Authors: Bianca Reichert; Adriano Mendonça Souza; Wesley Vieira Da Silva; Claudimar Pereira Da Veiga
Addresses: Department of Statistics, Federal University of Santa Maria (UFSM), 1000 Roraima Ave, Building 13, Office 1205 C, CCNE/UFSM, Santa Maria, RS, Brazil ' Department of Statistics, Federal University of Santa Maria (UFSM), 1000 Roraima Ave, Building 13, Office 1205 C, CCNE/UFSM, Santa Maria, RS, Brazil ' Federal University of Alagoas (UFAL), Av. Lourival Melo Mota, S/N, Tabuleiro do Martins, 57072-900, Maceió, Alagoas, Brazil ' Postgraduate Program in Organizations Management, Leadership and Decision (PPGOLD), Federal University of Parana (UFPR), 632 Lothário Meissner Ave, Jardim Botânico, 80210‐170, Curitiba, PR, Brazil
Abstract: The construction sector is one of the essential sectors for the country's economic development, in which the retail of building materials is the link between cement industries and consumers. This study aimed to identify the most accurate method to predict the cement purchase price for building materials retail in Brazil, using the individual autoregressive integrated moving average (ARIMA) models and their forecast combinations, and to evaluate peak periods of cement price using residues control charts. The study contribution proposed a forecast combination method based on the Akaike information criterion (AIC) and a price monitoring tool using residues control charts. As a result, the proposed combination - AIC module - presented better forecasting statistics (MSE = 1.15). The control charts analysis was possible to verify periods of high and low cement purchase price and its causes.
Keywords: ARIMA models; forecasting combinations; control charts; construction industry; times series; marketing intelligence; B2B; purchase.
International Journal of Business Forecasting and Marketing Intelligence, 2021 Vol.7 No.1, pp.1 - 12
Received: 10 Jul 2020
Accepted: 28 Jan 2021
Published online: 07 Jan 2022 *