We study whether commodity risk is efficiently priced in the cross-section of US equities. We first provide novel evidence on firm-level risk exposure to up to 42 individual commodities. We then construct a mispricing measure as the realized return less the commodity-risk-implied return, and find that it strongly predicts next-month return reversal beyond the traditional short-term reversal. This commodity-risk-mispricing reversal is substantially stronger among small, less analyst- or media-covered stocks, is weakly associated with investor sentiment, and is less prevalent among firms that engage in commodity hedging. In particular, underpricing among these hedgers often remains persistent and even intensifies after the 2004 financialization of commodity markets. This implies that despite the overall decline in mispricing, financialization has likely introduced more costs than benefits for corporate risk management.

This paper examines how major artificial intelligence (AI) breakthroughs affect the market pricing of firms’ pre-existing AI capabilities. Using Chinese listed firms’ 2023–2024 AI patents classified with a large language model, we estimate dynamic stock return and volatility responses to the DeepSeek-R1 release. Stronger pre-event AI capabilities are associated with higher post-release returns and greater short-term volatility, with volatility effects persisting longer. Positive return responses concentrate in algorithms and models, not computing infrastructure, while industry applications show later volatility increases without comparable return gains. These findings highlight technology-specific differences between return and volatility responses to AI breakthroughs.

This study investigates the role of contextual cues in memory-based decision-making within high-stakes trading environments. Using trade records from a large Chinese brokerage firm, I show that an everyday contextual cue—geographic location—triggers the recall of previously traded stocks: when an investor returns to a city she has visited before, her probability of retrading stocks associated with that city rises by 2.1 percentage points under the most saturated specification (up to 12.2 percentage points under less saturated ones). The effect displays the signatures of similarity-based recall: it is diluted by intervening trading experiences, decays with the age of the memory trace, survives when the city is not the investor's home, is not explained by mechanical recency, and is present for orders submitted via both smartphones and PCs. I then examine an extreme-event cue, COVID-19 quarantine, and find a smaller and less robust effect; the interference and decay signatures are absent precisely for these strongly encoded traces, as the memory account predicts. Memory-cued transactions underperform: stocks bought upon the cue earn about 50 basis points less per week than stocks bought for other reasons on the same day. The study provides causal field evidence for memory-based theories, particularly similarity-based recall, in real-world trading.

This paper asks whether a personally experienced bubble crash leaves a lasting imprint on investment style. We study this in China’s stock market, where two technology-growth booms —the 2014–2015 Internet boom and the 2025–2026 AI hardware boom—share a similar narrative but are separated by a decade of generational turnover. Following managers across the mutual funds they run from 2013Q1 to 2026Q1, we compare those who lived through the 2015 crash with otherwise similar managers who entered afterwards. We find that the caution of veteran managers reflects a scar rather than a symptom of ageing. Conditional on age and career length, managers who lived through the 2015 crash hold more value exposure and chase technology stocks less when prices have been rising, and this reluctance deepens with how much the crash hurt them—crash-window tenure, performance shortfalls, redemption pressure—rather than with age. These demand differences reach asset prices as well: stocks held disproportionately by managers without bubble memories carry significantly higher subsequent left-tail risk, whereas crash-experienced ownership stabilises the stock, particularly when ownership is broad. Heterogeneous institutional beliefs rooted in career experience therefore shape not only how managers invest but also the crash risk of the assets they hold.

Can fully agentic artificial intelligence predict equity returns in real time? We build a novel dataset of daily AI-generated attractiveness scores spanning the entire Russell 1000 universe, continuously updated since April 2025. Each trading day, the model freely searches the live internet and produces quantitative scores for every constituent stock before the next market open, eliminating both look-ahead bias and attribution bias. A daily value-weighted portfolio of the 20 highest-ranked stocks earns a Fama-French six-factor alpha of 16.5 basis points per day (41.5% per annum), with an average bid-ask spread of 1.92 basis points. Predictability is sharply asymmetric: the model reliably identifies future outperformers but not future underperformers. The evidence points to stock picking rather than market timing: performance is strongest on down-market days, and the alpha does not come from betting on aggregate market direction. The AI’s reasoning traces help explain this predictability. Alpha is concentrated in traces with richer live-web evidence and stronger analytical structure. In the theme attribution, most active alpha reflects firm selection under an investment lens, not broad theme choice.

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 identify choke points in global production networks as industries bridging flows across global value chains. These industries exhibit low substitutability: US firms exposed to Chinese choke points during the 2022 Covid-19 lockdowns experienced large and persistent sales declines. The Red Sea crisis demonstrates how negative shocks to water transport, a single choke point, propagate throughout the global network. This structural fragility is priced in global stock markets: firms in choke point industries earn annualized benchmark-adjusted returns exceeding 6%. The premium compensates for aggregate consumption risk, as downturns in choke point industries predict lower future US and global consumption growth.

This study investigates the relationship between the number of stock concepts and future returns, as well as the economic mechanisms underlying this association. Using novel collected data, we find that stocks with a greater number of concepts earn significantly higher returns in the subsequent month, generating a six-factor adjusted annualized alpha of approximately 9.6% for a long-short portfolio. Although these stocks exhibit higher turnover, return volatility, and investor attention - features commonly associated with speculative concept-driven trading - an analysis of cross-listed AH twin stocks reveals that concept counts do not widen the AH premium. Moreover, the return premium persists for up to 15 months without significant reversal, whereas firms that engage in opportunistic concept-chasing exhibit pronounced long-run reversals, suggesting that the positive concept-return relation is not driven by speculative motives. The higher returns are primarily attributable to strong industrial policy support and sustained above-expectation operating performance. Firms with more concepts are more likely to receive government subsidies and deliver positive earnings surprises, effects that are amplified when their core concepts receive stronger national policy backing. In contrast, firms that pursue concepts lacking substantive business relevance are significantly less likely to obtain government subsidies and fail to achieve above-expectation growth. Regarding investor composition, institutional ownership of high-concept stocks increases modestly, while ownership by government-guided funds rises substantially - particularly for stocks whose core concepts are strongly supported by national industrial policies. Conversely, concept-chasing behavior by listed firms significantly reduces the ownership ratio of government-guided funds. Overall, our findings indicate that concept stocks in China’s capital market are not purely speculative but signal underlying national policies.

This paper investigates the sales seasonality premium (Gustavo et al., 2020) in the Chinese stock market. We document a significant sales seasonality premium in the cross section of stocks listed on Chinese market. A long-short strategy of buying low-sales stocks and shorting high-sales season stocks generate a monthly return of 0.84% with a Newey-West t statistic of 2.94. The return spreads between low-sales season stocks and high-sales season stocks are robust to well-known anomalies and are larger in magnitude among large-cap companies. We also find that the return spreads are more pronounced within industries with relatively fixed patterns in product demand. The sales seasonality premium in the Chinese stock market is found to be a combined result of investor rationality and irrationality.

Under China’s transition toward high-quality financial development, this study investigates the tail dependence among media sentiment, investor attention, and stock market returns using the M-Clayton Copula. Comparative model fitting tests demonstrate that the M-Clayton Copula outperforms both single Copula models and M-Copula models in characterizing asymmetric negative dependence structures. It is particularly effective in capturing both upper-lower and lower-upper tail dependencies, thereby providing a more comprehensive analysis of their interdependencies. The empirical results reveal three key findings. First, a significant positive tail dependence exists between media sentiment and stock market returns, suggesting that they tend to move together under extreme conditions, with notable asymmetry in the strength of co-movements. Second, media sentiment and investor attention exhibit negative tail dependence, with the lower-upper tail dependence coefficient exceeding its upper-lower counterpart, suggesting higher probability of rising investor attention following media sentiment decline than vice versa. Third, a pronounced negative tail dependence emerges between stock market returns and investor attention, particularly showing strong lower-upper tail correlation, implying substantial likelihood of increased investor attention subsequent to market downturns.