Saladette Tomato Price Forecasts in Mexico With the Application of the Box-Jenkins Methodology

Abstract: The Box-Jenkins Methodology was used in this work in order to identify an autoregressive integrated moving average (ARIMA) model that might be suitable to the time series of average wholesale prices behavior of Saladette tomato in Mexico in order to predict future values of such product in the supply centers of the 31 states of the Mexican Republic from January 2011 to December 2020. Data was obtained from the National Market Information System of the Federal Government’s Ministry of Economy. The data processing was carried out with IBM SPSS V25 software and with Excel 2007 software. Results indicate that there is an increasing positive trend, seasonality, without cyclical effect and with random variations in monthly prices. The regression analysis for tomato prices indicates a correlation coefficient of 0.55, an average association. The analysis of variance rejected the null hypothesis of equality between prices. Once the regression equation has been calculated, it can be verified with the ARIMA Model (1, 0, 1), which firstly allows us to calculate a correlation adjustment, whose results indicate a 0.558 R² value. With this model the average prices per kg of Saladette tomato for the year 2020 were forecasted.

Key words: time series; trend; seasonality; average price

         JEL code: C22