This paper revisits two widely studied predictors of commodity futures returns, basis and basis momentum, whose conventional measures using first-nearby futures as proxies for spot prices may limit their ability to capture fundamental spot-market risks. Motivated by this limitation, we construct two spot-based signals from observed spot and futures prices, which are theoretically shown to contain incremental information beyond conventional measures. Using 41 Chinese commodity futures, we find that these signals robustly predict first-nearby contract returns and remain significantly priced in time-series and cross-sectional tests, even after controlling for their conventional counterparts. We then develop a spot-enhanced three-factor model, including the market factor and the two spot-based factors, which consistently outperforms three widely used benchmark models in pricing competing factors and explaining return anomalies.

This paper identifies the marginal investors in China’s stock market and examines their impact on stock pricing. To clearly distinguish between the equity constraint channel and the debt constraint channel, we construct the capital ratio factor and the debt constraint factor for banks and securities companies, the two most critical financial intermediaries in China’s stock market. Our results demonstrate that banks indeed serve as marginal investors and influence stock market efficiency primarily through the equity capital constraint channel. Furthermore, we find that the bank capital ratio factor significantly explains stock mispricing in China, with the single-factor model based on bank equity capital producing substantially smaller pricing errors compared to traditional multi-factor models.

The increasing institutional participation and deepening integration of physical trading and financial operations in commodity markets have elevated the interconnectedness of energy futures and equity markets to prominence in both scholarly discourse and industry analysis. Employing the Nelson-Siegel framework and Fama-French factor model, this study examines the dynamic relationships between energy futures holding cost variations and equity returns across coal and oil sectors. Our analysis yields three principal findings: First, the Fama-French three-factor model exhibits robust explanatory power in China's energy sector equity market, revealing significant statistical relationships between holding cost curve parameters—level, slope, and curvature—and industry excess returns. Second, holding cost variations manifest substantial heterogeneity in their impact on stock returns across coal and oil sectors. Third, carrying cost components demonstrate dominance over shock transmission effects in explaining industry stock return volatility, indicating complex, asymmetric interaction mechanisms between futures and equity markets. Drawing from these empirical results, we advance targeted policy prescriptions addressing futures market architecture and financial stability.

This study examines the pricing effect of global economic policy uncertainty (GEPU) in the cross-section of individual stocks and portfolios in the Chinese stock market. Employing the GEPU index as a systematic risk factor, our empirical analysis demonstrates that stocks in the lowest decile of βGEPU generate risk-adjusted annualized returns that are 5.16% higher than those in the highest decile. Our analysis reveals that this βGEPU premium is driven by the outperformance of stocks with negative βGEPU and the underperformance of those with positive βGEPU. These findings suggest that uncertainty-averse investors not only demand compensation for holding stocks with negative βGEPU exposure but are also willing to pay a hedging premium for assets that serve as positive βGEPU hedges. The results prove robust across multiple specifications, persisting in both bivariate portfolio sorts and Fama-MacBeth cross-sectional regressions that control an extensive set of classic pricing factors.

This study applies the generalized dynamic factor model (GDFM), TVPVAR-DY framework, and pattern causality to investigate spillover effect from international commodity idiosyncratic volatility co-movements to China's financial market risk, as well as the impact of a series of macroeconomic factors on such spillover effect. The empirical results indicate that the idiosyncratic volatility co-movements of energy, industrial metals, precious metals, soft commodities, and agricultural products all have significant spillover effects on China's financial market risk. The influence of commodity idiosyncratic co-movements on China’s financial market risk is relatively stable under normal economic conditions but intensifies significantly during periods of deteriorating economic fundamentals. Macroeconomic factors such as international capital flows, investor sentiment, geopolitical risks, economic conditions, and international freight rates predominantly exhibit a positive causal effect on the dynamic spillover effect.

While many studies document cross-stock predictability where returns of some stocks predict returns of other similar stocks, most evidence comes from US markets. Following Chen et al. (2019), we identify peer firms based on historical return similarity and construct a Peer Return Gap (PRG) measure, defined as the difference between a stock’s lagged return and its peers’ returns. Our empirical evidence from Chinese markets shows that past-return-linked peers strongly predict focal firm returns. A long-short portfolio sorted on PRG generates an equal-weighted monthly return of 1.26% (t = 3.81) and a Fama-French five-factor alpha of 1.10% (t = 2.86). These abnormal returns remain unexplained by several alternative factor models.

This study introduces an innovative approach for constructing multimodal investor sentiment indices and explores their varying impacts on stock market returns. We employ the RoBERTa model to quantify text-based sentiment, the Google Inception(v3) model for image-based sentiment measurement, and a multimodal semantic correlation fusion model to comprehensively consider the interplay between textual and visual sentiment features. These sentiment indices are further categorised into industry-specific investor sentiment and market-wide investor sentiment, enabling separate analyses of their effects on stock markets. Furthermore, we leverage these indices to build a multifactor stock selection model and timing strategies. Our research findings demonstrate that multimodal sentiment analysis yields superior predictive accuracy. Industry-specific investor sentiment exerts bidirectional positive influences on stock market returns, whereas market-wide investor sentiment indices exhibit unidirectional impacts. Integrating industry-specific investor sentiment into our multifactor stock selection model effectively enhances portfolio returns. Furthermore, combining market-wide investor sentiment with timing strategy optimisation further augments this advantage.

As China’s Zero-COVID policy has come to an end and travel restrictions have been removed, the country’s mobility patterns are very likely to become more heterogeneous than during the pandemic. Human mobility is a key mechanism through which economic activities emerge and viruses spread. It can bring both advantages and challenges to cities with different characteristics. This paper investigates intra-city mobility trajectories of 368 Chinese cities within a non-linear time-varying latent factor framework to uncover the evolution of heterogeneity in local travel behavior amidst that China has been approaching the turning point of the post-pandemic new normal. To this end, we compiled a novel panel on a weekly basis, using the latest Baidu Mobility Data and the risk-level data released by the State Council of the People’s Republic of China. We further examine the effects of exposure to high COVID-19 risk in the city on commuting behavior between May 17, 2021 and June 26, 2022. Our results provide stylized facts on stratified local travel across China: first, the 368 cities can be categorized into six clusters based on their mobility dynamics, and second, the gaps in intra-city mobility tend to narrow within each cluster but widen between different clusters. Moreover, exposure to high COVID-19 risk has a stronger impact on home-workplace commuting rates than on dining-, leisure, and recreational travel rates, persistently dampening commuting behavior. In addition, divisions in intra-city travel strength and commuting behavior between western regions and the rest of China are evident. In sum, this paper suggests that the daily life and economic activities which depend heavily on human mobility are recovering at different rates across China.

We introduce a deep visualization (DV) framework that turns conventional commodity data into images and extracts predictive signals via convolutional feature learning. Specifically, we encode futures price trajectories and the futures surface as images, then derive four deep‑visualization (DV) predictors, carry ($bs_{DV}$), basis momentum ($bm_{DV}$), momentum ($mom_{DV}$), and skewness ($sk_{DV}$), each of which consistently outperforms its traditional formula‑based counterpart in return predictability. By forming long–short portfolios in the top (bottom) quartile of each DV predictor, we build an image‑based four‑factor model that delivers significant alpha and better explains the cross‑section of commodity returns than existing benchmarks. Further evidence shows that the explanatory power of these image‑based factors is strongly linked to macroeconomic uncertainty and geopolitical risk. Our findings reveal that transforming conventional financial data into images and relying solely on image-derived features suffices to construct a sophisticated asset pricing model at least in commodity markets, pioneering the paradigm of image‑based asset pricing.

We utilize the Bayesian model scan approach to examine the best performing models in a set of 15 factors discovered in the literature, plus principal components (PCs) of anomalies unexplained by the initial factors in the Chinese A-share market. The Bayesian comparison of approximately eight million models shows that HML, MOM, IA, EG, PEAD, SMB, VMG,PMO, plus the four PCs, PC1, PC6, PC7, PC8 are the best supported specification in terms of marginal likelihoods and posterior model probabilities. We also find that the best model outperforms existing factor models in terms of pricing tests and out-of-sample Sharpe ratio.