogeneity issue.a)Explain and discuss what is the issue of endogeneity in the context of this question.[3marks]b)What are the implications of the presence endogeneity on OLS?[2marks]c)How would you accountfor the presence of endogeneity?[3marks][8Marks]
Q8. Estimate the model in Q5 using Two Stage Least Squares (TSLS). Answer the following questions:a)Define the set of instrumentsused to estimate the model. Justify the reasonsof your selected set of instruments.[4marks]b)Compare the qualitative and statistical interpretation of the TSLS estimated
4model to that estimated using OLSin Q5. Which model is more consistent and conform toeconomic theory?[4marks]c)Using this specification and TSLS estimates, derive the long run static model. How does it compare to (i) the estimated specification based on equation (1)in Q2 and (ii) to that derived based on (Q5–d).[4marks][12Marks]Q9. The model estimated in equation(1)may be spuriousdue to the presence of unit roots. a)Perform the ADF unit root test on all variablesin equation (1). State clearly the hypothesis being tested, the data generating process, the lag length selection criteriaand the critical values. Isthere evidence that the data containsa unit root? [6marks]b)What is the order of integration of the variables tested in part (a)above?.[4marks]c)Perform the Engle–Granger cointegration test. Are the variables cointegrated? State clearly hypothesis being tested, the data generating process,the lag lengthselection criteriaand the critical values.[5marks][15Marks]
Q10. Discuss how your findings can be useful to policy makers. What policy implications can be drawn from this analysis?
