Modelos de previsão de preços aplicados aos contratos futuros agropecuários
Arquivos
Data
2001-02-04
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Viçosa
Resumo
Esta pesquisa trata da aplicabilidade de modelos de previsão de séries temporais como ferramenta de decisão de compra e venda de contratos futuros da BM&F, em datas próximas ao vencimento. Para fins empíricos, foram consideradas as commodities boi gordo, café e soja. O objetivo geral foi verificar qual modelo fornece as previsões mais precisas para cada série de preços considerada no mercado físico. O objetivo específico foi calcular os retornos médios de cada modelo em operações de compra e venda nos mercados futuros das commodities analisadas, de modo a fornecer um indicativo do potencial ou da limitação de cada um deles. Os modelos estudados foram os de Box & Jenkins (ARIMA), Redes Neurais, Estruturais e Bayesianos. Os dados utilizados corresponderam às cotações semanais de boi gordo, café e soja nos mercados físico e futuro. A discussão se baseou na hipótese de que esses modelos são instrumentos viáveis de auxílio à tomada de decisão por parte de agentes ligados ao agronegócio, reduzindo a incerteza quanto ao comportamento futuro dos preços. A análise foi conduzida, primeiramente, em termos de Erro Percentual de Previsão da série de preços do mercado físico para, em seguida, verificar os retornos em simulações de compra e venda de contratos futuros de cada produto, utilizando-se o Índice Sharpe, além do viés positivo ou negativo dessa média, através da estatística de simetria e do grau de dispersão dos retornos, medido pela curtose da distribuição destes. De modo geral, os resultados indicaram que: a) os modelos de previsão de séries temporais captam, de modo coerente, o padrão de comportamento dos preços analisados; b) há, contudo, diferenças de desempenho preditivo entre os modelos e entre cada mercado; e c) os retornos financeiros se mostraram positivos na maioria dos contratos analisados, indicando o potencial de utilização desses modelos em negociações de contratos para datas próximas ao vencimento, com destaque para operações fundamentadas nas previsões dos Modelos ARIMA e Estruturais.
This research deals with the usefulness of times series forecast models as a tool for buy and sell decisions of the brazilian BM&F future contracts, in dates nearby the expiration. For this purpose, the commodities considered were live cattle, coffee and soybeans. The general objective is to verify which model generates the most accurate forecasts for each price series of the considered commodities in the spot market. The specific objective is to calculate the medium returns of each model in buy and sell operations in each market of the analyzed commodities, in way to provide an indication of the potentials or limitations of each one.The models considered are the Box & Jenkins (ARIMA), Neural Networks, Structural and Bayesians time series models. The data utilized correspond to the weekly quotations of live cattle, coffee and soybeans in the spot and futures markets. The discussion is based on the hypothesis that those models are viable instruments to support decisions of economic agents participating in the agribussiness, reducing the uncertainty related to the future behavior of the spot prices. The analysis is carried out, firstly, in terms of Percentage Forecast Error for the price series in the spot market. Then, it verifies the returns in simulated buy and sell of future contracts of each product, using the Sharpe Index as a tool for comparsion, as well as the symmetry and kurtosis statistics. In general, the results indicate that: a) the time series forecast models capture coherently the pattern of the analyzed prices; b) there is, however, differences of forecast performance among the models and markets; and c) the financial returns are shown positive in most of the analyzed contracts, indicating the potential use of those models in negotiations of contracts for dates close to the expiration, with prominence for operations based in the forecasts of the ARIMA and Structural models.
This research deals with the usefulness of times series forecast models as a tool for buy and sell decisions of the brazilian BM&F future contracts, in dates nearby the expiration. For this purpose, the commodities considered were live cattle, coffee and soybeans. The general objective is to verify which model generates the most accurate forecasts for each price series of the considered commodities in the spot market. The specific objective is to calculate the medium returns of each model in buy and sell operations in each market of the analyzed commodities, in way to provide an indication of the potentials or limitations of each one.The models considered are the Box & Jenkins (ARIMA), Neural Networks, Structural and Bayesians time series models. The data utilized correspond to the weekly quotations of live cattle, coffee and soybeans in the spot and futures markets. The discussion is based on the hypothesis that those models are viable instruments to support decisions of economic agents participating in the agribussiness, reducing the uncertainty related to the future behavior of the spot prices. The analysis is carried out, firstly, in terms of Percentage Forecast Error for the price series in the spot market. Then, it verifies the returns in simulated buy and sell of future contracts of each product, using the Sharpe Index as a tool for comparsion, as well as the symmetry and kurtosis statistics. In general, the results indicate that: a) the time series forecast models capture coherently the pattern of the analyzed prices; b) there is, however, differences of forecast performance among the models and markets; and c) the financial returns are shown positive in most of the analyzed contracts, indicating the potential use of those models in negotiations of contracts for dates close to the expiration, with prominence for operations based in the forecasts of the ARIMA and Structural models.
Descrição
Palavras-chave
Modelos de Previsão univariados, Modelos ARIMA, Modelos Estruturais de séries temporais, Modelos Bayesianos univariados, Modelos de Redes Neurais Artificiais, Mercados futuros agropecuários
Citação
BRESSAN, Aureliano Angel. Modelos de previsão de preços aplicados aos contratos futuros agropecuários. 2001. 152 f. Tese (Doutorado em Economia Aplicada) - Universidade Federal de Viçosa, Viçosa. 2001.