报告人： Yongmiao Hong，Chinese Academy of Sciences
时间：2021-04-29 14:00-15:30
地点：Guanghua Building No.2, Room 217
Abstract:
In this paper we propose a novel machine learning approach to estimating a class of high dimensional multivariate GARCH models, which preserve a channel for lagged values and past innovations to affect dynamic volatility. We propose a nuclear norm penalized constrained quasi-maximum likelihood estimation method, which is a one-step procedure and guarantees the positive definiteness property of the estimated volatility process, both asymptotically and in finite samples, with a fast rate of convergence. To ease its implementation, we further develop an ADMM algorithm and derive a generalized cross-validation criterion to optimally select the tuning parameter in practice. Simulation studies show that our proposed estimator provides superior finite sample performance over some existing dynamic variance estimators in the literature. We apply the proposed estimation method to vast portfolio selections, which enjoys significantly enhanced out-of-sample Sharpe ratios with largely reduced portfolio risks.
About the Speaker:
Yongmiao Hong, Center for Forecasting Science, Chinese Academy of Sciences, and School of Economics and Management, University of Chinese Academy of Sciences. This paper is coauthored with Liyuan Cui and Junbo Wang at City University of Hong Kong.
Yongmiao Hong is a Fellow of The World Academy of Sciences (TWAS) for advancement of science in developing countries, a Fellow of the Econometric Society, and a Senior Fellow of the Rimini Center for Economic Analysis (RCEA). He is a Distinguished Research Fellow in the Academy of Mathematics and Systems Science at Chinese Academy of Sciences, and a Distinguished Professor in the School of Economics and Management at University of Chinese Academy of Sciences. He publishes in Annals of Statistics, Biometrika, Journal of American Statistical Association, Journal of Royal Statistical Society Series B, Econometrica, Journal of Political Economy, Quarterly Journal of Economics, Review of Economic Studies, and Review of Financial Studies.
Your participation is warmly welcomed!
欢迎扫码关注北大统计科学中心公众号，了解更多讲座信息
