Supply Chain Pressure and US Inflation

Analysis of the relationship between the NY Fed Global Supply Chain Pressure Index (GSCPI) and US Consumer Price Index (CPI) inflation, including a Streamlit dashboard and VAR forecast model.

Overview

The data pipeline ingests monthly GSCPI readings from the NY Fed and CPI sub-indices from FRED, merges the dataset, and executes the following statistical procedures:

• Augmented Dickey-Fuller stationarity tests
• Granger causality tests (lags 1-12 months)
• Lagged cross-correlation analysis
• VAR(p) forecast 12 months ahead

Results are displayed via an interactive Streamlit dashboard featuring four analytics tabs: historical comparison, lagged correlations, Granger causality, and VAR forecast.

A GitHub Actions workflow automatically updates the underlying datasets on the 5th of each month.

Data Sources

Series	Source	Frequency
CPI (all items, core, food, energy, vehicles, shelter)	FRED (St. Louis Fed)	Monthly
Global Supply Chain Pressure Index	NY Federal Reserve	Monthly

Project Structure

.
├── data_pipeline.py          # Fetches and merges CPI + GSCPI data
├── statistical_analysis.py   # Stationarity, Granger, lagged correlation, VAR forecast
├── app.py                    # Streamlit dashboard
├── forecast_model.py         # Standalone VAR forecast script
├── requirements.txt
├── Dockerfile
├── .github/
│   └── workflows/
│       └── update_data.yml   # Monthly cron job
├── cpi_data.csv              # Raw CPI series (auto-updated)
├── gscpi_data.csv            # Raw GSCPI series (auto-updated)
├── merged_data_for_bi.csv    # Merged dataset used by the dashboard
└── forecast_data.csv         # 12-month VAR forecast output

Getting Started

Clone the repository and install dependencies:

git clone https://github.com/atharvasathaye/Supply-chain.git
cd Supply-chain
pip install -r requirements.txt

Refresh the data (optional, as the repository contains updated data):

python data_pipeline.py
python statistical_analysis.py

Run the dashboard:

streamlit run app.py

Run using Docker:

docker build -t supply-chain .
docker run -p 8501:8501 supply-chain

Key Findings

• GSCPI Granger-causes CPI at all lags 1-12 months (p < 0.0001 at lag 1).
• Peak lagged correlation occurs at r = 0.63 with a 10-month lag.
• The VAR model forecasts YoY CPI inflation at approximately 3-4% through mid-2027.

These empirical results align with economic literature indicating that supply chain disruptions impact consumer prices with a 6-12 month lag.

Automated Updates

The workflow defined in .github/workflows/update_data.yml executes on the 5th of every month. It retrieves the latest CPI and GSCPI data, regenerates the merged dataset and forecasts, and commits the updated CSV files back to the repository. Business intelligence tools such as Power BI or Tableau can connect directly to the raw GitHub CSV URLs.

Requirements

• Python 3.10+
• Dependencies listed in requirements.txt

Author

Atharva Sathaye