In response to empirical evidence, we propose a continuous-time model for multivariate asset returns with a two-layered dependence structure. The price process is subject to multivariate information arrivals driving the market activity modeled by nondecreasing pure-jump Lévy processes. A Lévy copula determines the jump dependence and allows for a generic multivariate information flow with a flexible structure. Conditional on the information flow, asset returns are jointly normal. Within this setup, we provide an estimation framework based on maximum simulated likelihood. We apply novel multivariate models to equity data and obtain estimates which meet an economic intuition with respect to the two-layered dependence structure.

中文翻译：

---

金融市场中的信息流依赖性

针对经验证据，我们提出了一个具有两层依赖结构的多元资产收益的连续时间模型。价格过程受多变量信息到达驱动市场活动的影响，该市场活动由非减纯跳跃 Lévy 过程建模。Lévy copula 确定了跳跃依赖，并允许具有灵活结构的通用多元信息流。在信息流通的条件下，资产收益是共同正常的。在此设置中，我们提供了一个基于最大模拟似然的估计框架。我们将新颖的多元模型应用于股票数据，并获得符合关于两层依赖结构的经济直觉的估计。