Modeling Pakistan’s Macroeconomic Dynamics

A Time Series Approach Using VAR, SVAR, and VECM

This repository presents an applied macroeconometric analysis of Pakistan’s economy using modern time series techniques. The study includes univariate analysis (time series plots, unit root tests, and ARIMA/ARMA models) as well as multivariate analysis, such as Vector Autoregression (VAR), Structural VAR (SVAR), and Vector Error Correction Models (VECM).

---

📈 Variables to be Used (Quarterly Data)

Variable	Source	Notes
Real GDP	PBS / SBP	Convert to log, adjust for inflation
Government Spending	SBP / IMF IFS	Central government expenditure (real)
Tax Revenue	MoF / SBP	Net revenue or tax-to-GDP ratio
Inflation (CPI)	PBS	Quarterly average of CPI
Interest Rate	SBP (Policy Rate)	Used as-is

---

🗃️ Data Sources

• State Bank of Pakistan (SBP)
• Pakistan Bureau of Statistics (PBS)
• IMF – International Financial Statistics (IFS) API
• World Bank – World Development Indicators (WDI)
• Ministry of Finance, Pakistan

---

🧭 Project Tasks & Workflow

1️⃣ Data Collection & Cleaning

• Collect at least 11 years (44 quarters) of data
• Convert variables into time series format
• Log-transform and seasonally adjust where necessary
• Plot and visually inspect all time series

---

2️⃣ Univariate Time Series Analysis

• Plot and describe patterns for each variable
• Conduct ADF and KPSS tests for stationarity
• Apply log transformation and/or differencing
• Fit ARIMA/ARMA models where relevant

---

3️⃣ Vector Autoregression (VAR)

• Ensure all series are stationary
• Select optimal lag length
• Estimate the VAR model
• Perform:
  • Granger causality tests
  • Impulse Response Functions (IRFs)
  • Forecast Error Variance Decomposition (FEVD)

---

4️⃣ Structural VAR (SVAR)

• Impose short-run identification (AB-model)
• Apply institutional restrictions (e.g., Blanchard–Perotti method)
• Plot IRFs for:
  • Government spending → GDP and inflation
  • Tax shocks → GDP
  • Interest rate → inflation

---

5️⃣ Cointegration and VECM

• Conduct Johansen cointegration tests
• Interpret long-run cointegrating relationships
• Explain short-run dynamics via error correction terms

---

📃 Deliverables

• Research Report including:
  • Introduction
  • Methodology
  • Results and interpretation
  • Policy implications for Pakistan
• Quarto Notebook
• Rendered HTML file with complete code and graphical outputs