Research at FAST

82 Department of Applied Mathematics Department of Applied Mathematics Qualification PhD (HKUST) ORCID ID 0000-0002-3656-9966 Dr DING Yi Research Assistant Professor Research Overview My research focuses on developing statistical methodology and theory to better understand financial markets. My area of interest includes Fintech, high-dimensional statistics, portfolio optimization, volatility modeling, high-frequency financial data, statistical machine learning, and asset allocation. Here are some recent results: High-dimensional portfolio optimization: We propose a high dimensional minimum variance portfolio (MVP) estimator under statistical factor models. Our proposed MVP estimator enjoys desirable risk ratio consistency in the sense that the ratio between the risk of the portfolio estimator and the theoretical minimum risk converges to one. The convergence properties are established under scenarios where the minimum risk either decays to zero as the number of assets increases or is bounded from below. Statistical learning for individualized asset allocation: We establish a statistical learning framework for personalized asset allocation. Methodologically, we propose is High-dimensional Q-learning for continuous decision making. The proposed approach addresses the continuous decision making problem in high-dimensional setting, we show that it enjoys desirable statistical properties, including proper inference for the optimal value function. In the fields of economics and finance, as a pioneer work, we study the individualized asset allocation problem using our proposed framework. High-frequency & High-dimensional volatility modeling: We investigate the cross-sectional properties of volatility and idiosyncratic volatility using high-frequency data under high-dimensional setting. Explicit conditions is given for the consistency of conducting principal component analysis on realized volatilities in identifying the factor structure in volatility. Empirically, strong empirical evidence from high frequency data of S&P 500 Index constituents, we propose a multiplicative single factor model for stock volatility. Representative Publication • Y. Ding , Y. Li and X. Zheng, High dimensional minimum variance portfolio estimation under statistical factor models, Journal of Econometrics, 2021 , 222(1): 502-515