时间:2016年12月7日(星期三)09:30-11:30
地点:广外南校区教学楼B304室
主办:
·betway体育
·“投资管理、期权定价和风险管理”科研创新团队
主讲人:
·09:30-10:30孔荫莹(广东财经大学教授)
·10:30-11:30赵钊(瑞士苏黎世大学与华中科技大学经济学联合培养博士)
主题:大数据hadoop平台在智慧校园和股票配对的应用(孔荫莹)
内容简介:
介绍大数据hadoop平台在智慧校园和股票配对的应用,包括:1)了解大数据hadoop平台构造的必要性;2)了解Hadoop平台原理和优点、基本原理、安装与运行情况; 3)介绍了解Hadoop平台在智慧校园的一些应用。4)平台对股票配对的想法。
Beyond Sorting: A More Powerful Test for Cross-Sectional Anomalies(赵钊)
内容简介:
Many researchers seek factors that predict the cross-section of stock returns. The standard methodology sorts stocks according to their factor scores into quantiles (terciles, quintiles, deciles, etc), then forms a portfolio that goes long the top quantile and short the bottom quantile, and aims to demonstrate that its out-of-sample performance is statistically significant. This procedure does not use any covariance matrix. Within the framework of Markowitz portfolio optimization, this omission is mathematically equivalent to assuming that the best estimator of the population covariance matrix is the identity matrix. Such an assumption may have been a reasonable course of action in the past, at least in large-dimensional settings, but recent advances in covariance matrix estimation have given birth to estimators that are capable of beating the identity matrix significantly. Two developments have been particularly impactful: (1) acknowledging the time-varying nature of the covariance matrix through a model such as Dynamic Conditional Correlation (DCC); and (2)fixing the distortions that systematically appear in large-dimensional settings by applying a transformation such as NonLinear (NL) shrinkage to the eigenvalues of the sample covariance matrix. We demonstrate that using the combined DCC-NL estimator of the covariance matrix substantially enhances the power of tests to detect factors that predict the cross-section of stock returns.
主讲人简介:
孔荫莹,广东财经大学教授,统计学硕导,现任广东财经大学数量经济研究中心主任,研究领域为大数据Hadoop 架构研究;单复变函数值分布;随机级数和Laplace-Stieltjes变换增长性;随机分支游动;人民币国际化。2008年聘为德国《数学文摘》的评论员,已评论文章70多篇。2016年聘为Mathematical Reviews评论员。《Complex Variables and Elliptic Equations》,《Chinese Annals of Mathematics》, 《Acta Mathematica Scientia》,《Turk J. Math.》,《数学学报》,《数学物理学报》审稿人。广州市科技局,广东省科学计划项目评审会会评专家;江西科技成果奖评审专家。主持完成国家项目1项,主持省自科基金自由申请项目2项,主持省级教改项目3项,主持大数据横向项目1项。2010年10公派到法国南布列塔里大学数学实验室进行一年的博士后工作,合作导师Quansheng LIU(刘全升)教授,研究分支过程及随机环境中的分支过程和分支随机游动,开始从事大数据研究。在《J. Math. Anal. Appl》、《数学学报》等期刊发表30多篇学术论文。个人主页:http://shx.gdufe.edu.cn/content.aspx?id=293540135131
最近,孔教授与美国NRG 电力公司欧阳永健工程师合作已经在广东财经大学经管实验教学中心搭建大数据Hadoop 平台实验室,这与美国NRG 电力公司数据处理中心是一样的架构;并完成调试以及运行,现在进行云储存及计算,电力、实时信号图形和证券等超大数据处理工作,为学校和企业提供大数据处理技术。
赵钊,瑞士苏黎世大学与华中科技大学经济学联合培养博士,研究领域为高维资产组合配置、资产定价。