摘要：Users of agricultural markets always need to establish accurate representations of future volatility. This paper investigates the properties of realized volatility in the soybean futures market. The results indicate that the distributional properties of realized volatility based on 5-minute returns largely correspond with existing literature. The findings of three volatility measures confirm that the Mixture of Distributions Hypothesis (MDH) is valid. In contrast, the standardized daily returns display some different properties compared with stock and exchange rate data. Moreover, the parametric ARFIMA and GARCH models reflect same patterns as described in nonparametric analysis.
