新葡京官方

讲座时间：2025年6月19日12:30-14:00

讲座地点：明德主楼729

主讲人：王一如 匹兹堡大学经济系助理教授

主持人：王霞教授

讲座主题：Estimating Large Covariance Matrices with Unobserved Clustering

讲座简介：

In this paper,we address two challenges in estimating large covariance matrices:unobserved heterogeneity and the curse of dimensionality.We propose a novel approach that incorporates latent clustering into multivariate generalized autoregressive conditional heteroskedasticity(MGARCH)models.The approach assumes diagonal GARCH parameters follow an unobserved group structure,while the unconditional covariance matrix shares this group assignment,yielding an equal-block structure.We combine a composite likelihood approach with a Sequential Binary Segmentation Algorithm to effectively identify group memberships and estimate group-specific GARCH parameters.Empirically,we find that incorporating latent group structures signifiCantly improves minimum-variance portfolio performance,particularly during stable market periods when asset risk heterogeneity is more pronounced.

主讲人简介：

王一如，美国匹兹堡大学助理教授。2020年获西班牙庞培法布拉大学经济学博士学位，主要研究方向包括计量经济学，时问序列、宏观计量、面板数据、实证宏观等。其论文发表在Journal of Econometrics、Journal of Business &Economic Statistics、Stata Journal等期刊。