This paper investigates the differential predictive power of overnight and intraday returns in China's A-share market. We find that overnight returns exhibit a significantly positive and robust relation with future stock returns at intermediate horizons, whereas intraday returns display no comparable predictive effect. Mechanism analysis indicates that this predictability is primarily driven by the T+1 trading rule, which restricts timely selling and leads to an underreaction to negative information. Further analyses show that overnight returns embed information about future earnings, suggesting that they capture the delayed incorporation of fundamental news resulting from institutional trading frictions. Our findings differ from patterns documented in developed markets and offer new insights into the price discovery process in emerging equity markets.

We examine whether the characteristics of economically linked firms predict stock returns in the Chinese A-share market. We construct two peer-based signals: a peeraverage (PA) signal that aggregates the characteristics of fundamentally similar firms, and a peer-deviation (PD) signal that summarizes a firm’s characteristic deviations from industry-peer averages. PA captures shared fundamental information, whereas PD benchmarks a firm’s relative position against economically comparable peers. Portfolio sorts, factor-spanning tests, and Fama–MacBeth regressions reveal economically meaningful and statistically significant return predictability for both signals, with generally larger portfolio returns among non-state-owned enterprises. Both signals retain significant incremental predictive power after controlling for the full set of own-firm characteristics and flexible machine-learning forecasts based solely on these characteristics. Changes in both signals also predict future earnings, while return predictability is generally stronger among firms receiving less investor attention, consistent with limited-attention-induced underreaction. These findings highlight the distinct informational and benchmarking roles of economic networks and their predictive value beyond firms’ own characteristics.

We posit and demonstrate that, in China’s retail-dominated market, quantitative trading over-relies on non-fundamental signals, thereby crowding out fundamental information from stock prices and increasing crash risk. Using trading data from quantitative mutual funds and Chinese A-share firms during 2009-2023, we find that greater exposure to quantitative trading is associated with higher future crash risk. Mediation analysis further reveals that reduced information efficiency constitutes a key channel through which quantitative trading elevates crash risk. The effect is stronger for stocks with more retail investors, consistent with our proposed mechanism. Overall, we identify a novel potential risk of quantitative trading in underdeveloped emerging markets.

We examine whether large language models (LLMs) can extract contextualized representation of Chinese public news articles to predict stock returns. Based on representativeness and influences, we consider seven LLMs: BERT, RoBERTa, FinBERT, Baichuan, ChatGLM, InternLM, and their ensemble model. We show that news tones and return forecasts extracted by LLMs from Chinese news significantly predict future returns. The value-weighted long-minus-short portfolios yield annualized returns between 35% and 67%, depending on the model. Building on the return predictive power of LLM signals, we further investigate its implications for information efficiency. The LLM signals contain firm fundamental information, and it takes two days for LLM signals to be incorporated into stock prices. The predictive power of the LLM signals is stronger for firms with more information frictions, more retail holdings and for more complex news. Interestingly, many investors trade in opposite directions of LLM signals upon news releases, and can benefit from the LLM signals. These findings suggest LLMs can be helpful in processing public news, and thus contribute to overall market efficiency.

To explore what kinds of roles foreign investors take in a gradually opening financial market, we propose the abnormal holding value ratio (AHVR) of northbound investors among stocks through China’s Stock Connect Mechanism. We find that AHVR positively predicts the expected stock returns and significantly relates to firms’ quality-related fundamental information, especially profitability. Foreign investors learn the firm fundamentals before they invest in the Chinese market, which is different from the trading behavior of domestic individual investors. The AHVR premium is larger among firms with higher attention of analysts who focus on effective information and with lower attention of individual investors who have behavioral bias. In all, the northbound inflows are smart money, which will increase the efficiency of the Chinese market instead of simply chasing stars that only grab investors’ attention.

Using novel data from a stock forum, we analyze return extrapolation in the cross-section. Our findings indicate that extrapolators overreact to the returns but underreact to the fundamentals. The post-earnings-announcement drift (PEAD) is more pronounced among firms with a high firm-level degree-of-extrapolation (DOX). Additionally, investors ask fewer questions about high-DOX firms’ fundamental information on official online interactive platforms. Extrapolation reduces the informativeness of stocks due to investors’ inattention to fundamentals. Furthermore, extrapolators’ overreaction to returns and underreaction to fundamentals increase stock price crash risks. These findings support explanations of extrapolation based on limited asymmetric attention.