EF308 – Econometrics and Forecasting
# Updated February 9th – to remove page limit
# Updated March 11th – to allow some group work
This is the second of two workbooks you are asked to complete for the EF308 module. This workbook
is worth 50% of the full module grade and relates to material covered in Part 3 (Financial Testing and
Forecasting) of the module.
You are asked to complete the workbook, including the written responses, in a Python document. See
the video recorded for the first assignment (in the Assessment Tab on Loop) for how to do this. When
you are finished the workbook, you should do the following:
• Download the .ipynb file (the python file)
• Print to PDF the .ipynb file with every cell run
• Upload both files to the Assignment 2 part of Loop
The assignment is due for submission by April 12th 2021.
A big part of using Python effectively is to use codes that others have developed, so as not to duplicate
efforts. You are, therefore, allowed to use code from elsewhere as part of answering this assignment.
You are allowed to work on this assignment with one other person from your class. You should name
that person in your submitted assignment. You are allowed to have 80% of words in common with
that person in your submitted assignment, but your grade awarded will be individual (but will
presumably be quite similar). What I’m therefore allowing you to do is to work in a team, but asking
you to put the finishing touches on the assignment yourself.
Please try not to ‘over-answer’ the workbook. The total writing should be about 1500 words. A good
answer should be quite comprehensive, but also not needlessly wordy.
For the dataset, IrishBanks.csv, load the dataset in Python and answer the following questions /
requests, including showing all code used. The dataset contains the stock prices of two Irish banks –
AIB and Bank of Ireland (BoI):
1. Describe both banks individually using any appropriate descriptive statistics and visuals,
including ARMA processes and unit roots. [30 marks]
2. Forecasting: [40 marks]
a. Show how you might forecast both series individually using their ARMA processes. [10
b. Analyse the volatility of both series individually and demonstrate how a GARCH
process can be used to model their volatility. [15 marks]
c. Develop a VAR model of the two series and show how one series might influence the
other. [15 marks].
Lastly, returning to the dataset that you used in the first assignment, HousePrices.csv, you are asked
3. Use appropriate machine learning techniques to develop a model to explain house prices. [30