Reference no: EM132400229

Assignment -

This assignment will be an individual portfolio of up to 2,000 words. This will require students to demonstrate an understanding of the main principles of econometrics.

Task 1 -

Download data from the UKDS database for a country of your choice and complete the following question:

a. Consider a log-lin (growth model) model

Ln(RGDPt) = B₁+B₂time + U_t

RGDP_t is real GDP of a country

Download country specific data for the relevant variables for a period of at least 30 years and run a regression using Eviews and estimate the model.

Explain and analyse the above model. Explain the estimated slope coefficient and test its significance. Comment on the R-squared value and test the significance of the overall model.

b. Consider the linear trend model:

RGDP_t = B₁+B₂time + U_t

RGDP_t is real GDP of a country

Use the same data as part (a) and explain and analyse the above model. Explain the estimated slope coefficient and test its significance. Comment on the R-squared value and test the significance of the overall model.

c. Consider the non-linear trend model:

RGDP_t = B₁+B₂time + B₃time² + U_t

Estimate the non-linear model using the same downloaded data. Explain and analyse the non-linear trend model and comment on the sign and significance of the estimated slope coefficients. Comment on the R-squared value and test the overall significance of the model.

Compare the linear and non-linear models and explain which one you think is the better model using the different comparative criteria.

Carry out the Ramsey Reset test for functional form and interpret your test results.

Task 2 -

Download data from UKDS for the country of your choice and estimate the Phillips curve with annual data 1992 -2016:

√(lnf_t = α + βUnp_t + u_t)

Where √(lnf_t) is the inflation rate (%) and √(Unp_t) is the unemployment rate (% annual).

(a) Explain the theoretical relationships underlying the specification of the model.

(b) Analyse the estimation results and comment on whether the Phillips Curve holds for the country sample you are testing.

(c) Carry out the Breusch-Godfrey Serial Correlation Lagrange Multiplier Test and explain the conclusions you arrive at with regards to Serial Correlation.

(d) Carry out the Heteroskdasticity tests - (BP and White) and comment on your test results.

Task 3 -

Use World Bank data to construct Real Household Consumption Expenditure (RCONS) and Real Disposable Income (RYD).

Estimate the following consumption function with annual data (30 years).

√(RCOS_t = α₀ + α₁RYD_t + ε_t)

a. Define the concepts of for stationarity and non-stationarity and identify the problems that arise from non-stationarity.

b. Using the model above, test whether the variables are stationary. Show all the steps you have had to take and justify the conclusion you have arrived at. State the null and alternative hypotheses of the tests and your conclusions.

c. Explain the concept of cointegration. Using the appropriate techniques to test for cointegration. What impact does cointegration have on the robustness of the model?

Task 4 -

Using the same model in Task 3 forecast future values of Rcons_t using either the regression equation method or the Box-Jenkins method or ARIMA method of forecasting. Explain in detail all the steps your followed in order to achieve the best possible forecast value or range.

Note - Chosen country is PORTUGAL, so every question mentioned the data has to be linked to Portugal.