Forecasting Food Import Volumes & Values – US Supply Chain (2023)

Project Title:

Predicting Import Trends for Edible Products Using Time Series Forecasting Course Project – SCH-MGMT 663: Supply Chain Analytics, UMass Amherst

Overview: This project applies quantitative forecasting techniques to predict the 2023 import values and volumes of edible products entering the United States. Using data from the U.S. Department of Commerce, the analysis enables grocery store distributors to make informed decisions regarding procurement, inventory, and supplier management. With over 30% of food in the U.S. being wasted annually, this data-driven forecasting effort aims to contribute to more sustainable, efficient supply chain practices.

Objectives: • Forecast import values and volumes for food categories (e.g., fruits, beverages, live meat) • Identify top-performing and underperforming product categories • Determine percentage change in import volume between 2022 and 2023 • Pinpoint key international suppliers by volume and value • Evaluate model credibility and discuss alternative qualitative forecasting approaches

Methodology: To develop robust and adaptive forecasts, a mix of time-series forecasting techniques were employed: • Exponential Smoothing: Chosen for its responsiveness to recent trends and simplicity • Moving Average Model: Used to smooth out fluctuations and validate longer-term patterns • Linear Trend Forecasting: Cross-validated predictions across datasets • Assumptions included data stationarity and the absence of major geopolitical shocks in 2023.

Key Findings: • Fruits had the highest forecasted value for 2023 at $30.4 billion • Live Meat Animals had the lowest forecasted value at $2.83 billion • Fruit import volume projected at 15.68 million metric tons, a 2.44% increase from 2022 • Mexico was forecasted as the largest import source for fruits, with $24.2 billion in import value • Qualitative backup forecasts included Delphi method, Scenario Analysis, and Expert Panels

Tools Used: • Microsoft Excel (forecast modeling, visualization, pivot analysis) • Time Series Forecasting (Exponential Smoothing, Moving Average) • Descriptive Statistics • PowerPoint & Report Writing (for presentation delivery)

Files Included: Project_2_Report_Anand_Gupta.pdf: Full forecasting report Project_2_Anand_Gupta.xlsx: Cleaned dataset and forecasting models FoodImports.xlsx: Original dataset Project 2 Report.docx: Working draft of the final report

Applications: Improve inventory planning and reduce overstock, Optimize supplier negotiations by identifying demand patterns, Enable more accurate shelf-life and replenishment scheduling, Support data-backed sustainability goals through waste reduction

Author: Anand Gupta MS in Business Analytics – University of Massachusetts Amherst