Simulação computacional para avaliação de cenários econômicos de cultivo agrícola

Abstract

Brazil is one of the biggest commodities’ producers and exporters, especially, regarding soybeans and corn. This segment has great influence on the international and national economy, as they are the main agricultural commodities traded in the world. However, due to globalization and other aspects inherent to agrobusiness, the agriculture has become an activity which presents several risks and uncertainties, influencing directly in the profitability of the enterprise. Facing that, the usage of tools to assess and minimize the risks is fundamental. In this view, this research aimed to develop a computational tool to simulate economic scenarios for the cultivation of soybeans and corn, in order to assist the producer in the decision-making task. For this, price forecasting models were developed using two different techniques, ARIMA and artificial neural networks. The forecasts and other information collected were consolidated into a single system which performed several simulations of economic scenarios which were analyzed using economic-financial and risk indicators. We found out that the models of price forecast using artificial neural networks were more accurate and presented a good predictive performance with MAPE between 1.5% and 4% and Willmott agreement index between 95% and 99%. Regarding the simulated scenarios, we concluded that the cultivation of soybeans has larger profitability with fewer chances of obtaining losses, when compared to the cultivation of corn. Besides that, it was observed that the risk management strategies proved themselves efficient, reducing the risks without compromising the profitability of the enterprise. The differential and the main contribution of this research was the development of a computer system for simulating economic scenarios of soybeans and corn cultivation, such tool showed great value as it allows the manager to estimate the profits of each cultivation and also to measure the financial risks involved, beyond assessing the impact of the usage of risk management strategies. All this group of information meaningfully supports the managers in the decision-making task.