Yubo Tao
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
Yubo Tao has a total of 29 co-authors. Here are the top co-authors by number of shared publications:
- Jun Yu
Singapore Management University
2 shared publications Li Guo
School of Economics, Fudan University, Shanghai, China
2 shared publications- Chengzhi Yang
Aviation University of Air Force, Changchun 130022
1 shared publication Dexin Zhou
1 shared publication- Hao Chen
China University of Mining and Technology
1 shared publication - Hunter Ng
Zicklin School of Business Baruch College The City University of New York New York New York USA
1 shared publication
Latest Publications
- Open Access
- 23 October 2026
- Published by Elsevier in SSRN Electronic Journal
Headline Emphasis and Stock Returns: Evidence from Counterfactual Headlines
Financial news headlines emphasize selected aspects of a broader story. We study the pricing implications of this choice by comparing published headlines with counterfactual headline distributions generated from the same articles using large language models. Among historical S&P 500 constituents during 2000-2024, a standard-deviation increase in headline divergence predicts 9 basis points higher next-month returns. Greater divergence is associated with an initial price decline followed by recovery over subsequent weeks. Predictability is stronger for longer articles and firms with less news coverage. The findings support a price-correction interpretation in which investors initially overweight the facts highlighted in headlines and gradually incorporate the article's remaining information.
- 23 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessHeadline Emphasis and Stock Returns: Evidence from Counterfactual Headlines
Financial news headlines emphasize selected aspects of a broader story. We study the pricing implications of this choice by comparing published headlines with counterfactual headline distributions generated from the same articles using large language models. Among historical S&P 500 constituents during 2000-2024, a standard-deviation increase in headline divergence predicts 9 basis points higher next-month returns. Greater divergence is associated with an initial price decline followed by recovery over subsequent weeks. Predictability is stronger for longer articles and firms with less news coverage. The findings support a price-correction interpretation in which investors initially overweight the facts highlighted in headlines and gradually incorporate the article's remaining information.
- Open Access
- 1 July 2025
- Published by Elsevier in SSRN Electronic Journal
Price Discovery and Trading in Modern Prediction Markets
This study provides the first evidence on price discovery dynamics across modern prediction markets. We study a unique dataset of common contracts traded on leading prediction markets, Polymarket, Kalshi, PredictIt, and Robinhood, during the period leading up to the 2024 U.S. presidential election. We find that more liquid prediction markets substantially outperform polls in predicting subsequent election results, yet there are significant price disparities across platforms. Polymarket leads Kalshi in price discovery, particularly when liquidity and trading activity are high, implying economically meaningful arbitrage opportunities. Moreover, net order imbalance from large trades strongly predicts subsequent returns, and the market experiencing greater directional order flow from large trades tends to lead price discovery. These results underscore how platform structure, liquidity, and informed trading interact to shape trading and prices.
- 1 July 2025
- Published by Elsevier in SSRN Electronic Journal
Open AccessPrice Discovery and Trading in Modern Prediction Markets
This study provides the first evidence on price discovery dynamics across modern prediction markets. We study a unique dataset of common contracts traded on leading prediction markets, Polymarket, Kalshi, PredictIt, and Robinhood, during the period leading up to the 2024 U.S. presidential election. We find that more liquid prediction markets substantially outperform polls in predicting subsequent election results, yet there are significant price disparities across platforms. Polymarket leads Kalshi in price discovery, particularly when liquidity and trading activity are high, implying economically meaningful arbitrage opportunities. Moreover, net order imbalance from large trades strongly predicts subsequent returns, and the market experiencing greater directional order flow from large trades tends to lead price discovery. These results underscore how platform structure, liquidity, and informed trading interact to shape trading and prices.
- 1 April 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 70 (4) , 2122-2143
Financialization and Commodity Markets Serial Dependence
Recent financialization in commodity markets makes it easier for institutional investors to trade a portfolio of commodities via various commodity-indexed products. We present several pieces of novel causal evidence that daily exposure to such index trading results in price overshoots and reversals, as reflected in negative daily return autocorrelations, only among commodities in that index. This is because index trading propagates nonfundamental noise to all indexed commodities. We present direct evidence for such noise propagation using commodity news sentiment data. This paper was accepted by Bruno Biais, finance. Funding: Z. Da acknowledges financial support from the Beijing Outstanding Young Scientist Program [Grant BJJWZYJH01201910034034] and the 111 Project [Grant B20094]. K. Tang acknowledges financial support from the National Natural Science Foundation of China [Grants 71973075 and 72192802]. Y. Tao acknowledges financial support from the Start-up Research Grant of University of Macau [Grant SRG2022-00016-FSS]. L. Yang acknowledges the Social Sciences and Humanities Research Council of Canada for financial support [Grants 430-2018-00173 and 435-2021-0040]. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2023.4797 .
- 1 April 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 70 (4) , 2122-2143
Financialization and Commodity Markets Serial Dependence
Recent financialization in commodity markets makes it easier for institutional investors to trade a portfolio of commodities via various commodity-indexed products. We present several pieces of novel causal evidence that daily exposure to such index trading results in price overshoots and reversals, as reflected in negative daily return autocorrelations, only among commodities in that index. This is because index trading propagates nonfundamental noise to all indexed commodities. We present direct evidence for such noise propagation using commodity news sentiment data. This paper was accepted by Bruno Biais, finance. Funding: Z. Da acknowledges financial support from the Beijing Outstanding Young Scientist Program [Grant BJJWZYJH01201910034034] and the 111 Project [Grant B20094]. K. Tang acknowledges financial support from the National Natural Science Foundation of China [Grants 71973075 and 72192802]. Y. Tao acknowledges financial support from the Start-up Research Grant of University of Macau [Grant SRG2022-00016-FSS]. L. Yang acknowledges the Social Sciences and Humanities Research Council of Canada for financial support [Grants 430-2018-00173 and 435-2021-0040]. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2023.4797 .
- 24 April 2023
- Published by Elsevier in Economic Modelling
- Vol. 124
Trend-based forecast of cryptocurrency returns
- 24 April 2023
- Published by Elsevier in Economic Modelling
- Vol. 124
Trend-based forecast of cryptocurrency returns
- 12 December 2022
- Published by Taylor & Francis in Journal of Business & Economic Statistics
- Vol. 42 (2) , 437-456
A Time-Varying Network for Cryptocurrencies
Cryptocurrencies return cross-predictability and technological similarity yield information on risk propagation and market segmentation. To investigate these effects, we build a time-varying network for cryptocurrencies, based on the evolution of return cross-predictability and technological similarities. We develop a dynamic covariate-assisted spectral clustering method to consistently estimate the latent community structure of cryptocurrencies network that accounts for both sets of information. We demonstrate that investors can achieve better risk diversification by investing in cryptocurrencies from different communities. A cross-sectional portfolio that implements an inter-crypto momentum trading strategy earns a 1.08% daily return. By dissecting the portfolio returns on behavioral factors, we confirm that our results are not driven by behavioral mechanisms.
- 12 December 2022
- Published by Taylor & Francis in Journal of Business & Economic Statistics
- Vol. 42 (2) , 437-456
A Time-Varying Network for Cryptocurrencies
Cryptocurrencies return cross-predictability and technological similarity yield information on risk propagation and market segmentation. To investigate these effects, we build a time-varying network for cryptocurrencies, based on the evolution of return cross-predictability and technological similarities. We develop a dynamic covariate-assisted spectral clustering method to consistently estimate the latent community structure of cryptocurrencies network that accounts for both sets of information. We demonstrate that investors can achieve better risk diversification by investing in cryptocurrencies from different communities. A cross-sectional portfolio that implements an inter-crypto momentum trading strategy earns a 1.08% daily return. By dissecting the portfolio returns on behavioral factors, we confirm that our results are not driven by behavioral mechanisms.
- Open Access
Joint News, Attention Spillover,and Market Returns
We analyze over 2.6 million news articles and propose a novel measure of aggregate joint news coverage of firms. The measure strongly and negatively predicts market returns, both in sample and out of sample. The relation is causal, robust to existing predictors, and is especially strong when market uncertainty is high or when market frictions are large. Using data on EDGAR downloads by unique IPs, we provide direct evidence that joint news triggers attention spillover across firms. Our results are consistent with the explanation that joint news generates a contagion in investor attention and causes marketwide overvaluations and subsequent reversals.
Open AccessJoint News, Attention Spillover,and Market Returns
We analyze over 2.6 million news articles and propose a novel measure of aggregate joint news coverage of firms. The measure strongly and negatively predicts market returns, both in sample and out of sample. The relation is causal, robust to existing predictors, and is especially strong when market uncertainty is high or when market frictions are large. Using data on EDGAR downloads by unique IPs, we provide direct evidence that joint news triggers attention spillover across firms. Our results are consistent with the explanation that joint news generates a contagion in investor attention and causes marketwide overvaluations and subsequent reversals.
- Open Access
- 19 September 2022
- Published by Elsevier in SSRN Electronic Journal
Trend-based Forecast of Cryptocurrency Returns
We systematically re-examine the efficacy of trend-based technical indicators in predicting cryptocurrency market returns at daily, weekly, and monthly horizons. It shows that the price-based signals are more effective than the volume-based signals in the short horizon (daily and weekly), while the volume-based signals are more powerful in the long horizon (monthly). We also document that machine learning techniques can significantly improve the performance of technical indicators both in and out of sample at all horizons. Further analysis reveals that leading cryptos are more predictable by technical analysis, and technical indicators based on different information respond differently to the COVID-19 outbreak.
- 19 September 2022
- Published by Elsevier in SSRN Electronic Journal
Open AccessTrend-based Forecast of Cryptocurrency Returns
We systematically re-examine the efficacy of trend-based technical indicators in predicting cryptocurrency market returns at daily, weekly, and monthly horizons. It shows that the price-based signals are more effective than the volume-based signals in the short horizon (daily and weekly), while the volume-based signals are more powerful in the long horizon (monthly). We also document that machine learning techniques can significantly improve the performance of technical indicators both in and out of sample at all horizons. Further analysis reveals that leading cryptos are more predictable by technical analysis, and technical indicators based on different information respond differently to the COVID-19 outbreak.
- Open Access
Regression-Adjusted Estimation of Quantile Treatment Effects under Covariate-Adaptive Randomizations
Datasets from field experiments with covariate-adaptive randomizations (CARs) usually contain extra covariates in addition to the strata indicators. We propose to incorporate these additional covariates via auxiliary regressions in the estimation and inference of unconditional quantile treatment effects (QTEs) under CARs. We establish the consistency and limit distribution of the regression-adjusted QTE estimator and prove that the use of multiplier bootstrap inference is non-conservative under CARs. The auxiliary regression may be estimated parametrically, nonparametrically, or via regularization when the data are high-dimensional. Even when the auxiliary regression is misspecified, the proposed bootstrap inferential procedure still achieves the nominal rejection probability in the limit under the null. When the auxiliary regression is correctly specified, the regression-adjusted estimator achieves the minimum asymptotic variance. We also discuss forms of adjustments that can improve the efficiency of the QTE estimators. The finite sample performance of the new estimation and inferential methods is studied in simulations and an empirical application to a well-known dataset concerned with expanding access to basic bank accounts on savings is reported.
Open AccessRegression-Adjusted Estimation of Quantile Treatment Effects under Covariate-Adaptive Randomizations
Datasets from field experiments with covariate-adaptive randomizations (CARs) usually contain extra covariates in addition to the strata indicators. We propose to incorporate these additional covariates via auxiliary regressions in the estimation and inference of unconditional quantile treatment effects (QTEs) under CARs. We establish the consistency and limit distribution of the regression-adjusted QTE estimator and prove that the use of multiplier bootstrap inference is non-conservative under CARs. The auxiliary regression may be estimated parametrically, nonparametrically, or via regularization when the data are high-dimensional. Even when the auxiliary regression is misspecified, the proposed bootstrap inferential procedure still achieves the nominal rejection probability in the limit under the null. When the auxiliary regression is correctly specified, the regression-adjusted estimator achieves the minimum asymptotic variance. We also discuss forms of adjustments that can improve the efficiency of the QTE estimators. The finite sample performance of the new estimation and inferential methods is studied in simulations and an empirical application to a well-known dataset concerned with expanding access to basic bank accounts on savings is reported.
- 15 April 2020
- Published by Emerald Publishing
- p. 73-103
Model Selection for Explosive Models
This chapter examines the limit properties of information criteria (such as AIC, BIC, and HQIC) for distinguishing between the unit-root (UR) model and the various kinds of explosive models. The explosive models include the local-to-unit-root model from the explosive side the mildly explosive (ME) model, and the regular explosive model. Initial conditions with different orders of magnitude are considered. Both the OLS estimator and the indirect inference estimator are studied. It is found that BIC and HQIC, but not AIC, consistently select the UR model when data come from the UR model. When data come from the local-to-unit-root model from the explosive side, both BIC and HQIC select the wrong model with probability approaching 1 while AIC has a positive probability of selecting the right model in the limit. When data come from the regular explosive model or from the ME model in the form of 1 + nα/n with α ∈ (0, 1), all three information criteria consistently select the true model. Indirect inference estimation can increase or decrease the probability for information criteria to select the right model asymptotically relative to OLS, depending on the information criteria and the true model. Simulation results confirm our asymptotic results in finite sample.
- 15 April 2020
- Published by Emerald Publishing
- p. 73-103
Model Selection for Explosive Models
This chapter examines the limit properties of information criteria (such as AIC, BIC, and HQIC) for distinguishing between the unit-root (UR) model and the various kinds of explosive models. The explosive models include the local-to-unit-root model from the explosive side the mildly explosive (ME) model, and the regular explosive model. Initial conditions with different orders of magnitude are considered. Both the OLS estimator and the indirect inference estimator are studied. It is found that BIC and HQIC, but not AIC, consistently select the UR model when data come from the UR model. When data come from the local-to-unit-root model from the explosive side, both BIC and HQIC select the wrong model with probability approaching 1 while AIC has a positive probability of selecting the right model in the limit. When data come from the regular explosive model or from the ME model in the form of 1 + nα/n with α ∈ (0, 1), all three information criteria consistently select the true model. Indirect inference estimation can increase or decrease the probability for information criteria to select the right model asymptotically relative to OLS, depending on the information criteria and the true model. Simulation results confirm our asymptotic results in finite sample.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 16, 2733-2737
An Optimization Method for Unit Commitment to Promote Wind Power Accommodation
Wind power has been promoted to mitigate the energy crisis worldwide. However, owing to the chaotic nature of atmospheric movement, wind power generation always exhibits nonlinear and non-stationary uncertainties. Designing an energy storage system for dispatch of a wind farm is an effective integration solution. This paper presents a framework for integrating the stochastic unit commitment, energy storage, dispatch formulations that account for wind power uncertainty. Generation cost and the fluctuation of load are proposed as multi-objective functions with the constraint functions. Numerical results on three cases including thermal generators, wind farm and energy storage and comparisons with results obtained using state-of-art optimization software, show that can be solved most efficiently and the energy storage device improves the flexibility of system operation.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 16, 2733-2737
An Optimization Method for Unit Commitment to Promote Wind Power Accommodation
Wind power has been promoted to mitigate the energy crisis worldwide. However, owing to the chaotic nature of atmospheric movement, wind power generation always exhibits nonlinear and non-stationary uncertainties. Designing an energy storage system for dispatch of a wind farm is an effective integration solution. This paper presents a framework for integrating the stochastic unit commitment, energy storage, dispatch formulations that account for wind power uncertainty. Generation cost and the fluctuation of load are proposed as multi-objective functions with the constraint functions. Numerical results on three cases including thermal generators, wind farm and energy storage and comparisons with results obtained using state-of-art optimization software, show that can be solved most efficiently and the energy storage device improves the flexibility of system operation.