Questions

2. (55 marks) From any (reliable) data source choose a univariate time series of your
interest, with at least 50 equally spaced data points. (If you choose a time series
that has already been provided to you in this module, either as part of an application
in the lectures or for the labs then you will be penalised with minus 25 marks on this
question.)
(a) [10 marks] Explain why your chosen time series is of interest and discuss its features
(for example, trend, cycle elements, structural breaks) in no more than 300 words.
Make sure to cite your data source.
(b) [10 marks] Determine if your time series is stationary or not using the Augmented
Dickey Fuller (ADF) test. Explain clearly what you are doing.
(c) [25 marks] Use the Box-Jenkins procedure to obtain the most suitable model for
your time series. Explain each step carefully. (Note that if the series is non-
stationary, you need to use the appropriate transformation to make it stationary.)
(d) [10 marks] Consider the ‘best’model you picked in part (c) and perform the
Lagrange Multiplier (LM) test on the residuals to determine whether or not the
residuals show conditional heteroscedasticity. Explain clearly what you are doing.
3. (15 marks) After reading Engle (2001), write a short re‡ection (400 words maximum)
addressing the following:
 What were the key points in the article that you found most interesting, and why?
 How could the …ndings and ideas of the article inform your own future research
involving macroeconomic or …nancial data?
Engle, R. (2001). GARCH 101: The Use of ARCH/GARCH Models in Applied Econo-
metrics, Journal of Economic Perspectives, 15(4), 157-168.
(Note: You should be able to get the article from JSTOR, via the library.)
2