Indicadores antecedentes compostos da agroindústria

Abstract

The interest and especially the need of today's global economy to understand the present and anticipate the future, even in the short term, makes the study of cyclical forecasting and the leading indicators of extreme importance. Thus, this study aims to create three composite leading indicators for GDP of the Brazilian Agribusiness, with the forecast horizon between 2-4 months, 5-8 months and 9-12 months respectively named short, medium and long term. For this purpose, a survey was made of the current state of the art, mainly produced in Brazil. Introductorily, we present the concept of the cycle and leading indicators, as the justification and importance of this issue. Then, a survey of the literature on business cycles, addressing seminal publications such as Burns and Mitchell (1946), and the current discussion between business cycles and growth cycles. By then, it was mentioned the concept of leading indicators, its origin, the main methods used and current work on the subject. Finally, we built a methodology, based on the model proposed in OECD (2008) with addition of VAR models, Granger Causality and Probit. Being tested and evaluated, for the monthly information of physical production of Agribusiness in Brazil and other 421 series candidates as leading indicators, for period between January 1995 and December 2011. Completing positively to the possibility of creating composite leading indicators, whether short, medium or long term, for Brazilian Agribusiness.