Only need (h)(i)

Note: In Eviews one can estimate the model >>- + B - Bo+BAH + by typing the following command in the Equation Estimation window:yes(-100-j) (a) The time series cpifr denotes the consumer price index in France i) Does epift appear to be a stationary time series? Briefly explain ) Does cpift appear to be a trend stationary time series? Briefly explain Place any supporting Eviews output in Appendix 2 (a). (4 marks) (b) Let INF_FRANCE, - 1200 - logo denote the annualized monthly change in the CPI in France, Transform the observations on cpifr into observations on INF_FRANCE. You should find that the Sample mean of IN FRANCE, 4.406853 I) Does INF FRANCE, appear to be a stationary time series! Briefly explain il) Is there any evidence of persistence in INF_FRANCE/? Briefly explain III) Are there any other notable features of INF FRANCE, Briefly explain (4 marks) Place any supporting Eviews output in Appendix 2 (b). (c) i) Estimate the following model for INF_FRANCE, INF_FRANCE, - Be + a time;+ B.INF_FRANCE,4+ Place the Eviews output in Appendix 2 (c). It is not necessary to report the estimated model in the main body of the assignment. ii) Does the individual significance of the regressors depend on whether you use t-statistics based on the conventional OLS Standard errors or HAC standard errors? Briefly explain. Place any supporting Eviews output in Appendix 2 (). iii) Test the joint significance of those lags of inflation which are individually insignificant ini). Identify the restricted and unrestricted models. State the null and alternative hypotheses, the form and distribution of the test statistic under the null, the sample value and critical value of the test statistic, your decision rule and your conclusion Place any supporting Eviews output in Appendix 2 (c) (12 marks) (d) Estimate the following model for INF_FRANCE, and report the estimated model: INF_FRANCE, - Bo + a time, + B INF_FRANCE, 1 + B12INF_FRANCE 12 +1. (4) Place the Eviews output in Appendix 2 (d) i) Is there any evidence of autocorrelation in the error term in (4). Briefly explain. il) Let Me-Pills-1 + 2W-2+ er Perform a Breusch-Godfrey test of the mall hypothesis that there is no first-order or second-order autocorrelation in the error termin (4). State the null and alterative hypotheses, the form and distribution of the test statistic under the wall the sample value and critical value of the test statistic, your decision rule and your conclusion (Use the form of the test). (8 marks) (e) Estimate the following model for INF_FRANCE, INF_FRANCE: Bo + @itime + BINF_FRANCEH4 + B12INF_FRANCE-12 + W. (5) Place the Eviews output in Appendix 2 () 1) Is there any evidence of autocorrelation in the error term in (5). Briefly explain ii) Use Eviews automated procedure to perform a Breusch-Godfrey test of the mull hypothesis that there is no first-order or second-order autocorrelation in the error term in (5) Report the p-value and your test conclusion. iii) Use Eviews automated procedure to perform a Breusch-Godfrey test of the wall hypothesis that there is no autocorrelation in the error term in (5) up to and including lag 12. Report the p-value and your test conclusion. (4 marks) (1) Estimate the following model for INF_FRANCE, INF FRANCE,- Bo + a time + B.INF_FRANCE +B/NF_FRANCE,-12 + ( - Be + Place the views output in Appendix 2 ( I) Is there any evidence of autocorrelation in the error term in (6). Briefly explain ) Use Eviews automated procedure to perform a Breusch-Godfrey test of the mull hypothesis that there is no first-order or second-order autocorrelation in the error termin (6) Report the p-value and your test conclusion iii) Use Eviews automated procedure to perform a Breusch-Godfrey test of the mull hypothesis that there is no autocorrelation in the error termin (6) up to and including lag 12. Report the p-value and your test conclusion. (4 marks) (@) Use appropriate criteria to choose between the models in (5) and (6) and justify your choice (4 marks) (h) When the value of a regression coefficient changes between one sub-sample and another, the coefficient is said to exhibit a structural break. Inspection of the line graph of the series INF_FRANCE, suggests that the behavior of the series is qualitatively different before and after 1986ml. This may be due to a structural break in one or more of the regression coefficients i) Extend the model in (6) to allow for the possibility that there is a structural break in both the intercept and the time trend in the first month of 1986 and report the extended model ii) Estimate the model specified in i) and report the estimated model. Place the Eviews output in Appendix 2 (h) iii) What is the estimated change in the intercept between the two sample periods? iv) What is the estimated change in the marginal effect of the time trend between the two sample periods? V) Is the estimated change in the marginal effect of the time trend what you expected? Briefly explain (8 marks) (h) When the value of a regression coefficient changes between one sub-sample and another, the coefficient is said to exhibit a structural break. Inspection of the line graph of the series INF_FRANCE, suggests that the behavior of the series is qualitatively different before and after 1986ml. This may be due to a structural break in one or more of the regression coefficients. i) Extend the model in (6) to allow for the possibility that there is a structural break in both the intercept and the time trend in the first month of 1986 and report the extended model. INF_FRANCE = Bo + a time, + +381 BINF_FRANCE,-i + B12INF_FRANCE-12 + up. (0 i=1 Let cpifrt-1 denote the annualized monthly % change in the CPI in France. Transform the observations on cpifr into observations on INF_FRANCE,. You should find that the sample mean of INF_FRANCE; is 4.406853. The data for Question 2 is contained in the Eviews workfile inflation.wfl in the Assignments folder. The workfile contains monthly observations on the consumer price index for France, cpifr, for the period 1981m1 to 1996m6. The workfile also contains a dummy variable, dum, and a time trend, time. Dum assumes the value 1 from 1981ml to 1985m12, and the value zero from 1986ml to 1996m6. The variable time assumes the value zero in 1981ml and increases by 1 in each subsequent time period. Note: In Eviews one can estimate the model >>- + B - Bo+BAH + by typing the following command in the Equation Estimation window:yes(-100-j) (a) The time series cpifr denotes the consumer price index in France i) Does epift appear to be a stationary time series? Briefly explain ) Does cpift appear to be a trend stationary time series? Briefly explain Place any supporting Eviews output in Appendix 2 (a). (4 marks) (b) Let INF_FRANCE, - 1200 - logo denote the annualized monthly change in the CPI in France, Transform the observations on cpifr into observations on INF_FRANCE. You should find that the Sample mean of IN FRANCE, 4.406853 I) Does INF FRANCE, appear to be a stationary time series! Briefly explain il) Is there any evidence of persistence in INF_FRANCE/? Briefly explain III) Are there any other notable features of INF FRANCE, Briefly explain (4 marks) Place any supporting Eviews output in Appendix 2 (b). (c) i) Estimate the following model for INF_FRANCE, INF_FRANCE, - Be + a time;+ B.INF_FRANCE,4+ Place the Eviews output in Appendix 2 (c). It is not necessary to report the estimated model in the main body of the assignment. ii) Does the individual significance of the regressors depend on whether you use t-statistics based on the conventional OLS Standard errors or HAC standard errors? Briefly explain. Place any supporting Eviews output in Appendix 2 (). iii) Test the joint significance of those lags of inflation which are individually insignificant ini). Identify the restricted and unrestricted models. State the null and alternative hypotheses, the form and distribution of the test statistic under the null, the sample value and critical value of the test statistic, your decision rule and your conclusion Place any supporting Eviews output in Appendix 2 (c) (12 marks) (d) Estimate the following model for INF_FRANCE, and report the estimated model: INF_FRANCE, - Bo + a time, + B INF_FRANCE, 1 + B12INF_FRANCE 12 +1. (4) Place the Eviews output in Appendix 2 (d) i) Is there any evidence of autocorrelation in the error term in (4). Briefly explain. il) Let Me-Pills-1 + 2W-2+ er Perform a Breusch-Godfrey test of the mall hypothesis that there is no first-order or second-order autocorrelation in the error termin (4). State the null and alterative hypotheses, the form and distribution of the test statistic under the wall the sample value and critical value of the test statistic, your decision rule and your conclusion (Use the form of the test). (8 marks) (e) Estimate the following model for INF_FRANCE, INF_FRANCE: Bo + @itime + BINF_FRANCEH4 + B12INF_FRANCE-12 + W. (5) Place the Eviews output in Appendix 2 () 1) Is there any evidence of autocorrelation in the error term in (5). Briefly explain ii) Use Eviews automated procedure to perform a Breusch-Godfrey test of the mull hypothesis that there is no first-order or second-order autocorrelation in the error term in (5) Report the p-value and your test conclusion. iii) Use Eviews automated procedure to perform a Breusch-Godfrey test of the wall hypothesis that there is no autocorrelation in the error term in (5) up to and including lag 12. Report the p-value and your test conclusion. (4 marks) (1) Estimate the following model for INF_FRANCE, INF FRANCE,- Bo + a time + B.INF_FRANCE +B/NF_FRANCE,-12 + ( - Be + Place the views output in Appendix 2 ( I) Is there any evidence of autocorrelation in the error term in (6). Briefly explain ) Use Eviews automated procedure to perform a Breusch-Godfrey test of the mull hypothesis that there is no first-order or second-order autocorrelation in the error termin (6) Report the p-value and your test conclusion iii) Use Eviews automated procedure to perform a Breusch-Godfrey test of the mull hypothesis that there is no autocorrelation in the error termin (6) up to and including lag 12. Report the p-value and your test conclusion. (4 marks) (@) Use appropriate criteria to choose between the models in (5) and (6) and justify your choice (4 marks) (h) When the value of a regression coefficient changes between one sub-sample and another, the coefficient is said to exhibit a structural break. Inspection of the line graph of the series INF_FRANCE, suggests that the behavior of the series is qualitatively different before and after 1986ml. This may be due to a structural break in one or more of the regression coefficients i) Extend the model in (6) to allow for the possibility that there is a structural break in both the intercept and the time trend in the first month of 1986 and report the extended model ii) Estimate the model specified in i) and report the estimated model. Place the Eviews output in Appendix 2 (h) iii) What is the estimated change in the intercept between the two sample periods? iv) What is the estimated change in the marginal effect of the time trend between the two sample periods? V) Is the estimated change in the marginal effect of the time trend what you expected? Briefly explain (8 marks) (h) When the value of a regression coefficient changes between one sub-sample and another, the coefficient is said to exhibit a structural break. Inspection of the line graph of the series INF_FRANCE, suggests that the behavior of the series is qualitatively different before and after 1986ml. This may be due to a structural break in one or more of the regression coefficients. i) Extend the model in (6) to allow for the possibility that there is a structural break in both the intercept and the time trend in the first month of 1986 and report the extended model. INF_FRANCE = Bo + a time, + +381 BINF_FRANCE,-i + B12INF_FRANCE-12 + up. (0 i=1 Let cpifrt-1 denote the annualized monthly % change in the CPI in France. Transform the observations on cpifr into observations on INF_FRANCE,. You should find that the sample mean of INF_FRANCE; is 4.406853. The data for Question 2 is contained in the Eviews workfile inflation.wfl in the Assignments folder. The workfile contains monthly observations on the consumer price index for France, cpifr, for the period 1981m1 to 1996m6. The workfile also contains a dummy variable, dum, and a time trend, time. Dum assumes the value 1 from 1981ml to 1985m12, and the value zero from 1986ml to 1996m6. The variable time assumes the value zero in 1981ml and increases by 1 in each subsequent time period