报告人：黎德元

报告时间：2024年5月24日14:00-15:00

报告地点：慎思楼120教室

摘要：Marginal expected shortfall (MES) was proposed to measure the expected loss of an equity subject to a system-wide stress event (SWSE). The strengths and weaknesses of MES depend on the characterization of the SWSE. This paper defines an SWSE as the scenario in which either a representative index return or the worst-performed individual equity's return in a system drops below their respective prescribed thresholds. With this broadened SWSE definition, we advance MES to an innovative systemic and systematic risks-driven marginal expected shortfall (SYS2MES). We construct estimators for SYS2MES and establish their asymptotic theories within the multivariate extreme value theory framework, including tail-dependent and tail-independent scenarios, rendering its applicability to various models. The finite sample performance of the estimators is investigated in a simulation study. The SYS2MES can have broad applications in finance, environmental sciences, and social networks, amongst many application areas. The advantages of SYS2MES in risk management are demonstrated by studying its application to Dow Jones' 30 stocks, which yield better and more meaningful results than MES alone.

报告人简介：黎德元，复旦大学管理学院统计与数据科学系教授，博士生导师。1997年、2000年毕业于北京大学数学科学学院概率统计系，分别获得学士学位和硕士学位；2004年毕业于荷兰Erasmus大学经济学院，获得博士学位；2005年至2007年在瑞士伯尔尼大学统计学系做博士后、研究助理；2008年至今任教于复旦大学管理学院统计与数据科学系。研究方向：极值统计、分位数回归、分布式统计、风险管理。目前已在Annals of Statistics, JASA, Biometrika，Sinica, JBES, Journal of Economic Theory, Econometric Theory等统计学和经济学期刊上发表高水平学术论文40余篇，主持国家自然科学基金项目四项、教育部科研基金一项，海关总署百日攻关项目两项。目前担任中国现场统计研究会教育统计与管理分会副理事长。