Japan's first comprehensive AI legislation places "promotion" at its centre. The orientation itself is not unique, since the United Kingdom and Singapore have adopted comparable postures. What distinguishes Japan is the social conditions under which it made this choice. The urgency of Japan's governance problem comes from the economic and social cost of AI going unused. Governance is being pursued in a society where adoption is low, public trust is likewise low, and an ageing population coincides with ageing industrial IT systems. In the year since the Act was passed in 2025, personal use and firm-level trial have both risen sharply, while deployment within firms remains sharply stratified by firm size. The government has updated its Basic Plan accordingly, placing its emphasis on organisational change. Promotional instruments reduce the cost of trying. They do not supply the capital, personnel and managerial authority that organisational integration requires. Japan pairs state coordination with voluntary compliance rather than binding regulation, its market does not confer the leverage that EU-style rule export depends on, and its industrial structure does not support the self-organising innovation on which the American model rests. This governance model has been shaped by Japan's particular policy problem and policy environment, and the difficulties it faces differ accordingly.
Modern societies are confronted with an unprecedented spectrum of complex risks, spanning geopolitical tensions, global meteorological disasters, environmental degradation, supply chain disruptions, financial instabilities, and technology mediated societal and institutional challenges. These risks are global in scope, systemic in nature, and characterized by deep uncertainty, with far-reaching impacts across multiple domains. This review summarizes key research progress on different risk categories and highlights emerging methodologies such as interdisciplinary integration, big data and artificial intelligence, system dynamics, network modeling, and scenario simulations. Effective responses typically follow a cyclical process of identification, assessment, management, and adaptive monitoring, tailored to the specific attributes of each risk. The findings emphasize the critical importance of global cooperation and policy coordination. Future research should focus on multidimensional risk assessment, the integration of emerging technologies, and international collaboration to develop forward-looking strategies for navigating increasingly intricate global challenges.
Health insurers must hold adequate reserves to cover healthcare costs during epidemics; however, traditional reserving methodologies, which are typically grounded in historical averages, are ill-suited to capturing the rapid, transmission-driven dynamics of pandemic disease spread. To address this limitation, we propose a framework that explicitly models infection duration-referred to as infection age-which critically influences the progression of medical expenditures from initial diagnosis through hospitalization and intensive care. Our approach structures infection and payment data into a triangular framework that enables projection of future liabilities for infected individuals requiring ongoing treatment. The analysis demonstrates that key epidemiological quantities, particularly the effective reproduction number, fundamentally shape reserve trajectories, premium adequacy, and ultimate cost ratios. A numerical illustration based on COVID-19 dynamics validates the framework and incorporates sensitivity analysis. The principal contribution of this work is establishing a formal linkage between epidemiological modeling and financial risk quantification, providing insurers, public health institutions, and regulators with a systematic method for incorporating real-time outbreak data into reserve estimation and stress-testing. This enhances financial preparedness and bolsters resilience within healthcare systems during global health emergencies.
The mechanism and optimal size of network formation remain challenging. We consider a simple interbank network and show that the network formation is characterized by banks’ trade-offs between profitability and survivability when maximizing risk adjusted expected profits, and between risk-sharing and free-riding when making reserving decisions. Depending on liquidity risk adjusted profit margin and bank heterogeneity, banks collude for risk-sharing but compete for free-riding. In Nash equilibrium network, a shifting dominance from risk-sharing to free-riding generates a hump-shaped bank’s expected profits in the number of banks, resulting in relatively small Pareto optimal interbank networks. The findings provide new insights into “too-big-to-fail” and “too-interconnected-to-fail”.
In an interview, renowned global investor Jim Rogers shares his philosophy of contrarian thinking, risk management, and life lessons from decades of investing. He stresses that true success requires independent judgment and the courage to act against consensus—“looking out the window” rather than at market screens. Rogers warns of mounting U.S. debt and the dangers of complacency, while remaining cautiously optimistic about China’s long-term resilience and India’s emerging potential. Skeptical of fiat and cryptocurrencies, he favors tangible assets like gold and silver. Rogers addresses disruptive technologies such as artificial intelligence, predicting transformative impacts akin to past industrial revolutions, yet cautioning investors to balance optimism with timely exits. Despite his global investment achievements, he considers his two daughters to be his greatest investment—more rewarding than any financial venture. The interview highlights the enduring importance of independent thinking, rigorous research, and provides a nuanced perspective on risk, conviction, and opportunity in an era marked by rapid technological change and geopolitical volatility.
This paper discusses the foundation of the science(s) of risk. A key issue addressed is the question of whether there is a science of risk—or several sciences. The recent launch of a journal titled Risk Sciences appears to endorse the plural view. This contrasts with the evolution of risk analysis over the past 40–50 years as a distinct discipline and science—one that encompasses foundational issues alongside the assessment, perception, communication, and management of risk—with the Society for Risk Analysis (SRA) and its members as primary contributors. The primary aim of this paper is to demonstrate that the SRA perspective provides a solid foundation for these studies and activities through its distinction between generic risk analysis (science) and applied risk analysis (science). While applied risk analysis/science is typically multidisciplinary or interdisciplinary, a distinct risk science exists, comprising the most justified knowledge on risk analysis as covered by the risk analysis field/discipline—as well as the practices for generating this knowledge.
The econometric theory and methods of financial risk are critical technical issues that consistently concern both the economic and mathematical communi- ties. Traditional risk measurement theories and methods are based on classical probability and statistics. However, frequent financial crises have demonstrated significant flaws in conventional risk measurement theories and methods, which cannot accurately depict the risks and uncertainties in incomplete financial mar- kets. Instead, the nonlinear expectation seems more suitable when depicting the uncertain phenomenon in economics and finance. This paper will introduce the latest developments in this field, including nonlinear expectation, the non-linear central limit theorems, nonlinear normal distributions, and limit theorems in the quantum realm.
The arid and semiarid pastoral and agropastoral systems across sub-Saharan Africa face increasing pressure from climate change, land degradation, and changing land use, endangering food security, livestock productivity, and ecosystem health. However, specific evidence on how these combined pressures interact locally is still limited. This study addresses this gap with a long-term case study of Tiaty, Baringo County, Kenya, where pastoral and agro-pastoral livelihoods are increasingly influenced by climate variability and land-use changes.Landscape dynamics from 1994 to 2024 were examined using Landsat imagery, CHIRPS and CRU TS climate records, agroecological data, and socioeconomic surveys. Land-use/land-cover changes were mapped using random forest classification and change detection in Google Earth Engine. Results revealed a marked shift from dense shrublands to sparse shrublands, grasslands, and croplands. Rainfall during the March–May season declined while heatwave frequency increased, altering farming and grazing practices. Goats and camels showed greater resilience than cattle, while settlements shifted farther from croplands, reflecting expanding rangeland.This studydepended on gridded climate datasets that may obscure microclimatic differences. The study also employed limited temporal observations that could hinder causal inference, and used spectral similarity among transitional land-use classes in medium-resolution imagery which may have introduced uncertainty in area estimates. The observed relationships therefore, should be seen as indicators of system responses rather than definitive causal effects.These findings underscore the need for targeted adaptation strategies, including drought-tolerant crops, climate-resilient livestock breeds, and sustainable rangeland management, supported by policies strengthening market access, irrigation infrastructure, and community capacity.
China is particularly prone to severe earthquake disasters. The Chinese government pays great attention to the monitoring and forecasting of earthquakes, the seismic fortification of engineering structures, and emergency rescues and resilience safety. This requires support from the academic community in the form of seismic risk management. In this study, the Chinese literature on famine relief and seismic risk management until the end of 2024 was analyzed and visualized using CiteSpace software. The progress of research and the current situation were tracked, and the relationship between seismic risk management and seismic resilience was discussed. According to the bibliometric analysis, the famine policy and relief measures in the Yuan and the Qing dynasties were the primary focus of recent research on famine relief. On seismic risk management, research trends in the last decade included disaster management, earthquake vulnerability, emergency management, and disaster risk management. The content of seismic risk management and seismic resilience partially overlapped, and there was more content related to the former than the latter. Moreover, the intensity of research declined in the most recent three years. The current research areas of interest mainly involve functional loss, performance assessment, and recovery strategies. Research on collaborative management promoted the formation of a more comprehensive management system. The progress of research on seismic risk management, traced through the Chinese literature, has produced clear development paths and state-of-the-art techniques for researchers and practitioners.
This study evaluates the overall risk exposure of manufacturing small- and medium-sized enterprises’ supply chains (SMESC) during plausible disruption scenarios, particularly in emerging economies where disruptions can have severe consequences. It develops a holistic, systematic, and quantitative framework to empower SMEs to assess and manage supply chain risks (SCR) effectively, thereby enhancing resilience and ensuring business continuity. A comprehensive literature review and expert consultations were undertaken to identify potential hazards. An integrated Analytic Hierarchy Process (AHP)-Hazard Identification and Risk Assessment (HIRA) methodology was employed, where AHP determined hazard weights and HIRA evaluated overall risk levels. The framework was illustrated through a case study of a manufacturing SME in India. The results revealed that SMESC are highly vulnerable to disruptions, reflected by an overall risk score of 63.476%. Key hazards included the scarcity of raw materials, distribution network breakdowns, and inventory stockouts, with procurement and production activities being particularly susceptible. The findings offer critical insights for managers and policymakers to proactively manage risks and bolster the resilience of SMESC in emerging economies. This research addresses existing gaps by proposing a structured and problem-driven integration of AHP and HIRA, enabling SMEs to quantify and prioritize both internal and external SCR. The combination of expert-driven weighting (AHP) and scenario-based risk scoring (HIRA) offers a practical decision-support framework suiting contexts with limited historical data and high uncertainty.
This article focuses on Rao’s damage model and a possible Markovian version of this model. The novelty of the formulation lies in the fact that the conditional survival distributions are defined from convolution-like identities satisfied by Appell-type polynomial families. Attention is first paid to probability distributions, some of which are non-standard, that can be generated in this way. A dynamic extension of the model is then proposed in the form of a non-stationary Markovian process that stops at the end of the damage process. The distribution of the final undamaged part is obtained by a martingale reasoning and using a family of Abel-Gontcharoff (A-G) pseudopolynomials. An application to epidemic modeling is presented, and the generalization to a splitting damage model is also provided. Finally, a few numerical examples illustrate some of the results.
In this study, we examine the impact of big data and corresponding prediction techniques on insurance risk assessment. We demonstrate that both big data and the Least Absolute Shrinkage and Selection Operator (LASSO) approach, a modern predictor selection technique, effectively improve insurance risk assessment accuracy. In an average of proxies, the out-of-sample health risk prediction accuracy improves by 106 %, in which big data, in addition to traditional insurance policy and demographic information, contributes 82.5 %, and the LASSO model contributes 17.5 %. We also demonstrate that big data obtained from smartphone use offers extra-predictive power in addition to past medical histories. We employ Adaptive Group LASSO to determine that the most fruitful data collection sources for health insurance underwriting include personal digital devices, recent travel experience, and insureds’ credit records.
In this paper, we present axiomatic characterizations of certain simple risk-sharing (RS) rules, such as uniform, mean-proportional and covariance-based linear RS rules. These characterizations facilitate a clearer understanding of the principles underlying the application of these rules; they typically include maintaining some degree of anonymity regarding participant data and/or incident-specific data, adopting non-punitive processes and ensuring the equitability and fairness of RS. By formalizing key concepts such as the reshuffling, source-anonymous contributions and strongly aggregate contributions properties, along with their generalizations, we develop a comprehensive framework that clearly expresses these principles and defines the relevant rules. To illustrate, we demonstrate that the uniform RS rule, a simple mechanism in which risks are shared equally, is the only RS rule that satisfies both the reshuffling and source-anonymous contributions properties. This straightforward axiomatic characterization of the uniform RS rule serves as the foundation for exploring similar principles in two broad classes of RS rules, which we call q-proportional RS rules and (q1, q2)-based linear RS rules, respectively. The framework also allows us to introduce specific new RS rules, such as scenario-based RS rules.
We present a general non-parametric statistical inference theory for integrals of quantiles without assuming any specific sampling design or dependence structure. Technical considerations are accompanied by examples and discussions, including those pertaining to the bias of empirical estimators. To illustrate how the general results can be adapted to specific situations, we derive – at a stroke and under minimal conditions – consistency and asymptotic normality of the empirical tail-value-at-risk, Lorenz and Gini curves at any probability level in the case of the simple random sampling, thus facilitating a comparison of our results with what is already known in the literature. Results, notes and references concerning dependent (i.e., time series) data are also offered. As a by-product, our general results provide new and unified proofs of large-sample properties of a number of classical statistical estimators, such as trimmed means, and give additional insights into the origins of, and the reasons for, various necessary and sufficient conditions.
Under current regulations, financial institutions are required to estimate the daily Value-at-Risk (VaR) or Expected Shortfall (ES) of their trading positions. Despite being widely studied and adopted, both risk measures have theoretical limitations: VaR is not coherent and ES is not elicitable. Expectile, the only law-invariant risk measure that is both coherent and elicitable, has gained considerable interest in both risk management and statistics recently. However, the backtesting of expectile has not yet received adequate attention, and existing backtests tend to suffer from size distortion or low test power. This paper proposes novel unconditional and conditional backtests for expectile. Simulation studies show that the proposed tests exhibit promising finite sample size performance. In addition, in an empirical study, we apply the proposed tests to S&P 500 data.
Deep learning is a powerful tool whose applications in quantitative finance are growing every day. Yet, artificial neural networks behave as black boxes and this hinders validation and accountability processes. Being able to interpret the inner functioning and the input-output relationship of these networks has become key for the acceptance of such tools. In this paper we focus on the calibration process of a stochastic volatility model, a subject recently tackled by deep learning algorithms. We analyze the Heston model in particular, as this model's properties are well known, resulting in an ideal benchmark case. We investigate the capability of local strategies and global strategies coming from cooperative game theory to explain the trained neural networks, and we find that global strategies such as Shapley values can be effectively used in practice. Our analysis also highlights that Shapley values may help choose the network architecture, as we find that fully-connected neural networks perform better than convolutional neural networks in predicting and interpreting the Heston model prices to parameters relationship.
With the evolution of empirical asset pricing theory and the increasing complexity of markets, the exploration of innovation - driven alternative factors has attracted growing attention. This study integrates network structure analysis methods into textual analysis to innovatively expand the concept of financial system vulnerability, introducing the notion of emotional vulnerability. This study employed the Moka massive mixed embedding (M3E) model to embed textual sentiment data to construct individual stock sentiment networks. The Ricci curvature of these networks was calculated to quantify emotional vulnerability factors, which were effectively demonstrated to account for excess weekly returns through rigorous validation using portfolio permutation tests, factor redundancy checks, and time series regression analyses. These factors maintain significant explanatory power even when adjusted for the Fama–French five factors. Emotional vulnerability factors enhanced asset risk assessment accuracy, improved market trend predictions, and increased investment decision-making precision and efficiency.
This article articulates a conceptual framework for risk sciences—an emerging interdisciplinary field dedicated to the study of risk and uncertainty across natural, social, economic, management, technological and other domains. We propose a three-dimensional structure for understanding risk sciences: (1) identification and assessment (systematic recognition, evaluation, and modeling of risks); (2) mechanism and strategy (functional tools and governance plans for risk intervention); and (3) behavior and decision (psychological, cognitive, and strategic responses to uncertainty). Using this framework, we survey critical research areas and identify key trends shaping the field’s future. The article concludes by introducing the new journal Risk Sciences. Its core hypothesis is that interdisciplinary synthesis can unify disparate risk research across fields, creating a knowledge base applicable to diverse risks. By bridging disciplines and fostering innovation, risk sciences build societal resilience in an era of escalating complexity.
