Chao Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Chao Zhang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Mihai Cucuringu
Department of Mathematics, University of California, Los Angeles, Los Angeles, USA
5 shared publications Mihai Cucuringu
Oxford-Man Institute of Quantitative Finance, University of Oxford , Oxford
3 shared publications- Rama Cont
Mathematical Institute, University of Oxford, Oxford, UK
2 shared publications Xiaowen Dong
Oxford-Man Institute of Quantitative Finance, University of Oxford , Oxford
2 shared publicationsXingyue Pu
Oxford-Man Institute of Quantitative Finance, University of Oxford , Oxford
2 shared publicationsYihuang Zhang
Mathematical Institute, University of Oxford , Oxford, UK
2 shared publications
Latest Publications
- Top Cited Papers
- 1 April 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 72 (4) , 2917-2936
T ail -GAN: Learning to Simulate Tail Risk Scenarios
The estimation of loss distributions for dynamic portfolios requires the simulation of scenarios representing realistic joint dynamics of their components. We propose a novel data-driven approach for simulating realistic, high-dimensional multiasset scenarios, focusing on accurately representing tail risk for a class of static and dynamic trading strategies. We exploit the joint elicitability property of Value-at-Risk and Expected Shortfall to design a Generative Adversarial Network that learns to simulate price scenarios preserving these tail risk features. We demonstrate the performance of our algorithm on synthetic and market data sets through detailed numerical experiments. In contrast to previously proposed data-driven scenario generators, our proposed method correctly captures tail risk for a broad class of trading strategies and demonstrates strong generalization capabilities. In addition, combining our method with principal component analysis of the input data enhances its scalability to large-dimensional multiasset time series, setting our framework apart from the univariate settings commonly considered in the literature. This paper was accepted by Kay Giesecke, finance. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.00936 .
- 1 April 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 72 (4) , 2917-2936
Top Cited PapersT ail -GAN: Learning to Simulate Tail Risk Scenarios
The estimation of loss distributions for dynamic portfolios requires the simulation of scenarios representing realistic joint dynamics of their components. We propose a novel data-driven approach for simulating realistic, high-dimensional multiasset scenarios, focusing on accurately representing tail risk for a class of static and dynamic trading strategies. We exploit the joint elicitability property of Value-at-Risk and Expected Shortfall to design a Generative Adversarial Network that learns to simulate price scenarios preserving these tail risk features. We demonstrate the performance of our algorithm on synthetic and market data sets through detailed numerical experiments. In contrast to previously proposed data-driven scenario generators, our proposed method correctly captures tail risk for a broad class of trading strategies and demonstrates strong generalization capabilities. In addition, combining our method with principal component analysis of the input data enhances its scalability to large-dimensional multiasset time series, setting our framework apart from the univariate settings commonly considered in the literature. This paper was accepted by Kay Giesecke, finance. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.00936 .
- Open Access
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 422-430
DeltaLag: Learning Dynamic Lead-Lag Patterns in Financial Markets
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 422-430
Open AccessDeltaLag: Learning Dynamic Lead-Lag Patterns in Financial Markets
- Open Access
- 9 November 2024
- Published by Oxford University Press (OUP) in Journal of Financial Econometrics
- Vol. 23 (2)
Graph-Based Methods for Forecasting Realized Covariances
We forecast the realized covariance matrix of asset returns in the U.S. equity market by exploiting the predictive information of graphs in volatility and correlation. Specifically, we augment the Heterogeneous Autoregressive model via neighborhood aggregation on these graphs. Our proposed method allows for the modeling of interdependence in volatility (also known as spillover effect) and correlation, while maintaining parsimony and interpretability. We explore various graph construction methods, including sector membership and graphical LASSO (for modeling volatility), and line graph (for modeling correlation). The results generally suggest that the augmented model incorporating graph information yields both statistically and economically significant improvements for out-of-sample performance over the traditional models. Such improvements remain significant over horizons up to 1 month ahead, but decay in time. The robustness tests demonstrate that the forecast improvements are obtained consistently over the different out-of-sample sub-periods and are insensitive to measurement errors of volatilities.
- 9 November 2024
- Published by Oxford University Press (OUP) in Journal of Financial Econometrics
- Vol. 23 (2)
Open AccessGraph-Based Methods for Forecasting Realized Covariances
We forecast the realized covariance matrix of asset returns in the U.S. equity market by exploiting the predictive information of graphs in volatility and correlation. Specifically, we augment the Heterogeneous Autoregressive model via neighborhood aggregation on these graphs. Our proposed method allows for the modeling of interdependence in volatility (also known as spillover effect) and correlation, while maintaining parsimony and interpretability. We explore various graph construction methods, including sector membership and graphical LASSO (for modeling volatility), and line graph (for modeling correlation). The results generally suggest that the augmented model incorporating graph information yields both statistically and economically significant improvements for out-of-sample performance over the traditional models. Such improvements remain significant over horizons up to 1 month ahead, but decay in time. The robustness tests demonstrate that the forecast improvements are obtained consistently over the different out-of-sample sub-periods and are insensitive to measurement errors of volatilities.
- 7 October 2024
- Published by Elsevier in International Journal of Forecasting
- Vol. 41 (1) , 377-397
Forecasting realized volatility with spillover effects: Perspectives from graph neural networks
- 7 October 2024
- Published by Elsevier in International Journal of Forecasting
- Vol. 41 (1) , 377-397
Forecasting realized volatility with spillover effects: Perspectives from graph neural networks
- Open Access
- 25 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 436-444
Cryptocurrency volatility forecasting using commonality in intraday volatility
We investigate the benefits of using intraday realized volatility (RV) commonality, and propose a novel non-parametric framework for forecasting one-day ahead intraday RV (1D-ahead intraday RV). Specifically, we train multiple models using machine learning (ML) techniques under various training settings (single-asset, cluster-driven, and cross-asset), where commonality gradually enters model dynamics as training schemes become more complex. We conclude that models that leverage the cryptocurrency commonality outperform models that do not explicitly account for it, regardless of the market regime considered. The source code of this project is available at: github.com/edjanga/crypto_volatility_commonality.
- 25 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 436-444
Open AccessCryptocurrency volatility forecasting using commonality in intraday volatility
We investigate the benefits of using intraday realized volatility (RV) commonality, and propose a novel non-parametric framework for forecasting one-day ahead intraday RV (1D-ahead intraday RV). Specifically, we train multiple models using machine learning (ML) techniques under various training settings (single-asset, cluster-driven, and cross-asset), where commonality gradually enters model dynamics as training schemes become more complex. We conclude that models that leverage the cryptocurrency commonality outperform models that do not explicitly account for it, regardless of the market regime considered. The source code of this project is available at: github.com/edjanga/crypto_volatility_commonality.
- Open Access
- 17 August 2023
- Published by Taylor & Francis in Quantitative Finance
- Vol. 23 (10) , 1373-1393
Cross-impact of order flow imbalance in equity markets
- 17 August 2023
- Published by Taylor & Francis in Quantitative Finance
- Vol. 23 (10) , 1373-1393
Open AccessCross-impact of order flow imbalance in equity markets
- Open Access
- 20 March 2023
- Published by Oxford University Press (OUP) in Journal of Financial Econometrics
- Vol. 22 (2) , 492-530
Volatility Forecasting with Machine Learning and Intraday Commonality
We apply machine learning models to forecast intraday realized volatility (RV), by exploiting commonality in intraday volatility via pooling stock data together, and by incorporating a proxy for the market volatility. Neural networks dominate linear regressions and tree-based models in terms of performance, due to their ability to uncover and model complex latent interactions among variables. Our findings remain robust when we apply trained models to new stocks that have not been included in the training set, thus providing new empirical evidence for a universal volatility mechanism among stocks. Finally, we propose a new approach to forecasting 1-day-ahead RVs using past intraday RVs as predictors, and highlight interesting time-of-day effects that aid the forecasting mechanism. The results demonstrate that the proposed methodology yields superior out-of-sample forecasts over a strong set of traditional baselines that only rely on past daily RVs.
- 20 March 2023
- Published by Oxford University Press (OUP) in Journal of Financial Econometrics
- Vol. 22 (2) , 492-530
Open AccessVolatility Forecasting with Machine Learning and Intraday Commonality
We apply machine learning models to forecast intraday realized volatility (RV), by exploiting commonality in intraday volatility via pooling stock data together, and by incorporating a proxy for the market volatility. Neural networks dominate linear regressions and tree-based models in terms of performance, due to their ability to uncover and model complex latent interactions among variables. Our findings remain robust when we apply trained models to new stocks that have not been included in the training set, thus providing new empirical evidence for a universal volatility mechanism among stocks. Finally, we propose a new approach to forecasting 1-day-ahead RVs using past intraday RVs as predictors, and highlight interesting time-of-day effects that aid the forecasting mechanism. The results demonstrate that the proposed methodology yields superior out-of-sample forecasts over a strong set of traditional baselines that only rely on past daily RVs.
- Open Access
Volatility forecasting with machine learning and intraday commonality
We apply machine learning models to forecast intraday realized volatility (RV), by exploiting commonality in intraday volatility via pooling stock data together, and by incorporating a proxy for the market volatility. Neural networks dominate linear regressions and tree-based models in terms of performance, due to their ability to uncover and model complex latent interactions among variables. Our findings remain robust when we apply trained models to new stocks that have not been included in the training set, thus providing new empirical evidence for a universal volatility mechanism among stocks. Finally, we propose a new approach to forecasting one-day-ahead RVs using past intraday RVs as predictors, and highlight interesting time-of-day effects that aid the forecasting mechanism. The results demonstrate that the proposed methodology yields superior out-of-sample forecasts over a strong set of traditional baselines that only rely on past daily RVs.
Open AccessVolatility forecasting with machine learning and intraday commonality
We apply machine learning models to forecast intraday realized volatility (RV), by exploiting commonality in intraday volatility via pooling stock data together, and by incorporating a proxy for the market volatility. Neural networks dominate linear regressions and tree-based models in terms of performance, due to their ability to uncover and model complex latent interactions among variables. Our findings remain robust when we apply trained models to new stocks that have not been included in the training set, thus providing new empirical evidence for a universal volatility mechanism among stocks. Finally, we propose a new approach to forecasting one-day-ahead RVs using past intraday RVs as predictors, and highlight interesting time-of-day effects that aid the forecasting mechanism. The results demonstrate that the proposed methodology yields superior out-of-sample forecasts over a strong set of traditional baselines that only rely on past daily RVs.