Comparison of models for estimating the planted area of Agave spp. for mezcal production in Mexico

Abstract

Objective: The objective of this study was to compare forecasting models for the planted area of agave mezcalero for the period 2023–2027, using data from 1982–2022 through autoregressive integrated moving average (ARIMA) models. This analysis aimed to promote the efficient management of natural resources.

Design/methodology/approach: To validate the forecasting efficiency, ARIMA models were compared with simple forecasting methods, such as the mean and naïve approaches, using out-of-sample evaluations.

Results: The findings revealed a growth scenario for the planted area of agave mezcalero during 2023–2027. The most accurate model, an autoregressive (4,0,0) specification, projected that the maximum planted area would be reached in 2024, followed by a gradual decline beginning in 2025.

Limitations on study/implications: Predictions are constrained by the availability and quality of databases. The primary limitation encountered in this study was the availability of historical data.

Findings/conclusions: The study concludes that the ARIMA model is an efficient tool for estimating the planted area of agave mezcalero. The cultivation of mezcal agave occurs in a context of overproduction. The generated information can support planning and investment by the industry, producers, and government entities.