联系我们： 手动添加方式: 微信>添加朋友>企业微信联系人>13262280223 或者 QQ: 1483266981

Dr. Stefano Soccorsi LUMS, 8 May 2023
Time Series
(Econ 420)
Coursework
Electronic submission due on May 15, 2023 by 12 noon
Guidelines
1. The submitted file must be in pdf format with the following name
stud# studname.pdf
where: stud# is your student number and studname is your name in the format sur_x005f name name.
2. Do NOT include R codes.
3. Word limits: no more than
(a) 500 words in Exercise 1
(b) 150 words in Exercise 2
(c) 150 words in Exercise 3
4. All results presented must be interpreted: include the figures and tables you comment and
nothing else.
5. Maximum filesize: 2MB.
6. For support with submissions refer to John Sharples (j.sharples8@lancaster.ac.uk).
Exercise 1
Import the data in the file ex1.txt there are 39 years (1970-01-30 to 2008-12-31) of monthly
returns of an imaginary stock. 1 Letting s be a scalar obtained summing the day and month of
your birthday – e.g. May 11 would give s = 5 + 11 = 16 – run the following command:
set.seed(s)
x=x+.01*rnorm(length(x),0,1)
Estimate an AR(1) and explain if it is an adequate model.
[50 marks]
1
Import the data using the following command: x =read.table(“ex1.txt”,header=T)[,1].
1
Exercise 2
Import the data in the file ex2.txt, and defining s as in Exercise 1 do
set.seed(s)
y=y+.15*cumsum(rnorm(length(y),0,1))
Estimate an appropriate ARIMA model motivating your decision.
[25 marks]
Exercise 3
Download the Dow Jones Industrial Average and the SP500 in the sample that starts on your
birthday in 2010 and ends on your birthday in 2018. 2 Are these two time series cointegrated
[25 marks]
2To do this in R use the command get.hist.quote with instrument =
′′ ^dji ′′ for the former and instrument
=
′′ ^gspc ′′ for the latter.
2