ECON3183: Time Series Data Analysis
Replication Study
This document provides questions and requirements for the Replication Study of ECON3183, which accounts for 20% of the total marks.
A statement regarding academic honesty and the stance on using generative AI tools in this course:
To achieve the desired learning outcomes for this course, students must complete individual assignments, a test, and a term project that involves conducting empirical investigations/studies related to economics or finance. Students are expected to obtain data from a reliable source, perform. exploratory data analysis, propose a causal inference strategy with justification, conduct empirical studies, and interpret the results independently.
To ensure that students meet the intended learning outcomes for this course, generative AI tools are not allowed for any submissions (including drafts or final versions) unless otherwise specified in the assessment instructions. All work (including assignment reports, test answers, the term project report, and Stata codes) must be the student’s own or adequately attributed to its source. Using ChatGPT or other AI tools for CA is considered equivalent to receiving help from someone else. It raises concerns that the work is not the student’s own unless the instructor has provided specific instructions to the contrary.
Penalties for unacceptable AI use may include resubmitting the work, partial mark deduction, or receiving zero marks for the corresponding CA component. Turnitin’s ‘Similarity Check’ and ‘AI detector’ features will be used to monitor the use of AI tools in this course.
Deadline: by Nov. 30, 2025.
Submission Method:
a) Submit your typing assignment report in one single PDF file to Turnitin ‘Submission Link: Replication Study - Report’ via iSpace. The file name of your PDF submissions should have the following format: ECON3183_Replication Study_Student ID_Name in Pinyin (e.g., ECON3183_Replication Study_190000001_Mi Lin) and
b) Save your data and .do file(s) in a zip file. Name your zip file as ECON3183_Replication Study_Program_Student ID_Name in Pinyin. Then, upload your file to ‘Submission Link: Replication Study - Stata Data and Program’ via iSpace.
Assignment Guidelines:
Please note that this is not an essay-type assignment. Answer each of the following questions one by one. Modify the STATA example programs from the Lectures on iSpace as needed to address each question. Name your STATA programs (.do files) as "Q1", "Q2", etc., so that it is easy to match each program with its corresponding answer.
Read the lecture materials and the paper by Chow and Wang (2010), “The Empirics of Inflation in China,” to answer the following questions. Chow and Wang’s (2010) paper could be downloaded from the iSpace.
Present your empirical results, formula, and graphs with your comments/observations in a written report.
1) On the first page, paragraph three, the authors report that “The Augmented Dickey- Fuller tests strongly suggest the presence of a unit root in log(p) and log(M2/y).” Use the data from the paper to verify this statement. Provide a detailed account of the steps and reasoning behind your empirical approach, including comments on the findings. (15%, with Stata code performance 5% out of 15%)
2) Replicate the empirical result reported right after paragraph three, i.e., “ … regress log(pt(*)) on a constant and log(M2t /yt ) ”. Justify that this regression produces a cointegration relationship between log(pt(*)) and log(M2t /yt ). (15%, with Stata code performance 5% out of 15%)
3) Replicate both Figure 1 and Figure 2. (Tip: To match the graphs precisely, use these Stata graph command options: ylabel, xlabel, yscale, xscale, xmtick, and yaxis. Type ‘help graph’ in Stata for more details). Then, share your comments and analysis of these graphs. What key insights can be drawn from them? (30%, with Stata code performance 10% out of 30%)
4) On the second page, in the second paragraph, the authors mention that “The following (table) reports the result of the Chow test for parameter stability using t = 1979 as the breakpoint. The result provides extremely strong support for parameter stability of this equation.” Reproduce the authors’ Chow test for structural stability result to confirm their statement here. Show your calculation steps and explain your understanding of the degrees of freedom involved in the F-statistic reported in the table. Why is it necessary to perform a structural break test in this scenario? What issues might arise ifa structural break actually exists in the time series? Are there any limitations to using the “Chow test for parameter stability” approach? (40%, with Stata code performance 20% out of 40%)