BU75101: an undergraduate finance student was interested in the relationship between salary, education level, and gender and gathered data on a cross-section of 30 workers: Financial Econometrics Assignment, TBS, Ireland

1. An undergraduate finance student was interested in the relationship between salary, education level, and gender and gathered data on a cross-section of 30 workers. She estimated the model using Ordinary Least Squares (OLS) in EVIEWS, some of the results from which are presented below.

(a) Write down and explain a theoretical equation that will represent the model
estimated above.

(b) Comment on the goodness of fit of the model, make sure to explain what this measure is attempting to achieve.

(c) Test the ‘EDUCATION’ coefficient using the 5% level of significance.

(d) Use the P-value (‘Prob.’ in EVIEWS) to comment on the ‘MALE’ coefficient.

(e) Construct and interpret a 90% confidence interval on the ‘EDUCATION’
coefficient.

(f) Suggest areas of improvement for the model outlined by the student, be
specific and use equations and/or examples to support your solution.

2. Select any econometric model that you have studied during your current or past studies and answer the following questions based upon this model. Ensure to use equations and to cite relevant journal articles where necessary.

(a) Write down and explain a regression equation that would allow you to estimate this model, explaining the functional form selected.

(b) Explain and describe what data you would need to carry out this study.

(c) How was the model estimated (e.g. a Probit model) and discuss the selection of this estimation procedure.

(d) Discuss any econometric testing issues that you would expect to encounter with this model, explaining why such issues would arise with your model and how you would test for the issue.

(e) Assuming the econometric issue that you tested for in part (d) was present in your project, explain how you would solve the issue arising.

3. (a) Assume that equation (1) below is a correctly specified econometric model. However, an undergraduate finance student mistakenly estimates equation (2) instead. Explain with the aid of equations what the implications for the OLS estimates of equation (2) are in this instance.

(b) Explain the use of dummy variables in econometrics. Ensure to use equations and diagrams in your solution.

(c) Below is output from EVIEWS where a test has been carried out for stationarity on data for the FTSE index. Interpret the outcome of this test ensuring to set up the null and alternative hypothesis initially, and also offer recommendations as to what the researcher should do next given this result.

(d) Describe with the aid of diagrams, equations and examples the differences
between panel and cross-sectional data sets.

(e) An undergraduate finance student is conducting a research project and develops equation (3) below. Explain using equations how the student could check the equation for the presence of multicollinearity.

4. (a) Explain with the aid of equations how the Durbin-Watson test is constructed. Ensure also to discuss the function of the test in econometrics.

(b) Equation (4) below violates a classical assumption. Describe with the aid of
equations how the equation can be formally tested for this violation.

(c) Demonstrate with equations how to test a model for potential ARCH effects.

(d) Explain with the aid of equations what an (Autoregressive Moving Average) ARMA (2,2) process is.

(e) Explain the aid of equations the role of the Ramsey test in econometric modelling.

5. (a) Explain the use of the Tobit model in financial econometrics. Use existing literature to provide an example of where the model has been applied.

(b) Explain with the aid of equations and existing literature when a Multinomial Logit model would be required in econometrics

(c) A student was conducting research using panel data and carried out a Hausman test. A summary of the results from the test is presented below.

Based upon these results and focusing upon panel data, what conclusion can you draw? Make sure to set up a null and alternative hypothesis for the test.

(d) Econometrics as a discipline is well positioned to handle the relatively recent interest in the analysis of ‘big data’. Discuss this statement and in so doing refer to research that supports your case.