Project: Global Commodity Forecasting (The Coffee Index)
Overview
For my Senior Capstone, I am building a Time Series Predictive Model to analyze Global Commodity Inflation, specifically focusing on the coffee supply chain.
The Problem
Coffee prices are volatile and heavily influenced by supply shocks in specific regions (e.g., frosts in Brazil or production booms in Vietnam). My goal is to determine if production data from the USDA can accurately predict the US Consumer Price Index (CPI) for coffee with a 6-month lag.
"A predictive dashboard that models how global events, like a frost in Brazil or a shipping crisis in Vietnam, impact the price of your morning coffee."
Methodology
I am developing an interactive R Shiny dashboard that allows users to simulate supply shocks (e.g., "What if Brazil's harvest drops 20%?") and forecast the resulting price impact. The model is trained on 30 years of historical data from:
- USDA Foreign Agricultural Service: Production, exports, and ending stocks.
- FRED (St. Louis Fed): US City Average CPI for Coffee.
- Cecafé & ICO: Brazilian Monthly Exports and Composite daily prices.
Current Status
I have gathered the core datasets and am currently cleaning the USDA production tables in R Studio to prepare them for regression analysis. My stretch goal is to incorporate Milk and Cocoa data to build a "Latte Index" that tracks the total cost of a cup of coffee.
Project Timeline
