Propuesta de un modelo econométrico para el pronóstico de ventas de la fábrica de balanceados EXIBAL-Riobamba

Abstract

This work proposes an econometric model for forecasting sales at the Exibal feed factory in the city of Riobamba using seasonal ARIMA models. For the investigation, the company’s monthly sales report in a period between January 2016 and October 2021, which were granted by the entity studied, was analyzed, being a total of 70 months used for the implementation of the Box-Jenkins methodology. Through an exploratory data analysis, the presence of atypical data was detected, in addition, the multiplicative decomposition of the time series was performed and the presence of stationarity was observed. With the use of the Box-Jenkins methodology, the model for the time series was found, which was diagnosed by means of the assumptions of normality, homoscedasticity and independence of its residuals, the model was presented as a SARIMA (1,1,2) (1,0,0) [12]. It is concluded that forecast sales in the period November 2021 - October 2022 decreased by $7328.02 on average with respect to the last year presented in the time series. The use of the econometric model is recommended as an internal evaluation tool of the company's financial processes in order to mitigate the causes and factors that lead to decrease or increase of the variable under study.