The impact of Artificial Intelligence (AI) on firm performance is an emerging issue in both practice and research. However, discussions surrounding the effect of AI on productivity are enshrouded in a paradoxical quandary. This study examines the relationship between AI implementation and total factor productivity (TFP), considering the moderation effects of digital infrastructure quality, business diversification, and demand uncertainty. Using data from 2155 Chinese firms over 2016-2021, our empirical analysis reveals a nuanced pattern: while moderate AI implementation achieves the best TFP, excessive and insufficient implementation yields diminishing returns. The curvature of this inverted U-shaped relationship flattens with higher levels of digital infrastructure quality but steepens when firms undertake diversified businesses and face heightened demand uncertainty. The findings suggest that the impact of AI on TFP is not universally beneficial, and the relationship between AI and TFP varies across different contexts. These findings also provide implications on how firms can strategically implement AI to maximize its value.

China's SMIC and Taiwan's TSMC are caught on opposite sides of the "Global Chip War." TSMC, despite having extensive commercial ties and fabs in the Mainland, is a beneficiary of U.S. efforts to stifle competition from Mainland competitors like SMIC. Geopolitical considerations, therefore, are increasingly influencing TSMC’s business decisions, as shown by TSMC’s construction of fabs in Japan and the United States despite founder Morris Chang’s longstanding opposition to overseas fabs due to their high costs. SMIC, meanwhile, is the Mainland’s best hope for creating a “red chip supply chain” and achieving 70% semiconductor self-sufficiency via domestic suppliers, which has taken on even more importance due to U.S. sanctions on advanced chips for AI model development. This article analyzes SMIC founder Richard Chang’s dream of building a red chip giant on the Mainland that can rival or even replace TSMC, which will directly conflict with Chang's former co-worker and fellow Taiwanese Morris Chang’s dream of solidifying TSMC and Taiwan’s position as the irreplaceable center of the semiconductor industry well into the 21st century.

Trade credit is a vital external source of financing, playing a crucial role in redistributing credit from financially stronger firms to weaker ones, especially during difficult times. However, it is puzzling that the redistribution perspective alone fails to explain the changes in trade credit when firms get involved in major litigation, which can be seen as an external shock for firms. Based on a firm-level dataset of litigations from China, we find that firms involved in major litigation not only exhibit an increased demand for trade credit but also extend more credit to their customers. Our further analysis reveals that whether as plaintiffs or defendants, litigation firms experience an increase in the demand and supply of trade credit. Moreover, compared to plaintiff firms, defendant firms experience a more pronounced increase in the demand for trade credit. Using firms’ market power and liquidity as moderators, we find that the increase in the demand for trade credit is more likely due to firms’ deferred payments rather than voluntary provision by suppliers, and the increase in the supply of trade credit appears to be an expedient measure to maintain market share. Generally, our results provide evidence of credit contagion effect within the supply chain, where the increased demand for trade credit is transferred from firms’ customers to themselves when they get involved in major litigations, while the default risk is simultaneously transferred from litigation firms to upstream firms.

China's tech companies are making waves with their recent achievements, including a "trifold" phone from Huawei and the revolutionary AI reasoning model from DeepSeek. Much discussion has centered on the founders of these companies and their ability to gain an edge on American rivals. But what is less appreciated or understood among foreign analysts of China’s tech giants is the role that innovation and transformation in corporate governance and organizational structure has played in these companies’ successes. Moreover, there are unique aspects of these companies from a corporate governance perspective that are not commonly seen in tech companies in other parts of the world or even within China itself. For instance, Huawei is 99% employee owned, while Alibaba is primarily governed by an "Alibaba Partnership." These unique corporate structures have arisen due to several factors, including the rapid changes to China’s regulatory landscape over the past three decades, distinct characteristics of Chinese business culture, geopolitical tensions and preoccupations with national security, and the “socialism with Chinese characteristics” model. In this article I overview some of the more distinctive corporate governance mechanisms of four Chinese tech companies: Huawei, Alibaba, Bytedance, and DeepSeek, and explain why these structures were adopted.

We combine a new collection of microdata from China with a natural policy experiment to investigate the extent to which reductions in rural-urban migration barriers affect flows of trade and investments between cities and the countryside. We find that increases in worker eligibility for urban residence registration (Hukou) across origin-destination pairs increase rural-urban exports, imports, capital inflows and outflows, both in terms of bilateral transaction values and the number of unique buyer-seller matches. To quantify the implications at the regional level, we interpret these estimates through the lens of a spatial equilibrium model in which migrants can reduce buyer- seller matching frictions. We find that a 10% increase in a rural county’s migration market access on average leads to a 1.5% increase in the county’s trade market access and a 2% increase in investment market access. In the context of China’s recent Hukou reforms, we find that these knock-on effects on market integration were on average larger among the urban destinations compared to the rural origins, reinforcing incentives for rural-urban migration.

This paper examines the impact of green mergers and acquisitions (M&As) on firms’ transition to green exports. We develop a “Technology-Qualification” theoretical framework and conduct the empirical analysis using a matched dataset of Chinese listed manufacturing firms and customs records. The findings show that green M&As significantly promote firms’ green exports, and this effect remains consistent across a series of robustness test. Mechanism analysis reveals that green M&As promote green exports through two key channels: green innovation spillovers and green qualification spillovers. Further heterogeneity analysis indicates that the positive impact of green M&As on green exports is more pronounced among firms with stronger operational performance, weaker green foundations, and those involved in processing trade. In addition, green M&As not only stimulate green exports but also prevent the entry of polluting products and reduce the exit of green product, thereby driving a green-oriented dynamic restructuring of firms’ export structure. This paper offers micro-level insights into how firms can navigate the dual challenges of enhancing green production capabilities and overcoming barriers to green trade during their transition to green exports.

This study explores the anchoring effect of previous bid premiums on acquirers’ bidding behavior in serial acquisitions. We demonstrate that, after controlling for deal characteristics, learning, and unobserved factors, the current bid premium is positively correlated with the acquirer’s previous bid premium. The strength of this anchoring effect diminishes with longer time intervals between acquisitions and increases with the industry similarity of targets. Notably, it remains unaffected by the acquirer’s state ownership or acquisition frequency. Additionally, the anchoring effect is less pronounced during periods of high economic uncertainty and can reverse following a change in the acquirer’s CEO. Our findings suggest that serial acquisitions are interrelated events, challenging the notion that each bid is an isolated occurrence. This research provides insights into the underperformance of serial acquirers compared to single acquirers and the declining trend in announcement returns across successive deals.

Background Economic stability depends on the ability to foresee financial risk, particularly in markets that are extremely volatile. Unbalanced financial data is difficult for traditional Support Vector Machine (SVM) models to handle, which results in subpar crisis detection capabilities. In order to improve financial risk early warning models, this study combines Gaussian SVM with stochastic gradient descent (SGD) optimisation (SGD-GSVM). Methods The suggested model was developed and assessed using a dataset from China's financial market that included more than 2,000 trading days (January 2022–February 2024). Missing value management, Min-Max scaling for normalising numerical characteristics, and ADASYN oversampling for class imbalance were all part of the data pretreatment process. Key evaluation metrics, such as accuracy, recall, F1-score, G-Mean, AUC-PR, and training time, were used to train and evaluate the SGD-GSVM model to Standard GSVM, SMOTE-SVM, CS-SVM, and Random Forest. Results Standard GSVM (76% accuracy, 1,200s training time) and CS-SVM (81% accuracy, 1,300s training time) were greatly outperformed by the suggested SGD-GSVM model, which obtained the greatest accuracy of 92% with a training time of just 180 seconds. Additionally, it showed excellent recall (90%) and precision (82%), making it the most effective and efficient model for predicting financial risk. Conclusion This work offers a new method for early warning of financial risk by combining SGD optimisation with Gaussian SVM and employing adaptive oversampling for data balancing. The findings show that SGD-GSVM is the best model because it strikes a balance between high accuracy and computational economy. Financial organisations can create real-time risk management plans with the help of the suggested technique. For additional performance improvements, hybrid deep learning approaches might be investigated in future studies.

Shill bidding, the use of non-genuine bids to inflate prices, undermines auction market integrity. Exploiting China’s online judicial housing auctions as a laboratory, we identify 2% of participants as suspected shill bidders, affecting 8% of auctions. They raise price premium by 14.3%, causing an annual deadweight loss of ¥570 million for homebuyers. Mechanism analysis reveals they create bidding momentum and intensify competition. We establish causality using a difference-in-differences analysis leveraging a 2017 regulatory intervention and an instrumental variable approach using dishonest judgment debtors. These findings offer actionable insights for policymakers and auction platforms to combat fraud in online high-stake auctions.

This study examines how housing purchase restrictions (HPRs) affect firms' robotics adoption through labor cost increases. Exploiting policy-driven housing price shocks across Chinese cities, we find firms significantly accelerate robot adoption in response to higher labor costs. Effects are pronounced among financially unconstrained firms, state-owned enterprises, and firms with skilled or educated workforces. Automation investments subsequently improve firm productivity, profitability, and market positions. Our findings highlight unintended spillovers from housing regulations to firm-level technological decisions and suggest policymakers consider these indirect effects when designing local market interventions.