# ECONOMETRICS

ECONOMETRICS

please answer the first and the fourth questions in the attached paper using the xl applications and the spss to do the regresion

• Groups of four people will be formed (although it is expected that this assignment can be completed by three people if there are not exact groups of four at the time the groups are formed)

• The handup should include a hard copy of a Word document (report on your answers to the four questions), and soft copies of the relevant Excel, SPSS (if used) and Powerpoint files (these can be uploaded to Moodle (Group Assignment Submissions before the start of class on 11th December). These submissions are worth 30% of the marks for this module.

• Each group will present for no more than 15 minutes on Tuesday – there are four questions and group one will present Q1, group two will present Q2, etc. (For the GFIS Class, Group 5 will present Q1 and Group 6 will present Q2). A draw for the presentations will be made in session seven. All group members are asked to speak during their group’s presentation (otherwise, marks may be lost).

• The relevant data file for question 3 of the assignment will be on Moodle (Group Assignment CPI data).

• In general, the assignment is designed to test your knowledge and understanding of material covered in the opening six sessions as well as some material to be covered in subsequent sessions (hence it will take some time before you are in a position to complete all questions), and marks will also be given for the overall presentation of your written report.

Q1. Generation of a Regression Model for a Recent Sports Event

Each group will work on a recent sports event and collect data for 32 countries as follows:

Group 1 2010 World Cup Soccer (South Africa) 32 countries

Group 2 2006 World Cup Soccer (Germany) 32 countries

Group 3 2008 Olympic Games 2008 (Beijing) Top 32 countries

Group 4 2004 Olympic Games 2004 (Athens) Top 32 countries

Group 5 2002 World Cup Soccer (Japan & South Korea) 32 countries

Group 6 2000Olympic Games 2008 (Sydney) Top 32 countries

Come up with a quantitative variable called OUTCOME to measure a country’s success in the relevant event. For example, this could be 1 (winner), 2 (runner-up), 3 (3rd), 4 (4th), 8 (beaten quarter finalist), etc. for the World Cup, and 1 (overall total number of medals), 2 (second highest number of medals) etc for the Olympic games.

Each group will regress OUTCOME on five independent variables as follows:

• Population (year in question)

• GDP ($ PPP) per capita (year in question)

• Three other variables which you propose to be related to OUTCOME (explanations as to the direction/nature of this link to be provided in your written report). Note that one of these three additional variables should be a dummy variable.

Note that each group will need to collect data from a variety of sources (IMF WEO Database, Wikipedia etc) to get the relevant data for this question.

(a) Present the relevant data file of all the variables in the model. Include a Population Regression Model (PRM) in this presentation and comment on the expected relationships between Y and the X variables (before the regression is performed)

(b) Present the results of the regression in standard format, and comment on the overall

explanatory power of the regression model

(c) Interpret the regression coefficients and comment on their statistical significance.

(d) Comment on the extent of multicollinearity in the data

(e) Is there any evidence of heteroscedasticity in the data

(f) Offer brief suggestions for an “improved” regression model

Q2. Determination of the Average Standard of Living

Use the WEO Database for advanced economies (n=35). Each group has a different current year.

PRM: Yi = ? + ?1X1i + ?2X2i + ?3X31i+ ei

where Y = GDP per capita ($000, PPP) in the current year

X1 = Total investment (% of GDP): Mean of the preceding 10 years

X2 = Government spending (% of GDP): Mean of the preceding 10 years

X3 = Crude labour productivity growth rate (%): Mean of the preceding 10 years [Hint: X3 is a derived variable got from annual average growth in real GDP minus annual average growth in employment]

(a) Set up the data set and include your workings as a separate sheet in your Excel submission.

(b) Present the regression results in the standard format

(c) Verify showing the relevant workings and calculations the value of R squared and adjusted R squared.

(d) Interpret R squared, the regression slope coefficients and their associated 95% confidence intervals. Comment briefly on the goodness of fit of the regression equation.

(e) Plot a histogram of the residuals (from SPSS) and carry out the relevant test for normality (Jarque-Bera test) in the error term

(f) Carry out the following hypothesis tests and comment on their findings:

1. Tests of the three regression slope coefficients

2. Test of the overall goodness of fit of the regression model

(g) Add a dummy variable (D) for euro countries and

1. Specify the revised PRM and SRM

2. Write out the regression results in the standard format

3. Interpret the new regression coefficient and do the relevant statistical test on it

4. Comment on the improvement or otherwise in the regression model and the impact on the original slope coefficients arising from the inclusion of the additional variable.

Current Year: Group One (2011), Group Two (2010), Group Three (2009), Group Four (2008), Group Five (2007), Group Six (2006)

Q3. Augmented Okun’s Law

Firstly, go to the latest Statistical Annexe of the OECD’s Economic Outlook (Excel files) – these can be found on the following link:

http://www.oecd.org/document/61/0,3746,en_2649_34573_2483901_1_1_1_1,00.html

Run the following augmented Okun’s Law regression for 2006 (Group 1), 2007 (Group 2) 2008 (Group 3), 2009 (Group 4), 2010 (Group 5) and 2011 (Group 6) across OECD member countries.

Where ?UR = Percentage point change in the annual unemployment rate

GDP = Annual percentage change in real GDP i.e. economic growth rate

RIR = Real interest rate (Short term interest rate* minus CPI inflation)

LPG = Labour productivity growth (for the total economy)

* Euro area countries will have the same interest rate as that for the euro area as a whole

(a) Present the regression results in standard format

(b) Interpret all the regression coefficients

(c) Comment on the possible existence of multicollinearity in this data

(d) Comment on the extent of heteroscedasticity in this data

(e) Conduct a Ramsey RESET test of the model. What do you conclude from the test?

Q4. Autocorrelation Analysis of CPI Data

The data sets CPI 1, 2, 3, 4, 5 and 6(on Moodle) contain annual CPI data for Ireland for the years 1976 through 2011 using base years 1986, 1991, 1996, 2001, 2006, and 2011 respectively.

Group 1, 2, 3, 4, 5 and 6 use CPI1, 2, 3, 4, 5 and 6 respectively.

(a) For your data set present the autocorrelation functions using a maximum of lag of 6 for CPI and ?CPI and comment on your results

(b) For your data set run the AR(1) model. Show the relevant graph and indicate whether the series is stationary or non-stationary

(c) For your data set run AR(4) with Deterministic Trend model. Comment on the regression results and conduct the relevant statistical tests of the regression coefficients. Indicate whether or not this model is “complete” and why this is so.

(d) Comment on the effect of the downward pressure on prices in the years 2009/2010 on the above findings