Traditional threshold-based stock networks suffer from subjective parameter selection and inherent limitations: they constrain relationships to binary representations, failing to capture both correlation strength and negative dependencies. To address this, we introduce statistically validated correlation networks that retain only statistically significant correlations via a rigorous t-test of Pearson coefficients. We then propose a novel structure termed the largest strong-correlation balanced module (LSCBM), defined as the maximum-size group of stocks with structural balance (i.e., positive edge-sign products for all triplets) and strong pairwise correlations. This balance condition ensures stable relationships, thus facilitating potential hedging opportunities through negative edges. Theoretically, within a random signed graph model, we establish LSCBM’s asymptotic existence, size scaling, and multiplicity under various parameter regimes. To detect LSCBM efficiently, we develop MaxBalanceCore, a heuristic algorithm that leverages network sparsity. Simulations validate its efficiency, demonstrating scalability to networks of up to 10,000 nodes within tens of seconds. Empirical analysis demonstrates that LSCBM identifies core market subsystems that dynamically reorganize in response to economic shifts and crises. In the Chinese stock market (2013–2024), LSCBM’s size surges during high-stress periods (e.g., the 2015 crash) and contracts during stable or fragmented regimes, while its composition rotates annually across dominant sectors (e.g., Industrials and Financials).

This study develops a comprehensive methodological framework for identifying and quantifying investor echo chambers in online stock discussion forums. Motivated by a dynamic model of endogenous echo chamber formation, which formalizes how investors optimally allocate attention and update beliefs under cognitive and informational constraints, we construct a two-layer multiplex investor network that integrates common-attention similarity and semantic similarity to jointly capture the informational and cognitive linkages among investors. This framework enables the systematic examination of how shared information sources and convergent opinions emerge within investor communities. We compute both community-level and individual-level (node-level) echo-chamber intensity by integrating measures of social homophily, semantic reinforcement, and community insularity. At the firm level, we further aggregate these micro-level indicators using attention-weighted indices, community concentration (HHI), and semantic polarization metrics to characterize how echo-chamber dynamics manifest in firm-related discussions. In addition, we propose a general empirical panel framework to examine the relationship between investor echo-chamber intensity and firm-level outcomes. Overall, this paper provides a methodological foundation for the broader Investors’ Echo Chamber Project, offering scalable tools for network-based behavioral analysis and laying the groundwork for future research linking online social dynamics, financial market efficiency, and corporate decision-making.

Cross-firm momentum effects via shared analyst coverage are well-documented in de-veloped markets, but their robustness remains unclear in emerging markets, where information diffusion is asymmetric and analyst coverage is highly concentrated. Our work revisits this effect in an environment of extreme informational frictions — the Chinese market. We reconstruct the information transmission channel within the an-alyst coverage network by introducing a novel weighting scheme based on strength centrality (SC). This measure identiffes inffuential leader firms that command dis-proportionate attention from both analysts and the market. Our results demonstrate that SC-weighted connected-firm returns robustly predict cross-sectional stock returns, yielding significant and persistent profits even under a rigorous stock filter. This per-formance cannot be subsumed by strategies based on alternative weighting schemes or by explanations such as intra-industry cross-firm momentum and information discreteness. Further analysis reveals that the superiority of the SC-based approach stems from its ability to effectively identify firms with stronger cross-period fundamental linkages. In addition, high-SC stocks are characterized by higher investor attention, more efficient information processing, lower arbitrage costs, and greater internationa exposures. With this evidence, we further confirm a directional spillover: cross-firm momentum effects flow exclusively from these high-SC leaders to low-SC laggards, and there is no reverse spillover. Our findings suggest that cross-firm momentum may be systematically underestimated in many international markets due to methodological limitations rather than economic irrelevance. The SC-based framework therefore of-fers a portable tool for global investors and researchers operating in environments with asymmetric information.

We propose a time-varying framework for tail risk contagion based on conditional higher co-moments (Co-HCM), derived from a DCC-GARCH-MGH model that provides closed-form expressions for dynamic co-moments. Applying this CoHCM approach, we construct tail contagion networks across Belt and Road Initiative (BRI) stock markets. Our ffndings indicate that covariance-based metrics underestimate the ex-tent of epidemic transmission, while the CoHCM metrics reveal China’s pivotal role in spreading outbreaks and identify a distinct cluster of core transmission hubs, particularly during the 2015 Chinese stock market crisis. Dynamic contagion further exhibits cross-country heterogeneity that the Southeast Asian markets synchronize tightly with China during crises, while smaller and resource-driven markets display more inter-mittent contagion patterns. These ffndings highlight the importance of higher co-moment dependence for monitoring systemic risk in interconnected emerging markets.

随着汽车金融下沉市场的拓展与多源异构数据的爆发，传统信用评分模型在兼顾预测精度与特定场景泛化能力时遭遇瓶颈。本文提出一种基于多模态混合专家模型（Multimodal Mixture of Experts, MMoE）的深度风控框架。该框架依托企业级AI中台，通过动态门控网络（Gating Network）将借款人的结构化征信、非结构化文本语义及动态行为特征智能路由至专属专家网络。基于 LendingClub 公开数据集的实证研究（有效映射汽车金融多模态场景）表明，MMoE 模型在 AUC 与 KS 指标上显著优于 LightGBM 等主流基准模型，且其期望校准误差（ECE）降至 0.015。研究证实，门控路由机制不仅提升了长尾人群的逾期预测准度，更为深度学习在金融领域的应用提供了宏观可解释性视角。本研究为金融机构构建高并发、易扩展的下一代智能风控底座提供了系统性的工程路径与理论支撑。

We develop a deep-visualization framework for timing the factor zoo. Historical factor return trajectories are converted to two complementary image representations, which are then learned by convolutional neural networks (CNNs) to generate factor-specific timing signals. Using 206 equity factors, our CNN-based forecasts deliver significant economic gains: timed factors earn an average annualized alpha of about 6\%, and a high-minus-low strategy yields an annualized Sharpe ratio of 1.22. The outperformance is robust to transaction costs, post-publication decay, and factor category-level analysis. Interpretability analyses reveal that CNNs extract predictive signals from path boundaries and regime shifts, capturing patterns orthogonal to investor attention.

This study develops a novel deep learning framework for bank distress prediction, designed to overcome the limitations of static network analysis and to enhance model interpretability. We propose a Spatio-Temporal Attention Network that uniquely captures the time-varying nature of systemic risk. Methodologically, it introduces two key innovations: (1) a dynamic interbank network whose connection weights are adjusted by the volatility of the Shanghai Interbank Offered Rate (SHIBOR), reflecting real-time market liquidity changes; and (2) a dual spatio-temporal attention mechanism that identifies critical time steps and pivotal contagion pathways leading to a distress event. Empirical results demonstrate that the model significantly outperforms traditional benchmarks across key metrics including accuracy and F1-score. Most critically, the architecture proves exceptionally effective at reducing Type II errors, substantially minimizing the failure to identify at-risk banks. The model also offers high interpretability, with attention weights visualizing intuitive risk evolution patterns. We conclude that incorporating dynamic, liquidity-adjusted networks is crucial for superior predictive performance in systemic risk modeling.

Purpose: High investment efficiency is a key dimension of high-quality enterprise development. As critical nodes embedded in supply chain networks, corporate investment behaviors are profoundly shaped by the structural characteristics of their supply chains. Concentrated supply chain configuration, as one of the core structural features, has not yet been systematically examined in terms of its impact on corporate investment efficiency and the underlying mechanisms, leaving an important research gap. Design/methodology/approach: Based on a sample of China’s A-share listed enterprises from 2007 to 2023, this study empirically examines the effect of concentrated supply chain configuration on corporate investment efficiency. Findings: First, concentrated supply chain configuration exerts a significant inhibitory effect on corporate investment efficiency, a conclusion that remains robust after a series of tests. Second, mechanism tests indicate that this influence operates primarily through three channels: exacerbating financing constraints, crowding out working capital, and deteriorating the information environment. Third, heterogeneity analysis shows that both supplier concentration and customer concentration inhibit investment efficiency, with the latter having a slightly stronger negative effect. The adverse impact is more pronounced in over-investing enterprises, non-state-owned enterprises, smaller firms, and those in growth or decline stages. Furthermore, regional factor market development, external market power, and internal control quality are found to effectively mitigate the negative effect of concentrated supply chain configuration on corporate investment efficiency. Originality: This study extends the research on determinants of corporate investment efficiency from a supply chain structure perspective, providing new theoretical insights and empirical evidence for understanding corporate investment behavior in China.

We compile a Chinese counterpart to the TRACE dataset and provide the first trade-level analysis of China’s wholesale corporate bond market—the second largest in the world. In contrast to the dealer-dominated, core–periphery networks typical of over-the-counter markets in developed economies, China’s corporate bond market shows limited dealer intermediation. Designated dealers are reluctant to intermediate trades,and non-dealers supply the majority of liquidity, leading to wide price dispersion and low trading activity. This weak dealer participation is not driven by information asymmetry but stems from balance sheet constraints among smaller dealers and large state-owned banks’ privileged access to profitable lending opportunities.

We examine how social networks shaped by high-speed rail connections influence investor attention and market reactions to earnings announcements in China. Firms in high-centrality cities exhibit stronger immediate and subsequent responses in investor attention, stock price, and trading volume to earnings news. Further analysis shows that earnings-induced local attention predicts future attention spillovers to intercity investors, amplifying both price and volume reactions after announcements. Overall, these findings indicate that high-speed rail networks foster investor social networks that facilitate the dissemination of firm news and help explain predictable patterns in investor behavior and market pricing.