辅导ECON3031、辅导Python/C++编程
- 首页 >> C/C++编程 ECON3031 Applied Econometrics 2022-23
EXERCISE 1
1. The EViews work?le oil5.wf1 (available on Blackboard) contains quarterly ob-
servations on the price of oil (West Texas Intermediate, in US$) from 1980Q1
to 2016Q1.
(a) First plot the observations, and comment on what you observe. Using
data from 1980Q1 to 2015Q2 only, test whether the oil price series is
stationary or nonstationary. [View, Unit Root Tests, Standard Unit Root
Tests, and accept all the default options] What do you conclude is the
order of integration of the oil price series?
(b) Using the information from part (a) and the sample period 1980Q1 to
2015Q2, specify and estimate AR(1) models for (in turn) the price of oil,
and the quarterly change in the oil price. Is either of these models sensible?
(c) Use each model estimated in part (b) to forecast the price of oil for 2015Q3,
2015Q4 and 2016Q1. [Use a spreadsheet to make the calculations, or in
EViews: Forecast, change the forecast sample to 2015Q3 2016Q1, change
the method to ?Static forecast?and untick the box ?Insert actuals for out-
of-sample observations.] How (in)accurate are your forecasts?
2. The EViews work?le freddie1.wf1 (available on Blackboard) contains a monthly
index for the price of houses in Beckley, West Virginia, USA (labelled bekly),
and the monthly value of Australian exports to China (labelled xchina) from
January 1988 to December 2015.
(a) Estimate the regression equation XCHINAt = 1 + 2BEKLY t + et, and
comment on the results.
(b) Now plot the series BEKLY t, XCHINAt, ln(XCHINAt), and the residu-
als bet from your earlier regression, and describe the graphs. [To obtain
ln(XCHINAt), choose Genr and then lnxchina=log(xchina) under ?En-
ter equation?] Do they provide any insights into your results from part
(a)?
(c) Now estimate the equation ln(XCHINAt) = 1+ t+2BEKLY t+et, and
comment on the results. Suggest a reason why ln(XCHINAt) rather than
XCHINAt was chosen as the left-hand-side variable.
EXERCISE 1
1. The EViews work?le oil5.wf1 (available on Blackboard) contains quarterly ob-
servations on the price of oil (West Texas Intermediate, in US$) from 1980Q1
to 2016Q1.
(a) First plot the observations, and comment on what you observe. Using
data from 1980Q1 to 2015Q2 only, test whether the oil price series is
stationary or nonstationary. [View, Unit Root Tests, Standard Unit Root
Tests, and accept all the default options] What do you conclude is the
order of integration of the oil price series?
(b) Using the information from part (a) and the sample period 1980Q1 to
2015Q2, specify and estimate AR(1) models for (in turn) the price of oil,
and the quarterly change in the oil price. Is either of these models sensible?
(c) Use each model estimated in part (b) to forecast the price of oil for 2015Q3,
2015Q4 and 2016Q1. [Use a spreadsheet to make the calculations, or in
EViews: Forecast, change the forecast sample to 2015Q3 2016Q1, change
the method to ?Static forecast?and untick the box ?Insert actuals for out-
of-sample observations.] How (in)accurate are your forecasts?
2. The EViews work?le freddie1.wf1 (available on Blackboard) contains a monthly
index for the price of houses in Beckley, West Virginia, USA (labelled bekly),
and the monthly value of Australian exports to China (labelled xchina) from
January 1988 to December 2015.
(a) Estimate the regression equation XCHINAt = 1 + 2BEKLY t + et, and
comment on the results.
(b) Now plot the series BEKLY t, XCHINAt, ln(XCHINAt), and the residu-
als bet from your earlier regression, and describe the graphs. [To obtain
ln(XCHINAt), choose Genr and then lnxchina=log(xchina) under ?En-
ter equation?] Do they provide any insights into your results from part
(a)?
(c) Now estimate the equation ln(XCHINAt) = 1+ t+2BEKLY t+et, and
comment on the results. Suggest a reason why ln(XCHINAt) rather than
XCHINAt was chosen as the left-hand-side variable.