Otimização de carteiras de commodities agropecuárias no mercado brasileiro como ferramenta de gestão de risco

Abstract

Commodity exports boost Brazilian economic performance, causing an increase in government tax revenues and serving as an important link with global markets. In 2015, UN data on trade and development revealed a dependence on commodities in emerging countries, showing that 2/3 of these, including Brazil, are dependent on raw material exports, with emphasis on agricultural commodities. The present study aimed to develop a portfolio of minimum variation, based on agricultural commodities and to compare it in terms of risk to other representative indices of the Brazilian financial market. The sample consisted of monthly observations available on the BM & FBOVESPA data base comprising the years 2004 to 2013, such as Beef Cattle, Coffee, Soy and Corn. The data from 2013 are no longer available, and the study used the series until a last available collection. The strategy adopted was divided in two periods: the first encompasses the observations until December 2010, the moment when the portfolio of minimal variation was presented by the methodology proposed by Markowitz (1952); The risk indicators and the monthly portfolio results were then analyzed, monthly reweighted, until the end of the sample in June 2013. We compared the results obtained in order to position it in terms of risk, with the IBOVESPA, CDI and ICB indicators, as well as a portfolio with equal weighting among the segments. The results indicated that in the portfolio until December 2010 Soja stake is dominant. With the rebalancing carried out after January 2011, it was verified that the updated minimum variance portfolio on a monthly basis represents a lower risk than the IBOVESPA and CDI indices, and its allocation can be used as a risk management strategy.