Ensaios sobre eficiência nos mercados agropecuários
Data
2015
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Escola Superior de Agricultura “Luiz de Queiroz” - Universidade de São Paulo
Resumo
A sinalização, formação e descoberta de preços agrícolas são adequadas se refletem rapidamente todas as informações recebidas pelos seus participantes. Então, quando o mercado é eficiente, possibilita eficiência alocativa, redução de imprecisão nas decisões dos agentes e dos custos informacionais. Entretanto, os agentes do agronegócio podem tomar decisões errôneas de produção, comercialização e estocagem, sujeitas ao conjunto de informações incompletas contidas nos preços passados, se os mercados forem não eficientes. Nesse contexto, o objetivo geral foi analisar a eficiência dos mercados futuros de commodities. Para atingi-lo, estruturou-se esta pesquisa em três ensaios. No primeiro, objetivou-se testar a hipótese de passeio aleatório a contratos futuros agropecuários negociados na BM&FBOVESPA. Refutá-la significa possível previsibilidade e, por conseguinte, os mercados não seriam fracamente eficientes. Correlações seriais e testes de razão de variância foram utilizados para verificá-las. Os resultados deram suporte à hipótese de passeio aleatório nos mercados futuros de café e da soja, eficientes na forma fraca, e evidências contrárias foram encontradas nos mercados do boi gordo, milho e etanol. No segundo, o objetivo foi investigar a eficiência e formações de clusters nos contratos futuros do complexo soja (soja, farelo de soja e óleo de soja) negociados nas bolsas de commodities: argentina, brasileira, chinesa, indiana, japonesa, norte-americana e sul-africana. Com base na métrica obtida por distância euclidiana de razões de variância, evidenciaram-se dependências similares dos mercados, as quais podem ser interpretadas como efeito espraiamento da eficiência informacional. Os agentes devem, portanto, manter percepções em relação aos diversos mercados devido às sinalizações interdependentes dos preços. No terceiro, objetivou-se analisar a eficiência dos mercados futuros agropecuários brasileiros, sob a hipótese adaptativa de mercado. Utilizando propostas recentes à não linearidade e razão de variância, encontrou-se que as elevadas rejeições à hipótese de diferença martingal se encontram nos mercados em que as intervenções governa- mentais se fazem presentes: milho e etanol. Nos mercados de café, boi gordo e soja ocorreram menores rejeições à hipótese martingal e, portanto, houve maior eficiência informacional. Essas evidências — consistentes com a hipótese adaptativa dos mercados — justificam operações de hedge dinâmicas, bem como a gerência de carteiras de investimentos de forma ativa.
Agricultural prices’ formation, discovery and signalling only are accurate when they can rapidly reflect all new information faced by its market agents. Thereby, when a given market is efficient, it allows for allocative efficiency, reducing inefficiencies both in decision-making process and in informational costs. On the other hand, when markets are said not to be efficient, agribusinesses’ agents can make mistaken production, marketing and storage decisions, once such decisions are due to incomplete information contained in past prices. In this context, the main purpose of this study is to analyze the efficiency in future markets of commodities. In order to achieve its final goal, the study has been structured in three essays. In the first essay, the random walk hypothesis has been tested for agricultural future contracts from Brazilian Securities, Commodities and Futures Exchange (BMF&BOVESPA). Refusing the hypothesis for a given commodity implies some degree of predictability, therefore inconsistent even with a weak notion of efficiency. These tests were carried out using serial correlations and variance ratios. The results show the presence of random walks in coffee and soybean future markets, and contrary evidences (absence of random walks) in live cattle, corn and ethanol future markets. In the second essay, it has been analyzed the efficiency and the presence of clusters in the soybean complex future contracts (soybean, soybean meal and soybean oil) traded in the following future markets: Argentina (MTB), Brazil (BVMF), China (DCE), India (NCD), Japan (TKT), US (CBT) and South-Africa (SAF). Based on the metrics obtained by Euclidian distances of variance ratios, similar dependencies have been found for all markets, which suggest informational efficiency spreading. Agents, therefore, shall maintain perceptions over several international markets, given the interdependence found for prices in distinct future markets. In the third one, the adaptive market hypothesis has been tested for agricultural future markets in Brazil. Applying more recent approaches to Nonlinearity and Variance Ratio tests, high rejections to martingale difference hypothesis took place in agricultural markets which governmental interference is highly persistent, i.e., corn and ethanol. In coffee, live cattle and soybeans markets, weaker rejections to the martingale hypothesis hint higher informational efficiency. These evidences, consistent with the adaptive market hypothesis, justify dynamic hedge operations, as well as an active management of investment portfolios in such markets.
Agricultural prices’ formation, discovery and signalling only are accurate when they can rapidly reflect all new information faced by its market agents. Thereby, when a given market is efficient, it allows for allocative efficiency, reducing inefficiencies both in decision-making process and in informational costs. On the other hand, when markets are said not to be efficient, agribusinesses’ agents can make mistaken production, marketing and storage decisions, once such decisions are due to incomplete information contained in past prices. In this context, the main purpose of this study is to analyze the efficiency in future markets of commodities. In order to achieve its final goal, the study has been structured in three essays. In the first essay, the random walk hypothesis has been tested for agricultural future contracts from Brazilian Securities, Commodities and Futures Exchange (BMF&BOVESPA). Refusing the hypothesis for a given commodity implies some degree of predictability, therefore inconsistent even with a weak notion of efficiency. These tests were carried out using serial correlations and variance ratios. The results show the presence of random walks in coffee and soybean future markets, and contrary evidences (absence of random walks) in live cattle, corn and ethanol future markets. In the second essay, it has been analyzed the efficiency and the presence of clusters in the soybean complex future contracts (soybean, soybean meal and soybean oil) traded in the following future markets: Argentina (MTB), Brazil (BVMF), China (DCE), India (NCD), Japan (TKT), US (CBT) and South-Africa (SAF). Based on the metrics obtained by Euclidian distances of variance ratios, similar dependencies have been found for all markets, which suggest informational efficiency spreading. Agents, therefore, shall maintain perceptions over several international markets, given the interdependence found for prices in distinct future markets. In the third one, the adaptive market hypothesis has been tested for agricultural future markets in Brazil. Applying more recent approaches to Nonlinearity and Variance Ratio tests, high rejections to martingale difference hypothesis took place in agricultural markets which governmental interference is highly persistent, i.e., corn and ethanol. In coffee, live cattle and soybeans markets, weaker rejections to the martingale hypothesis hint higher informational efficiency. These evidences, consistent with the adaptive market hypothesis, justify dynamic hedge operations, as well as an active management of investment portfolios in such markets.
Descrição
Tese de doutorado defendida na Escola Superior de Agricultura “Luiz de Queiroz” - Universidade de São Paulo
Palavras-chave
Hipótese de eficiência de mercado, Hipótese adaptativa de mercado
Citação
RODRIGUES, M. A. Ensaios sobre eficiência nos mercados agropecuários. 2015. 91 f. Tese (Doutorado em Economia Aplicada) - Escola Superior de Agricultura “Luiz de Queiroz”, Universidade de São Paulo, Piracicaba. 2015.