Dr. Chao Zhang
张超
张超
Assistant Professor, FinTech Thrust at HKUST (GZ)
Research Interest: Machine Learning in Finance
News: I am currently looking for PhD students for 2026 and full-time RAs (always welcome), with a strong interest in FinTech research and a solid background in machine learning, finance, statistics, or related fields. If you are interested, please feel free to contact me at chao.zhang94@outlook.com.
Please note that I will not be accepting PhD applications for Fall 2025.
with Ruslan Goyenko, Bryan Kelly, Tobias Moskowitz, Yinan Su
Portfolio optimization chiefly focuses on risk and return prediction, yet implementation costs also play a critical role. Predicting trading costs is challenging, however, since costs depend endogenously on trade size and trader identity, thus impeding a generic solution. We focus on a key, yet general, component of trading costs that abstracts from these challenges -- trading volume. Individual stock trading volume is highly predictable, especially with machine learning. We model the economic benefits of predicting stock volume through a portfolio framework that trades off portfolio tracking error versus net-of-cost performance -- translating volume prediction into net-of-cost portfolio alpha. We find the benefits of predicting individual stock volume to be substantial, and potentially as large as those from stock return prediction.
We introduce a novel method to forecast the realised covariances of multidimensional time series. The central innovation of our method is a two-step procedure which first filters the training data by relevance before using non-parametric or linear approaches to make predictions. The relevance filter uses the recent histories of both the test input and each individual training input to generate a pairwise measure of dissimilarity between the two points. Highly dissimilar training data which do not share the observed dynamics of the test input are discarded from the regression procedure, reducing the complexity of the forecasting task. The second step uses simple non-parametric or linear methods to perform regression with the relevant data from the first step. Our method produces fast, interpretable predictions which achieve state-of-the-art performance, and functions well in limited-data availability and high dimensional settings such as large covariance matrices because it does not depend on highly parameterised models. Furthermore, we propose a dynamic ensembling scheme to combine our method with existing models. The scheme uses information about the recent performance of forecasts from different models to produce a weighted average forecast which heavily discounts models ill-suited to the current market dynamics. We test our methods on realised covariance data of U.S. equities and find significant improvements over widely used benchmarks.