在随机波动模型的基础上, 结合交叉数据的特点, 提出了交叉数据随机波动模型, 综合对比与分析金融市场相关品种的价格波动关系.首先, 应用离散小波变换, 对数据进行滤波;其次, 采用伪极大似然估计方法 对系统模型进行参数 估计;然后, 运用遗传算法, 对各个子模型进行参数寻优;最后应用交叉数据随机波动模型对我国期货市场上大豆、豆油、豆粕三品种进行实证分析.实证表明, 交叉数据随机波动模型可以较理想地反映豆类三者之间价格波动的关系, 说 明相关品种之间存在较强的波动持续性.

On the basis of stochastic volatility model (SV), combining the characteristics of cross-data, stochastic volatility model with cross data (SV-CD) is presented to compare and analyze the relationships between price fluctuations of related species in financial markets. Firstly, discrete wavelet transform (DWT) is applied to filter market data; Secondly, the quasi-maximum likelihood estimation is used to Calculate parameter estimation; then, genetic algorithms (GA) are designed to search optimal parameter of each sub-model; Finally, application SV-CD is employed to empirically analyze soybeans, soybean oil, soybean meal in futures market. Empirical evidence shows that SV-CD can ideally reflect the relationship between price volatility of beans, indicating related species have a certain fluctuation persistence.