Demand forecasting for agrochemicals: fungicides, herbicides, and pesticides

Demand forecasting and regression analysis for agrochemicals such as fungicides, herbicides, and pesticides is essential for optimizing supply chain management and ensuring the timely availability of critical crop protection chemicals. This independent study presents predictive modeling techniques which are the constant model (using Moving Averages of 3 and 6 periods), and trend-seasonal models, including Time Series Decomposition (Simple and Ratio-to-Trend methods) and Holt-Winters’ multiplicative seasonal model and to enhance demand forecasting and accuracy was evaluated using Mean Absolute Deviation (MAD) and Mean Absolute Percentage Error (MAPE) to determine the best-fitting model. In addition, regression analysis was conducted to explore the factors influencing demand for each category of agrochemicals, recognizing that different products are affected by different variables. Historical sales data and relevant external factors were used to support the analysis. The results indicate that the trend seasonal model in demand forecasting is most suitable for three of the product categories, but for different types of models (TSD (simple) for fungicides, Holt-Winters model for herbicides, and TSD (Ratio-to-Trend) for pesticides). Furthermore, the study found that external factors influencing demand vary across categories. This study highlights the importance of data-driven forecasting in minimizing supply shortages, avoiding overstock, and understanding the external factors for different agrochemical products.