The fundamental methodologies of machine learning and econometrics are reviewed. We also discuss the challenges of integrating the data-driven and model-based causal approaches and conjecture how it may yield new insights to empirical economic studies.

Digital finance, as one of the “five major areas” of finance, plays a crucial role in improving the efficiency of financial services, promoting inclusive finance, and empowering high-quality economic development. Based on relevant literature from core English and Chinese-language databases spanning 2014-2025, this paper constructs a research framework for digital finance using bibliometrics and qualitative analysis, analyzes the shortcomings of existing research, and proposes potential future research directions. The results show that: First, Chinese-language literature focuses on policy guidance and local practices, while English-language literature focuses on sustainable development and global comparisons. Second, the research framework for digital finance can be systematically summarized through a logical thread of “measurement indicators, influencing factors, economic and social effects, innovative practices”. Third, future research can focus on refining measurement indicators, accurately identifying influencing factors, comprehensively addressing economic and social effects, and advancing the research on innovative practices. This paper broadens the perspective for further research on digital finance and provides insights for promoting its high-quality development in practice.

Climate change has emerged as a new source of instability in the global financial system, making the scientific identification and assessment of its transmission channels to the financial sector a critical issue in the field of climate finance. Currently, climate-related financial risk modeling and practical applications still face numerous obstacles. In this context, this paper reviews several key developments in climate-related financial risk studies, including the characteristics of climate risks in financial markets, the methodologies and practices for assessing climate financial risks, and future research directions. To be specific, this study first elaborates on three crucial features of climate financial risks. Second, it systematically reviews three streams of approaches for climate financial risk assessment developed in recent years, analyzes their applicability and limitations, and examines relevant practices adopted by central banks and financial regulators across different countries. Finally, the paper identifies promising directions for future research to support both theoretical advancement and practical implementation in the field of climate financial risk assessment.

This paper provides a classification framework to integrate multi-source information, including available platform firm data, survey data, and industry reports, to estimate the size of platform employment. We first define and conceptualize platform employment, which is then classified into cloud-based and location-based platform employment. We then estimate the total size and distribution of platform employment in 2023, and correct for duplicates arising from workers engaged on multiple platforms. We also estimate full-time and part-time workers separately. The results show that the total number of platform workers in China has reached 247 million in 2023, including approximately 118 million part-time and 129 million full-time workers, the latter accounts for 14.9% of China’s working-age population. Notably, cloud-based platforms employ significantly more workers than location-based platforms, with cloud-based roles predominantly part-time whereas location-based positions are primarily full-time. The paper provides a consistent framework for incorporating updated information from various sources to provide timely estimates of China’s platform workforce.

The rapid development of artificial intelligence has profoundly reshaped both the substantive focus and the methodological paradigms of behavioral science. This paper systematically synthesizes three frontier research directions that have emerged from these changes: 1) Investigations of human attitudes and behaviors during interactions with AI, and the mechanisms through which AI affects human decision-making and preferences; 2)behavioral experimental studies examining the behavioral characteristics and preference patterns of large language models; 3) methodological innovations enabled by AI technologies, including the use of AI agents as surrogates for human participants in surveys and experiments, and complex-systems research that builds dynamic interactive systems based on multi-agent frameworks. The paper concludes with a discussion of the challenges confronting, and future directions for, interdisciplinary research at the intersection of AI and behavioral science.

Short video marketing has become a core scenario for e-commerce conversion. While accelerated information dissemination and information overload have reduced consumer patience thresholds, the non-linear mechanisms through which video duration affects sales conversion remain unclear. This study examines the inverted U-shaped relationship between short video duration and product sales from the perspectives of information effectiveness and overload, revealing how this relationship is moderated by product attributes (price and type). Key findings include: 1) Information effectiveness (positive mediation) and information overload (negative mediation)constitute dual mediation pathways that jointly drive the inverted U-shaped relationship; 2) Higher-priced products reduce the optimal duration threshold and strengthen the relationship, whereas experiential products exhibit lower thresholds and weaker relationships compared to search-oriented products. The research establishes a dynamic trade-off framework for digital content marketing, demonstrating that the interplay between information density and cognitive load determines duration threshold effects. These conclusions provide empirical support for optimizing platform algorithms and innovating content production paradigms.

Due to the prevalence of “greenwashing” and similar phenomena, the actual environmental impact of green bonds remains subject to skepticism. This study investigates the effect of green bond issuance on corporate green transformation by employing a multi-period difference-in-differences (DID) model using a sample of 2,975 listed Chinese firms from 2012 to 2023. Furthermore, it conducts dual verification from the perspectives of green discourse power and the authenticity of green information disclosure. The findings reveal that: 1) Green bond issuance significantly promotes corporate green transformation, with the alleviation of financing constraints and the enhancement of reputational incentives serving as key mechanisms. 2) The intrinsic motivation of firms for digital transformation, along with external factors such as analyst attention, environmental regulation intensity, and market power, positively moderates the genuine “greening” effect of green bonds. 3) Heterogeneity analysis shows that green bonds have shown a certain positive effect on enterprises with different geographical locations, life cycles, industrial pollution attributes, ownership, and financing structures. Further differential testing revealed statistically significant differences between groups in terms of pollution level and corporate ownership, with green bonds showing a more significant incentive effect in high polluting and state-owned enterprises. Based on these findings, the study recommends improving the top-level design of green bond issuance, enhancing the quality of corporate environmental information disclosure, and adopting differentiated issuance strategies to further support comprehensive green transformation in enterprises.

New quality productive forces is an important engine for high-quality economic development and a crucial force in promoting the great rejuvenation of the Chinese nation in the new era. This article constructs a corresponding indicator evaluation system based on the connotation of the theory of new quality productive forces, and uses the entropy method to calculate the level of new quality productive forces in 236 cities in China from 2010to 2020, by constructing a theoretical model and conducting empirical research, the impact of new quality productive forces on economic growth is explored. The results indicate that new quality productive forces significantly promotes economic growth, and the three dimensions of new quality productive forces have heterogeneous effects on promoting economic growth. Mechanism analysis shows that environmental regulation intensity plays a regulatory role in promoting economic growth through new quality productive forces. In further analysis, by constructing a spatial lag model to test the spatial spillover effect of new quality productive forces, it was found that new quality productive forces can not only promote the macro economy of local cities, but also promote the economic growth of surrounding cities.

This paper systematically studies the impact and mechanism of digital government development on household consumption based on cross-border panel data from 118 economies worldwide from 2012 to 2022. Research has found that the improvement of the level of digital government construction significantly promotes the growth of household consumption. This conclusion still holds true after a series of robustness tests such as instrumental variable method, sub sample adjustment, and increasing control variables. Mechanism analysis shows that digital government mainly unleashes residents’ consumption potential through three paths: optimizing the business environment, promoting non-agricultural employment, and advancing inclusive finance development; Heterogeneity analysis reveals that the promotion effect of digital government construction on residents’ consumption is more significant in economies with high social system stability and strong macro governance capabilities, as well as in developed economies. However, in low-income economies, its effect is constrained by insufficient social stability and weak macro governance capabilities. It is necessary to enhance social resilience and improve the governance system in order to fully unleash the effectiveness of digital governance. The research conclusion of this article provides theoretical support and empirical evidence for promoting the construction of digital government in China and promoting household consumption.

Based on a nested model of IRIOT and WIOT, and combining Kronecker product to construct a spatiotemporal weight matrix, this paper empirically explores the mechanism of dual value chain embedding on“first rich driving later rich” using dynamic spatial panels. The regression results for the segmented manufacturing industry in various provinces of China show that: 1) The embedding of global value chains has widened economic disparities, while the embedding of domestic value chains can leverage industrial linkages to realize vertical governance, effectively weakening the further expansion of economic disparities. 2) The embedding of dual value chains is based on the active “spatial catch-up effect” and passive “spatial spillover effect” to promote the “first rich driving the second rich” of China’s manufacturing industry, with the former becoming stronger and the latter becoming weaker. At the same time, the inhibitory effect of domestic value chain embedding on economic disparities is gradually increasing under the catch-up effect, which can significantly offset the amplification effect of global value chain embedding on economic disparities. 3) The spatial effect decomposition results for the embedding of dual value chains indicate that the indirect effects generated by “third-party associations” are greater than the direct effects generated by “direct associations”. Under the spatial overflow mechanism, the indirect effect exceeds 65%, and even under the more spontaneous spatial catch-up mechanism, the indirect effect is as high as 55%. In view of this, deepening the construction of the domestic value chain system, strengthening inter provincial industrial linkages, and making full use of the dynamic mechanism of spillover and catch-up have important policy implications for promoting the Chinese path to modernization process of “first rich driving later rich, and ultimately achieving common prosperity”.

Under the “dual carbon” strategy, systemic risk prevention and control between China’s carbon market and commodity futures market is one of the important issues to maintain financial stability, but existing studies only focus on the low-order moment risk spillovers, and there is little literature to study the key driving mechanism. In view of this, this paper first constructs a high-order moment risk spillover measurement framework of“carbon-commodity futures” to measure the dynamic time-varying spillover index. On this basis, the SHAP algorithm and cutting-edge machine learning model are further integrated to deeply explore the driving mechanism of each influencing factor on the high-order moment risk spillovers between carbon-commodity futures markets. The findings reveal that risk spillovers between these markets exhibit significant time-varying characteristics, with kurtosis spillovers displaying the highest volatility. Financial market volatility indicators play a pivotal role in driving risk spillovers: Oil market volatility contributes 32% to volatility spillovers, EU carbon market volatility accounts for 37% of skewness spillovers, and oil market volatility explains 28% of kurtosis spillovers. The impact of various factors on risk spillovers demonstrates significant nonlinear features. Notably, economic policy uncertainty exhibits a threshold effect, where exceeding critical values triggers substantial increases in risk spillovers, while climate policy uncertainty shows an inverse nonlinear relationship. Moreover, significant interactions exist among the driving factors, with the most pronounced being the synergistic effects between financial market volatility indicators, which can potentially amplify the cross-market risk spillovers.

High-tech zones serve as crucial platforms for implementing China’s innovation-driven strategy and act as key engines for achieving high-quality economic development. The implementation of the “upgrading for promotion” policy in high-tech zones has profound implications for advancing technological innovation and fostering emerging industries. Based on panel data from Chinese cities from 2003 to 2022, this study conducts an in-depth analysis of the relationship between high-tech zone upgrades and the entry of strategic emerging enterprises. The findings indicate that the policy significantly attracts the entry of strategic emerging enterprises, and this conclusion remains robust after a series of robustness tests. Mechanism analysis reveals that the policy promotes enterprise entry through channels such as digital talent reserves, inclusive digital finance, government technology support, and fiscal and tax incentives. Heterogeneity analysis shows that the policy’s effect is more pronounced in regions with higher levels of artificial intelligence development, stronger official incentives, and greater utilization of foreign capital. Furthermore, the study finds that the policy enhances the total factor productivity (TFP) of strategic emerging enterprises within high-tech zones. These findings provide empirical evidence for promoting the advanced development of high-tech zones and accelerating the cultivation and expansion of emerging industries.

In financial markets, information sources are diverse and structurally complex, and how to systematically integrate such information into tradable investment signals has become a central issue in asset pricing research and practice. This study constructs an LLM-based decision-making system, incorporating multi-source data such as news texts, stock prices, and macroeconomic conditions. Using the constituents of the CSI 300 Index from 2020 to 2024 as the sample, we systematically evaluate the predictive power of investment signals generated by LLMs. First, we construct three types of portfolios based on buy, hold, and sell signals. Analysis of cumulative returns and risk-return metrics shows that the buy-signal portfolio achieves significantly higher excess returns (23.83%) than the hold (6.96%) and sell ($-$15.08%) signal portfolios, validating the directional predictive power of the generated signals. Second, to further enhance performance, we introduce a buy-score mechanism that classifies buy signals into finer categories and builds portfolios accordingly. Results show that the high-score long portfolio achieves a cumulative excess return of 99.20%, a Sharpe ratio of 0.65, and a turnover rate of 41.16%, indicating that the buy score has strong incremental predictive power and that the holdings exhibit persistence. Finally, we fine-tune the LLMs using in-depth research reports from brokerages, resulting in a significant improvement in strategy performance. Overall, this study demonstrates the value of financial LLMs in empirical asset-pricing applications, validates their effectiveness in signal identification, strategy construction, and model tuning, and provides a practical path for integrating AI into financial decision-making.

Interval-valued data is widely present in the economic and financial fields. Compared to point data, it provides richer information but also faces challenges such as model uncertainty and potential structural changes. To address this issue, this paper proposes, for the first time, a time-varying model averaging (IMA) method based on interval-valued data. This method dynamically selects time-varying weights by minimizing a locally quadratic loss function based on the $D_K$ distance, effectively capturing potential structural changes in interval-valued data. Moreover, the IMA method introduces a penalty term in the weight selection criterion, effectively preventing overfitting and enhancing its robustness and generalization ability. We demonstrate that the IMA method can achieve asymptotic optimality in selecting time-varying weights when all candidate models are misspecified. In the empirical study, we apply the IMA method to forecast the monthly price intervals of COMEX gold futures. The results show that the IMA method significantly outperforms existing model selection and averaging methods in predicting interval endpoints, midpoints, and ranges, especially in long-term forecasting. By dynamically adjusting weights and integrating information from multiple models, the proposed method mitigates the prediction bias that may arise from single-model selection, offering valuable insights for analyzing and forecasting complex and dynamic economic and financial systems.

As a core driver of new quality productive forces, artificial intelligence (AI) has attracted growing attention for its potential to promote industrial carbon reduction. However, existing studies largely remain at the levels of theoretical qualitative discussion and empirical quantitative analysis, lacking a systematic mathematical framework grounded in the production function. As a result, the underlying mechanisms through which artificial intelligence affects carbon emissions are not yet fully or rigorously explained. This study applies the classical task-based model developed by Acemoglu and Restrepo to construct a theoretical framework that elucidates how industrial AI affects carbon-emission intensity through task allocation mechanisms. Using panel data from A-share industrial listed firms in Shanghai and Shenzhen between 2012 and 2022, the theoretical model is further tested empirically. The results indicate that: 1) AI significantly reduces firms’ carbon-emission intensity, primarily through technological and energy-efficiency effects; and 2) the magnitude of the energy-efficiency effect depends on the optimization of the input structure between AI capital and traditional physical capital — Specifically, as AI adoption increases, the relative share of AI capital rises, thereby improving overall energy efficiency. This study contributes a novel theoretical analysis framework and empirical evidence to understanding the economic and environmental impacts of industrial AI, supporting progress toward China’s “dual-carbon” goals.

Green innovation, as the integration of the “green” and “innovation” themes under the concept of high-quality development, is an important pathway to drive China’s economic growth and enhance its international industrial competitiveness. This paper based on panel data from 214 prefecture-level cities from 2007 to 2022, this study first employs a multi-period synthetic difference-in-differences approach to examine the impact of low-carbon city pilots, innovative city pilots, and dual “low-carbon–innovation” policy pilots on green innovation. The study then further compare the policy effects of the dual pilot sequence on green innovation, and analyze in depth the heterogeneous characteristics of the impact of the pilot sequence on green innovation in different city locations, scientific and educational levels, and city levels. Finally, a spatial difference-in-differences model is constructed to explore the spatial spillover effects of the dual policy pilots. The findings reveal that low-carbon city pilots, innovative city pilots, and dual policy pilots all significantly promote green innovation, with the dual policy pilots demonstrating stronger policy effects compared to single pilots, and the effects of both single and dual pilots progressively strengthen over time. Overall, the policy of innovation first and low carbon later exhibits a greater promoting effect on green innovation levels compared to “low carbon first, then innovation”. The impact of the implementation sequence of the dual policies on green innovation varies depending on geographical location, urban level of science and education, and urban level heterogeneity. Specifically, the policy of low carbon first and innovation later fosters green innovation in eastern regions and cities with moderate levels of scientific and educational development, while the policy of innovation first and low carbon later shows larger innovation effects in eastern regions, cities with high levels of scientific and educational development, and key cities. Furthermore, the dual policy pilots demonstrate significant positive spatial spillover effects on neighboring regions and areas with similar levels of economic development. The research conclusions enrich the study of the economic effects of dual-policy pilot programs, while providing significant empirical evidence and policy references for the country to achieve high-quality development and comprehensive transformation and upgrading.

Finance is the lifeblood of the national economy and plays a pivotal role in China’s modernization. This paper conducts an in-depth analysis of the 2015 SSE 50ETF options market and reveals the profound impact of institutional investors’ options trading behavior on stock market pricing efficiency during periods of significant financial risk. The findings show that, first, institutional investors’ options trading significantly amplified the deviation of stock prices from fundamentals before the market crash. Although this effect weakened after the crash, it persisted. Second, due to heterogeneous beliefs and short-selling constraints, put and call options exhibited markedly asymmetric effects on the stock market, delaying the absorption of negative information into prices. Third, the analysis of dynamic feedback and momentum strategies demonstrates the existence of positive feedback effects in the put options market. However, informed investors, through “riding the bubble” behavior, absorbed the excess returns from momentum trading in advance. This paper uncovers the complex mechanisms by which institutional investors influence stock price formation through the options market during financial turmoil, offering a new analytical perspective for understanding the interaction between China’s derivatives and stock markets.

Synthetic control method (SCM) has emerged, as a mainstream causal inference technique, gaining popularity in policy evaluation due to its robust performance with small-scale data, transparent counterfactuals, and interpretable results. Firstly, the origin of SCM is traced back to the treatment effect model, clarifying its core principles. Secondly, the development of SCM is systematized in five aspects, including relaxation of basic assumptions, improvement of estimation methods, model structure and asymptotic properties, hypothesis testing and interval estimation, and emerging themes. Thirdly, the connection and differences between SCM and other causal inference methods are clarified. Fourthly, based on representative applications of SCM both domestically and internationally, the applicability and validity of SCM in concrete practice are analyzed. Two main reasons for using this policy evaluation tool in research are summarized: One is to estimate the individual treatment effect, and the other is to allow for the existence of individual heterogeneity and unobserved time-varying confounders. Lastly, the future of SCM is envisioned from both theoretical and applied perspectives.

Stock price jumps serve as a key indicator of macroeconomic shifts and are often viewed as a barometer of national economic conditions. The release of macroeconomic policies can significantly impact stock markets through various channels such as asset liquidity, market expectations, and investor confidence. As a result, understanding how policy announcements influence stock market jumps has become a central topic in financial volatility research, attracting the attention of policymakers, financial regulators, investors, and academics alike. This paper develops a policy model that incorporates both positive incentives and negative suppressive effects to capture the influence of policy announcements on the intensity of market jumps. Empirical results show that policy announcements have a more persistent impact on positive jumps, and the policy model outperforms traditional self-exciting point process models (e.g., Hawkes models) that do not consider policy factors, which demonstrates the effectiveness of incorporating policy information in analyzing market jumps. Furthermore, the study finds that under lower jump thresholds, the policy model exhibits enhanced fitting accuracy and stability by more effectively integrating policy-related information to capture market dynamics. Using high-frequency data, this paper evaluates the effects of policy releases from the perspectives of parameter estimation and jump prediction, offering practical insights and policy recommendations based on the findings.

Adequate and stable power supply serves as the foundation for enterprise production. It not only safeguards employment in local firms but also attracts potential entrants to create additional job opportunities. This paper develops a firm production model incorporating power supply and investigates the employment-stabilizing effects of power supply by utilizing the China urban power supply index and firm-level tax survey data. The findings reveal that power supply significantly stabilizes employment through two primary mechanisms at the firm level: expanding production scale (output effect) and reducing intermediate input procurement (intermediate input substitution effect). At the regional level, the effect manifests mainly through attracting new firm entry and enhancing local employment capacity. Further analysis demonstrates heterogeneous effects across regions, industries, and firm characteristics, with more pronounced impacts observed in economically less-developed regions, non-energy-intensive industries, electricity-intensive firms, small and medium-sized enterprises, and private firms. These findings provide critical policy insights for employment protection and the realization of high-quality economic development alongside employment stabilization goals.

Accurately predicting stock market volatility holds significant theoretical and practical importance for portfolio analysis, financial risk management, and derivative pricing, and has thus attracted extensive attention from scholars. Within the framework of the mainstream HAR-type volatility forecasting models, this paper examines the impact and predictive power of an informed trading index-constructed based on high-frequency trading data and ensemble neural networks – On the daily realized volatility of the A-share market. The study finds that the informed trading index serves as a positive predictor of market volatility. Its predictive capability remains robust across various model specifications and both in-sample and out-of-sample tests, and it performs particularly well during periods of high market volatility and overall economic downturn. Further research confirms that incorporating the informed trading index significantly enhances the out-of-sample forecasting accuracy of existing HAR-type models for market volatility. This not only contributes to the body of research on the market impact of informed trading factors but also provides valuable insights for advancing volatility forecasting modeling and analysis.

Taking the study of national conditions as its perspective and “observing national affairs, studying national conditions, and seeking national policies” as its logical thread, this paper systematically sorts out the origin, constituent elements, and disciplinary nature of statistics. By expounding on the connotations and interrelationships of the three official statistical elements — Data, facts, and policies — Behind “observing national affairs, studying national conditions, and seeking national policies”, it reveals the essential characteristics of statistics as a methodological science that takes data as its research object and uses facts to support policy–making. In light of the complexity of problems faced by human social development and the diversification of data forms in the era of big data, the paper proposes that statistics should pursue a path of interdisciplinary integration. It should redefine relevant concepts, reconstruct the logical framework of statistical analysis, and innovate statistical methods through algorithmic breakthroughs to continuously enhance the ability to analyze and study big data, thereby continuing to play its unique role in national governance, global governance, and the sustainable development of human society.

This paper employs a heterogeneous spatial factor model to accommodate both strong and local dependencies simultaneously. The spatial matrix is approximated by a news co-mention network constructed through textual analysis of business news reports. We introduce relevant methods for the estimation and inference of asymptotic no-arbitrage theory under this framework, applicable to both tradable and non-tradable factors. Empirically, we apply the proposed model and methodology to the listed stocks in the A-share market from 2012 to 2021 and compare the performance of the news co-mention network with other competing asset linkages. Results show that the assets in the A-share market exhibit significant heterogeneity in local dependency, and the news co-mention network outperforms in many aspects due to its quick updating and comprehensive non-industrial connectivity information. Additionally, non-tradable factors have non-negligible effects.

The matching problem between data characteristics and model structures is a critical factor that constrains the application of neural networks in economic and financial forecasting. Existing studies primarily focus on residual analysis, while relatively little attention has been paid to the exploration of model structure stability. To address this research gap, this study proposes a novel research paradigm based on neural network parameter analysis to quantify and evaluate the degree of matching between data and models. A model distance measurement method based on network parameters is designed, and a comprehensive evaluation framework combining model stability testing and residual analysis is established. Empirical results from the foreign exchange market demonstrate a significant nonlinear relationship between the stability of the LSTM model and the length of the lookback period. Among multiple evaluation dimensions, a 36-month lookback period achieves the best performance, effectively balancing the impact of short-term fluctuations and long-term trends. This study not only provides new analytical tools to enhance the reliability of neural network forecasting models but also further deepens the theoretical foundation of machine learning methods in financial applications.

The depth integration of manufacturing and logistics industries is an inevitable trend for promoting industrial upgrading and achieving high-quality economic development. Smart technologies provide an important resource foundation and technical support for this. However, the current level of smartization in the logistics industry remains relatively low, making it difficult to effectively support the flexible production and service extension of the manufacturing industry, which seriously hinders the depth and breadth of the integration of the two industries. Therefore, this paper, from the perspective of the two-way interaction between manufacturing and logistics industries, introduces process factors and conducts empirical analysis based on the improved TOEP-TAM model to explore the influencing factors and mechanisms of smart technology application in logistics enterprises. The research findings show that at the technical level, the superiority of smart technologies significantly promotes the willingness of logistics enterprises to adopt smart technologies, while the security of smart technologies inhibits it; at the organizational level, the scale and strength of logistics enterprises and the mutual trust between the two industries significantly promote the willingness of logistics enterprises to adopt smart technologies; at the environmental level, competitive pressure and government policies significantly promote the willingness of logistics enterprises to adopt smart technologies; at the process level, the customization level of smart logistics significantly promotes the willingness of logistics enterprises to adopt smart technologies. This study provides important decision support for promoting the smart transformation of logistics enterprises and facilitating the in-depth integration of manufacturing and logistics industries.

In the global exploration of central bank digital currency (CBDC), the effectiveness of promotion ultimately depends on whether the public can develop stable and sustainable usage demand. This paper constructs a structural money demand model to characterize the endogenous demand for the digital renminbi (e-CNY) through key features, including remuneration, credit risk, and payment convenience, and estimates the model using microdata from the China Household Finance Survey. The results indicate that the remuneration feature exerts a significant positive effect on potential holding and usage intentions, with interest-rate sensitivity substantially higher than that of other design features. The potential share of the e-CNY in households’ liquid asset portfolios ranges from 2.77% to 60.89%. Further analysis incorporating pilot program data shows that short-term transaction growth is mainly driven by exogenous incentives such as consumption subsidies, while the pattern of high balances and low transaction intensity suggests that endogenous demand has not yet been fully activated. The divergence between theoretical predictions and observed practice reveals a misalignment between exogenous policy interventions and endogenous utility in e-CNY adoption, providing quantitative identification to inform the design and promotion of the digital Renminbi.

Africa is one of the regions most severely impacted by global climate change. However, there are few accurate methods for assessing climate vulnerability in Africa due to data challenges. This paper, based on 3.2million news event texts from 2015 to 2023, uses a large language model to extract climate-related news texts. A fine-tuned climate-themed large language model is employed to identify the impact of climate shock events on dimensions such as the economy, health, and politics. A climate change vulnerability assessment index system for Africa is constructed, and the spatial-temporal variation characteristics of climate change vulnerability in the Great Lakes region of Africa are calculated. The study also tests the effectiveness of the climate vulnerability index using a mixed-frequency dynamic single-factor model. The results show that the climate vulnerability index significantly improves the predictive accuracy of national conflicts. Climate change vulnerability has become a key driving factor for national and regional conflicts in Africa. The climate vulnerability index measurement method proposed in this paper contributes to enhancing the effectiveness of international engagement in climate cooperation and regional conflict prediction in Africa.

Based on the data of A-share listed companies in Shanghai and Shenzhen from 2009 to 2023, this paper empirically analyzes the impact of data element sharing, characterized by the launch of data trading platforms and public data opening platforms, on the efficiency of enterprise resource allocation and its mechanism of action. The findings demonstrate that data element sharing significantly enhances resource allocation efficiency, a conclusion validated through robustness tests including instrumental variable methods, expectation effect exclusion, and dual machine learning approaches. Furthermore, reducing transaction costs and boosting digital technology innovation are key pathways through which data sharing improves resource efficiency. Notably, data sharing proves particularly effective for capital-intensive enterprises, coastal regions, and competitive industries. The study also reveals that data sharing not only elevates resource allocation efficiency within local enterprises but also positively influences neighboring regions, with economic development-related data sharing demonstrating stronger efficiency-enhancing effects. These conclusions provide both theoretical and empirical evidence to support China’s digital economy development and resource allocation optimization.

Major event shocks cause significant fluctuations in macroeconomic variables, leading to a notable decrease in the parameter stability of traditional parameter-driven mixed-frequency dynamic factor models(PD-MFDFM). This severely impacts the stability of the extracted business cycle coincident index, potentially affecting economic situation analysis and macroeconomic policy formulation. To address this issue, this paper proposes a new class of score-driven mixed-frequency dynamic factor models (SD-MFDFM) and provides the corresponding maximum likelihood estimation method, where the distribution can be either normal or Student’s $t$. Particularly, the introduction of heavy-tailed distributions allows the model to determine, in a data-driven manner, whether abnormal fluctuations in macroeconomic variables stem from heavy-tailed noise in the data or signals of economic downturn. This distribution can also handle the instability of model parameters, yielding robust measurements of the business cycle coincident index against extreme events. Simulation studies show that the maximum likelihood estimation performs well under correct model specification, while ignoring the heavy-tailed characteristics of macroeconomic data leads to biased parameter estimates and significant deviations in factor estimation. Finally, using economic data from China and the United States as examples, this paper compares the estimation results of the SD-MFDFM model with traditional methods and extends the model to consider conditional heteroskedasticity. The results reconfirm the advantages of the proposed model and the stability and accuracy of the extracted business cycle coincident index.

We employ the LM-Test and ABD-Test as tools for jump detection to examine the effect of Federal Open Market Committee (FOMC) meetings on cryptocurrency price jumps. The main findings are as follows:First, FOMC meetings increase the likelihood of price jumps in cryptocurrency markets. Second, unscheduled FOMC meetings are more likely to lead to price jumps. Third, smaller-cap cryptocurrencies are more susceptible to price jumps due to FOMC meetings compared to larger-cap cryptocurrencies. Fourth, a classification based on cryptocurrency use cases reveals that currency, smart contract platform, and computing tokens are more prone to FOMC-induced price jumps. Stablecoins, Culture and Entertainment, and decentralized finance (DeFi) tokens also show some response to FOMC meetings, though the effect lacks statistical significance. Finally, by incorporating economic uncertainty and geopolitical risk as environmental variables, this study finds that high economic uncertainty amplifies the impact of FOMC meetings on cryptocurrency price jumps, while high geopolitical risk tends to dampen the effect of FOMC information on cryptocurrency price volatility.

Using city-level panel data from 2002 to 2020 in China, this study quantitatively examines the impact of extreme temperature on the commercial health insurance market using fixed-effects models. The results demonstrate that a one-standard-deviation increase in the number of days with average temperatures exceeding 33$^{\circ }$C leads to a 3.37-percentage-point rise in the growth rate of per capita commercial health insurance premiums, but the impact of the number of days with average temperatures below 0$^{\circ }$C is statistically insignificant. Mechanism analysis reveals that health risk exposure and risk belief channels play dominant roles, whereas income volatility and supply-side channels show statistically insignificant effects. Heterogeneity analysis further indicates that the climate risk effect is more pronounced in regions with higher levels of financial market development, educational attainment, and healthcare infrastructure.

Pledging company shares to financial institutions for financing is a common but high-risk financing method among shareholders of listed companies. Against this backdrop, reducing the proportional risk associated with equity pledge financing and constraining the scale of shadow banking are important and practically relevant research topics. This paper uses firm-level data of listed companies from 2012 to 2023 to examine the effectiveness of the policy regulating asset management businesses of financial institutions — The Guiding Opinions on Regulating the Asset Management Business of Financial Institutions (hereinafter referred to as the “New Asset Management Rules”). It analyzes the transmission mechanisms and heterogeneous effects through which the New Asset Management Rules influence the proportional risk of equity pledge activities. The findings show that the New Asset Management Rules can effectively reduce the proportional risk of equity pledge financing. Mediation analysis reveals that the rules significantly suppress such risk by limiting the scale of shadow banking. Further heterogeneity tests indicate that the impact of the New Asset Management Rules on equity pledge risk varies across regions, ownership structures, and the professional backgrounds of senior executives. It is recommended to strengthen the implementation of the New Asset Management Rules, reduce proportional risk of equity pledge financing and the scale of shadow banking, apply differentiated regulatory measures to different types of firms, moderately broaden financing channels, ease financing constraints, and ensure the stable operation of both the real economy and the capital markets.

Based on the micro data of Chinese high-tech enterprises, this paper studies the impact of US strengthening of high-tech export control on Chinese high-tech enterprise and analyzes the response effect of Chinese monetary policy and fiscal policies. We find that US high-tech export control has negative effect on Chinese high-tech enterprise by restricting the export of chip, industrial mother machines, and advanced manufacturing technology. The implementation of fiscal policies (represented by the National Integrated Circuit Industry Investment Fund) and structural monetary policies (represented by technology innovation refinancing, equipment renewal refinancing, and inclusive small and micro business loan support tools) can actively respond to the negative impact of the US high-tech export control policy. The policy recommendation is to pay attention to the new changes in the US high-tech export control policy, increase the research and development and domestic substitution of advanced chips, industrial mother machines, and advanced manufacturing technologies, increase the support of fiscal policies such as the National Integrated Circuit Industry Investment Fund for Chinese high-tech enterprises and implement structural monetary policies to accurately support Chinese high-tech enterprises.

The heterogeneous input-output (IO) table disaggregates industrial sectors along specific dimensions, enabling the depiction of differentiated production technologies within sectors and revealing important insights that would otherwise be obscured by the assumption of homogeneity. However, the construction of heterogeneous IO tables is constrained by the limited detail in available statistical data, necessitating the use of non-survey techniques, such as proportional assumptions, which introduce significant errors. This creates an inherent trade-off between resolution and accuracy, and the uncertainty resulting from this trade-off has not been adequately assessed due to the lack of necessary reference data for accuracy evaluations. This paper distinguishes between two aspects of accuracy in heterogeneous IO analysis: the accuracy of the heterogeneous IO table and the accuracy of the heterogeneous IO model. It explores the impact of errors introduced by non-survey table construction methods on the accuracy of IO model applications. To address the absence of reference data for accuracy assessment, we propose a method that combines Monte Carlo simulation with TRAS techniques to generate a simulated true value matrix that reflects the structural characteristics of heterogeneous IO tables. By comparing the simulated true values with results from non-survey methods, this study evaluates the accuracy of non-survey techniques across various dimensions of IO model applications. Using the 2020 and 2017 Chinese non-competitive IO table distinguished between domestic and foreign investment as an example, 10,000 simulations were conducted for the elements of the intermediate flow matrix under three distribution scenarios. The distances between the simulated true values and non-survey method results were calculated at different levels of model application, including the direct consumption coefficient matrix, Leontief inverse matrix, output multipliers, and export value-added. The simulations reveal that the accuracy of heterogeneous IO models constructed using non-survey methods significantly improves as the integration level of IO operations deepens. Moreover, the sensitivity of model accuracy to the validity of non-survey assumptions is low. Further regression analysis of the accuracy of matrix or vector elements in the model results shows that diagonal elements of the Leontief inverse matrix exhibit higher accuracy, as do elements corresponding to homogeneous consumption relationships. Additionally, there is a positive correlation between the accuracy of elements and their sensitivity to changes in direct consumption coefficient. These findings provide valuable insights for evaluating the accuracy of heterogeneous IO models constructed using non-survey techniques and offer important guidance for more effective use of model results.

The increasing severity of climate change has led to more frequent adjustments in climate policies, and the uncertainty of these has brought new risks to the production and operation of enterprises. This paper takes China’s A-share listed industrial enterprises from 2007 to 2021 as samples to examine the impact of provincial-level climate policy uncertainty in China (CCPU) on enterprise green total factor productivity (GTFP). The study shows that CCPU has a significant negative impact on enterprise GTFP, and the effect is particularly pronounced among state-owned enterprises and among firms with smaller sizes, low quality of internal control, and low analyst attention. Meanwhile, CCPU can inhibit enterprise GTFP by increasing financing constraints, enhancing environmental regulations, and reducing green investment, while improving enterprise risk-taking capacity can mitigate the inhibitory effect of CCPU on enterprise GTFP. The results of this paper provide a useful reference for policymakers to construct a reasonable climate policy system to guide enterprises’ green transformation.

Against the backdrop of the rapid development of the low-altitude economy and the absence of a well-established statistical accounting system, accurately identifying its industrial structure and economic linkage effects is of important practical significance. Within the existing input-output statistical framework, this study constructs an input-output table for the low-altitude economy based on a sector-splitting approach, disaggregating it into low-altitude manufacturing, low-altitude operations, low-altitude infrastructure and information services, and low-altitude supporting industries, and systematically estimates their industrial linkage characteristics and final demand effects. The results show that the influence coefficient of low-altitude manufacturing remains significantly above one over time, exerting a strong pulling effect on upstream high-technology manufacturing sectors such as communication equipment, general and special equipment, and electrical machinery, whereas the influence coefficients of low-altitude operations, low-altitude infrastructure and information services, and low-altitude supporting industries are generally lower, indicating that their industrial driving effects have not yet been fully released. The sensitivity coefficients of all four low-altitude sectors remain at relatively low levels, suggesting that the low-altitude economy has not yet been widely embedded in the regular production system of the national economy, although the gradual increase in sensitivity reflects the progressive expansion of application scenarios. From the perspective of final demand, the production-inducement coefficients of the low-altitude economy are generally low, with a demand structure dominated by capital formation, while final consumption demand is still in the cultivation stage and the role of exports remains limited. Overall, the low-altitude economy exhibits typical characteristics of an emerging industry, namely relatively strong supply-side driving capacity but insufficient demand-side penetration, and the industry as a whole is still in a development stage dominated by investment-driven construction and application-scenario incubation. This study provides a new statistical basis for the quantitative evaluation of the low-altitude economy and offers a methodological reference for input-output measurement of emerging integrated industries.

This paper examines how the horizon orientation of capital markets shapes real economic outcomes, focusing on firms’ environmental, social, and governance (ESG) performance in the context of China’s stock market. Building on insights from market microstructure theory, the study proposes a novel measure of stock price horizon orientation to capture the extent to which market prices reflect long-term versus short-term fundamentals. Empirically, we find that as stock prices become more forward-looking, firms exhibit lower ESG performance. Further analysis suggests that long-horizon prices enhance managerial learning by conveying richer information about long-term fundamentals, thereby reducing managers’ incentives to engage in ESG activities as a form of precautionary risk hedging. These findings underscore the dual role of capital markets — as both monitors of corporate behavior and providers of information — and offer new insights into how financial market shape firms’ ESG performance.

With the rapid development of generative artificial intelligence technology, the readership of corporate annual reports has expanded from human readers to artificial intelligence (AI), making the limitations of traditional readability measurement methods increasingly evident. This study constructs an AI readability index for annual reports under large language models from three dimensions — Long-text capability, table parsing, and other factors — Using random sampling Q&A and a random forest model. Based on an analysis of companies listed in China’s mainland from 2004 to 2023, the results show that: 1) Significant differences exist between the AI readability index and traditional readability measures in terms of time trends, numerical distribution, industry characteristics, and regional distribution. 2) The AI readability of different types of companies exhibits heterogeneous performance, which is opposite to traditional readability trends. Non-state-owned enterprises, high-tech industries, and companies with higher accounting information quality demonstrate significantly better AI readability in their annual reports compared to others. 3) Compared to traditional readability, AI readability more effectively improves corporate information transmission efficiency. This study innovatively proposes the concept and measurement method of AI readability for annual reports, providing a new perspective and approach for readability research in the context of large language models, thereby enhancing the understanding of corporate information disclosure motivations and economic consequences. It also offers insights for companies to adjust their annual report writing paradigms to make them more AI-friendly, while providing important references for regulators and investors to accurately evaluate corporate information disclosure levels under large language models.

Can the new generation of AI reconstruct the global value chain of service industry? By introducing CES production function and Dixit Stiglitz model framework, based on the transnational input-output and service industry data from 2010 to 2024, this paper uses the text mining method to construct the development level index of the new generation of artificial intelligence, calculates the development level of the new generation of artificial intelligence combined with the dictionary of the development plan of the new generation of artificial intelligence, and discusses its mechanism of action on the status of the global value chain of the service industry combined with the production decomposition framework of the forward and backward associated global value chain (WWYZ2017). The research shows that the global value chain status index and resilience index will increase by an average of0.101 units and 0.960 units respectively for each unit increase in the level of the new generation of artificial intelligence. The new generation of artificial intelligence has significantly driven the rise of China’s service industry’s position in the global value chain and enhanced its resilience through three core channels: industrial intelligence agglomeration, intelligent innovation efficiency and information transparency. The expansion study found that the positive role of the new generation of AI in promoting the status and toughness of the global value chain was more prominent in non AI pilot areas and labor-intensive industries. Further analysis showed that there were clear multiple threshold constraints in capital, labor and knowledge intensive service industries. For the status of the global value chain, the new generation of AI needed to cross a high technology and capital threshold to fully release the momentum of service value upgrading; For the resilience of global value chain, it also shows certain threshold constraints. Only when the corresponding factor accumulation threshold is reached, the toughness enhancement effect is significantly displayed. This study provides multiple evidences for understanding the development path and value upgrading of service industry in the era of artificial intelligence.

In the globalized economic system, supply chain resilience is crucial for the stable and sustainable development of enterprises and the security and stability of the domestic economy. In the current booming development of new quality productive force, whether and how to strengthen supply chain resilience is an important research topic. This article is based on the data of A-share listed companies in the new energy vehicle industry chain from 2015 to 2022, and uses a two-way fixed effect model to empirically test the impact of the development of new quality productive force on the resilience of enterprise supply chains. The research results indicate that the improvement of new quality productive force in enterprises has a significant promoting effect on the resilience of the supply chain, and the conclusion still holds after alleviating endogenous problems and undergoing robustness testing. Heterogeneity analysis shows that the improvement of new quality productive force in enterprises has a more significant promoting effect on supply chain resilience in high-tech enterprises, small enterprises, and midstream enterprises in the new energy vehicle industry chain. Mechanism analysis has found that the development of new quality productive force in enterprises promotes the improvement of supply chain resilience through two channels:Reducing operating costs and enhancing bargaining power. The research findings enrich existing research on new quality productive force and provide useful empirical evidence for enterprises to enhance supply chain resilience and promote high-quality development through new quality productive force.