MANG6480 Advanced Time Series Modelling Seminar 1 Contacts - Lectures • Prof. Tapas Mishra • Email Address: T.K.Mishra@soton.ac.uk • Dr. Soumyatanu Mukherjee • Email Address: S.Mukherjee@soton.ac.uk • Dr. Ahmad Maaitah • Email Address: Ahmad.Maaitah@soton.ac.uk Contacts – Teaching Assistants • Miss Yue Shi • Email Address: ys1u16@soton.ac.uk • Mr. Yifu Li • Email Address: yl12y18@soton.ac.uk Who We Are? • Miss Yue Shi • 3rd Year PhD student in the department of finance, University of Southampton. • MSc in Finance from the University of Southampton. • My research interests are mainly in the following area: • Corporate Finance • Office hours : Appointments by email. • Email: ys1u16@soton.ac.uk Who We Are? • Mr. Yifu Li • 1st Year PhD student in the department of finance, University of Southampton. • MSc in Finance from the University of Southampton. • My research interests are mainly in the following area: • Betting Market • Office hours : Appointments by email. • Email: yl12y18@soton.ac.uk Outline of the Seminar 1 • Part 1: Multiple Choices • Part 2: Lab session (Based on STATA) Part1: Multiple Choices • Question: • Which are correct description of Hodrick-Prescott Filter (a) removes the secular trend; (b) removes the seasonal trend; (c) regards any deviation from the potential level is only transient; (d) results in spurious dynamics that are not found in the underlying data. • Given the above, which of the following options is correct? • A: (1) Only (a) and (d) • B: (2) Only (b) and (c) • C: (3) Only (a) and (c) • D: (4) Only (c) and (d) Part1: Multiple Choices • Answer of the Question is D: Only (c) and (d). • Brief Explanation: • The Hodrick–Prescott filter is a mathematical tool used in macroeconomics, especially in real business cycle theory, to remove the cyclical component of a time series from raw data. • Restrictions of Hodrick–Prescott filter • Data exists in a I(2) trend. (If one-time permanent shocks or split growth rates occur, the filter will generate shifts in the trend that do not actually exist.) • Noise in data is approximately normally distributed. • Analysis is purely historical and static. Why we need filtering? • In real life, many economic/financial variables, such as GDP, grow over time. • To analyze the cyclical properties of such variables, as, for example, the business cycle, they are often corrected for a trend. • For this purpose, there are several filtering techniques available. • Filtering techniques help us to separate economic/financial time series data into two parts: • (1) The growth (trend) component • (2) The cyclical component • While, Hodrick-and-Prescott filter (HP filter) and Hamilton filter (proposed by James D. Hamilton in 2017) are the most popular tools for separating time series data. Part 2: Lab session (Based on STATA) • STATA commands which would be used for today’s seminar: • tsset • rename • tsfilter • tsline • tsfilter • hamiltonfilter (Unofficial Stata command) • ssc • help Part 2: Lab session (Based on STATA) • Use the dataset hamiltonfilterdquarterly.dta • Exercise One: • Set the data into quarterly frequency format. • Exercise Two: • Try to rename the variables. • Exercise Three: • Perform the Hodrick-Prescott filter on the variable called Log RGDP. Output the graph of cyclical component. Part 2: Lab session (Based on STATA) • Exercise Four: • Try to compute the Hamilton filter for this dataset. • Exercise Five: • Try to plot the lgrgdpusaq variable with its Hamilton filtered trend. • Exercise Six: • Try to plot the Hamilton filtered cyclical component. • Exercise Seven: • Try to compare the Hamilton filter with the Hodrick-Prescott filter. Exercise 1 Exercise 1 (contd), Exercise 2 Exercise 3 Exercise 4 Exercise 5 Exercise 6 Exercise 7: Part I Exercise 7: Part II Exercise 7: Part III