主讲人Speaker: 李莹莹    讲座教授      香港科技大学

时间Date & Time: 2026年4月21日（周二），10:30--12:00

地点Venue:粤海校区汇星楼565会议室

内容简介/ Abstract:

This paper addresses the challenge of constructing efficient portfolios in a large investment universe composed exclusively of risky assets. We derive a linearly constrained regression representation of the efficient portfolio that circumvents the need to estimate the mean vector and covariance matrix. Instead, we apply constrained sparse regression techniques to estimate portfolio weights directly. Theoretically, we establish the asymptotic mean-variance efficiency of the estimated portfolio as both the number of assets and the sample size grow proportionally. In extensive simulations and empirical studies using S&P 500 constituents, our method yields portfolios that meet specified risk levels, achieve higher Sharpe ratios, and outperform various benchmarks, both gross and net of transaction costs. In addition, we present an extension that generates efficient tracking portfolios subject to a fixed tracking-error budget against an index.

主讲人介绍/Biography of the speaker：

李莹莹，香港科技大学金融系及ISOM系讲座教授、冯氏商学教授。博士毕业于芝加哥大学统计系。她曾任香港科技大学(广州)金融科技学域署理主任，也曾在普林斯顿大 Bendheim Center for Finance 及 Operations Research and Financial Engineering 系 担任讲师与博士后研究员。

李教授的研究方向包括:金融大数据的统计学习、大规模投资组合优化、个性化金融决策、高频金融数据、波动率估计与预测等。李教授在统计学、金融学与经济学领域，如Econometrica、Review of Financial Studies、Journal of Financial Economics、Annals of Statistics、Journal of American Statistical Association、Journal of Econometrics 等顶级期刊发表多篇论文。

李教授是国际金融计量学会会士(SoFiE Fellow)、Journal of Econometrics 会士/Fellow、现主持香港研资局高级研究学者(RGC Senior Research Fellow)项目、国家自然科学基金青年科学基金项目 (A类)。现任期刊Journal of American Statistical Association, Journal of Econometrics, Management Science的副主编。