Time Series Econometrics & Macroeconomic Forecasting

EC614 · School of Economics, Quaid-i-Azam University, Islamabad

Time series, cointegration and macroeconomic forecasting

Course materials, hands-on R labs and applied Pakistani case studies from Prof. Dr. Zahid Asghar, School of Economics, Quaid-i-Azam University.

Everything here is reproducible: each lecture and lab is a Quarto document that pulls real data — SBP, PBS, World Bank, Yahoo Finance, UN Comtrade — and runs the estimation in front of you.

14 weekly lectures · 10 hands-on labs · 90+ datasets · all code in R

Start here

Step 1 Course outline Learning outcomes, assessment, textbooks and the week-by-week plan. Step 2 Lectures Week by week, from stationarity and unit roots through ARIMA, ARCH, cointegration, ARDL and structural VARs. Step 3 Labs Work the same methods yourself in R, on Pakistani and international data.

What the course covers

The sequence is deliberate. We begin with the single question that decides everything else — is this series stationary? — because getting it wrong produces regressions that look magnificent and mean nothing. From there the course builds outward: modelling a single series, quantifying how wrong the forecast is likely to be, then moving to systems of series where the interesting economics lives.

Block	Topics	When
Foundations	Stationarity, unit root testing, spurious regression	Weeks 1–5
Univariate modelling	ARIMA, forecast uncertainty and evaluation, ARCH/GARCH volatility	Weeks 6–8
Multivariate systems	Cointegration and VECM, ARDL bounds testing, VAR, structural VAR	Weeks 9–13
Panel data	Fixed and random effects, panel unit roots	Week 14

Software

All work is done in R with Quarto. The core stack is the tidyverse plus fable/feasts for forecasting, urca and tseries for unit root and cointegration testing, vars and svars for VAR work, ARDL for bounds testing, and plm for panels. Selected Python equivalents live in R/methods/ for students who prefer them.

install.packages(c("tidyverse", "fpp3", "fable", "feasts", "tsibble",
                   "urca", "tseries", "vars", "svars", "ARDL", "plm",
                   "readxl", "quantmod", "WDI"))

How this repository is organised

Folder	What lives there
`lectures/`	The 14 weekly lecture documents, named `week-01-…` to `week-14-…`
`labs/`	Hands-on R walkthroughs, including `labs/r-foundations/` for students new to R
`applications/`	Applied Pakistan studies — remittances, exchange rate, exports, tax, solar prices, KSE
`assessments/`	Exams, assignment briefs and the term paper specification
`slides/`	Standalone decks (PPTX and revealjs)
`R/`	Standalone scripts: `setup/`, `textbook/` (Enders, Stock–Watson), `methods/`, `applications/`
`data/`	Datasets used across the course; `data/large/` holds files too big for git
`references/`	Reading material, IMF forecasting module, xts/zoo chapters
`shiny/`	Small interactive apps
`docs/`	Generated site output — never edit by hand

Rebuilding the site

quarto render          # writes the full site into docs/

GitHub Pages serves docs/ on the main branch. Rendered chunk output is frozen under _freeze/, so a rebuild only re-executes documents you have actually changed.