Dynamic risk evaluation in insurance and finance is typically formulated on fixed integrability classes, such as L^p spaces or Orlicz hearts generated by a single Young function. A uniform tail regime is thereby imposed across scenarios, which is difficult to reconcile with actuarial portfolios in which loss severity and admissible model perturbations depend on observable risk factors. Regional catastrophe exposures and lapse or surrender behaviour under stress are concrete instances in which tail control should vary locally with the realised state.
Admissible terminal payoffs at time t are modelled on a conditional Musielak–Orlicz heart $M^{\Phi_{S_{t}}}(\mathcal{F}_{T}\mid\mathcal{F}_{t})$ generated by an $\mathcal{F}_{t}$-measurable random Young function driven by an adapted state process S. Dynamic evaluation is developed in the stopping-time two-parameter sense on adapted càdlàg processes, using the $L^{0}$-module viewpoint and the conditional expectation pairing to handle scalarisation cleanly in conditional duality.
A robust dual representation is obtained on state-dependent hearts as an essential supremum of penalised conditional expectations over a state-indexed dual family. Time consistency is characterised by pasting the stability of the dual classes and, in the convex case, by a cocycle property of the minimal penalty. The $\sigma$-order continuous conditional Köthe dual is identified and supports a stable weak topology under which attainment of a worst-case measure is obtained. A discrete-time Snell envelope on varying hearts follows from a backward-admissibility mechanism controlled by a conditional modular bound; it yields a minimal dominating $\mathcal{E}$-supermartingale and a first-hitting optimal exercise rule.
The construction integrates state-dependent tail control with time-consistent dynamic programming for unbounded rewards, providing a rigorous foundation for actuarial and financial evaluation under heterogeneous tail regimes. It covers situations where classical fixed-integrability frameworks are restrictive, and it supports explicit dynamic-programming implementations on the admissible state-dependent domain.

This study examines the application of machine learning techniques to predict policyholder renewal behavior in motor vehicle insurance. Accurately identifying customers likely to lapse is crucial for pricing strategies and customer retention in actuarial practice. Using a dataset of motor insurance policies, three classification models, Logistic Regression, Random Forest, and LightGBM, were developed and compared. Exploratory analysis revealed a moderate class imbalance, with approximately 79.6% renewal and 20.4% lapse observations. Feature engineering was performed to construct variables such as age and driving experience. The models were evaluated using multiple performance metrics, including accuracy, precision, recall, specificity, F1-score, and the area under the ROC curve (AUC). The results show that Logistic Regression achieved the highest accuracy (79.4%) and recall (99.5%), but exhibited extremely low specificity (2.1%), indicating poor performance in identifying lapse cases. Random Forest provided a more balanced performance, with an AUC of 0.663 and improved specificity (7.1%), though still limited. LightGBM achieved the best overall discrimination ability, with the highest AUC (0.683) and a more balanced trade-off between recall (63.0%) and specificity (63.3%), despite lower overall accuracy. These findings suggest that while traditional models, such as Logistic Regression, may perform well on aggregate metrics, they can be misleading in imbalanced insurance datasets. Ensemble methods, particularly gradient boosting, offer superior capability in capturing complex patterns and improving classification balance. The study highlights the importance of using appropriate evaluation metrics beyond accuracy and demonstrates the practical relevance of machine learning methods in actuarial modeling and policyholder retention analysis.

Factor momentum has attracted significant attention in the recent asset pricing literature, as it has been proposed as one of the potential mechanisms through which the empirical puzzle of the momentum anomaly manifests itself in financial markets. Factor momentum exploits the tendency of anomalies that have recently exhibited strong performance to continue outperforming, and of anomalies that have recently underperformed to continue doing so, in a manner analogous to cross-sectional price momentum. While the literature has devoted considerable effort to examining this phenomenon in equity markets, the question of whether a similar mechanism operates in the corporate bond market remains largely unexplored. This market represents a particularly relevant testing ground, as it has been at the center of a broader debate regarding the replicability of anomalies proposed to explain the cross-sectional variation in corporate bond returns.

Leveraging a novel, manually cleaned, and error-free dataset that consolidates information from the most widely used databases in the literature, including TRACE among others, we investigate whether a factor momentum premium exists in the corporate bond market, examining both its cross-sectional and time-series manifestations across factor subsamples comprising up to more than 140 corporate bond anomalies.

Our results show the following: (i) for certain factor subsamples, both cross-sectional and time-series factor momentum generate positive and statistically significant raw payoffs with modest economic magnitude of approximately 2.60% per year; (ii) these payoffs survive transaction cost adjustments and standard factor model risk exposures; (iii) they are not subsumed by traditional price momentum, suggesting they capture a distinct source of return variation.

This study contributes to the corporate bond literature by providing evidence of a factor momentum effect, challenging the notion that momentum is a pervasive feature of all cross-sectional anomalies, and contributing to the growing literature on factor momentum.

Introduction

This work introduces a novel risk-sharing rule called serial risk sharing (SRS), inspired by the serial cost-sharing method of Moulin and Shenker (1992, Econometrica 60(5), 1009–1037). Using a serial mechanism, which involves a quantile-based decomposition of initial losses, and a uniform allocation of the resulting components to the relevant agents, we develop an equitable and intuitive risk allocation tailored for heterogeneous losses, serving as a computationally efficient alternative to other risk-sharing rules like conditional mean risk sharing (CMRS) due to the availability of a closed-form formula. In addition to numerical illustrations, the analytical properties of SRS are shown to consolidate its theoretical foundation.

Methods

The quantile-based decomposition of losses is constructed via a novel concept of quantile equivalent loss, defined via the probability levels of agents’ initial losses. By probability and quantile transformations, the distributional properties of the resulting components remain tractable, enabling theoretical analysis of the behavior of SRS. Particularly, when the losses are ordered by first-order stochastic dominance, substantial simplification of the rule is available, further facilitating the analysis of the rule and improving its transparency.

Results

Various theoretical properties of SRS are shown, such as actuarial fairness, full allocation, scale invariance, monotonicity, risk fairness, and convex-order improvement, under suitable assumptions. Additionally, through numerical studies, we demonstrate that SRS exhibits good properties and performs well relative to other risk-sharing rules. The numerical results also reinforce our theoretical findings.

Conclusion

By adapting the serial cost-sharing method put forward by Moulin and Shenker (1992) to the risk-sharing context, we design a novel risk-sharing rule known as serial risk sharing, where losses are decomposed into components based on quantiles. It carries a clear and interpretable structure, with favorable theoretical properties and supportive numerical results.

This work studies the problem of shortfall risk minimization in incomplete financial markets where the payoff of a contingent claim depends on a non-tradable asset. The trading is allowed only in a correlated tradable asset. This setting naturally leads to market incompleteness and makes perfect hedging impossible.

We study the problem in a continuous-time Brownian framework under different informational structures. First, we consider the case where the non-tradable asset is fully observable. In this setting, the optimal trading strategy is derived using stochastic calculus techniques, including Itô’s formula and a change of probability measure. The resulting strategy incorporates a term depending on the market price of risk and a term reflecting the dependence of the payoff on the non-tradable asset.

Second, we study the case where the non-tradable asset is unobservable until the terminal time. Thus, the problem is reformulated by conditioning on the hidden factor. As a result, an optimal strategy is expressed in terms of conditional expectations with respect to this hidden factor. The conclusions demonstrate the impact of information availability on hedging performance.

The approach is further extended to a mixed model driven by both Brownian motion and fractional Brownian motion, with the Hurst parameter H>3/4, ensuring that the model remains within the semi-martingale framework.

Systemic financial risks, arising from macroeconomic fluctuations, market volatility, policy shifts, and international economic events, present significant challenges for investors, particularly in rapidly evolving and highly interconnected markets such as China. Despite substantial research on risk perception and investor behavior, a comprehensive and psychometrically rigorous instrument specifically designed to measure investors’ sensitivity to systemic risks has been lacking. This study aims to fill this gap by developing the systemic risk sensitivity scale for Chinese investors, grounded in financial theory and designed to assess the impact of systemic risks on investment behavior. An empirical study was conducted using structural equation modeling method, which demonstrated the scale’s methodological rigor, high reliability, and strong internal consistency. Moreover, the scale exhibited excellent cross-group consistency across different types of investors, suggesting its applicability across diverse investment populations. Robustness tests further confirmed the stability and validity of the results, highlighting the practical and theoretical soundness of the instrument. This study thus develops the systemic risk sensitivity scale for Chinese investors, grounded in financial theory and behavioral finance principles, aimed at assessing how systemic risks affect investment decisions. The scale not only provides valuable insights for portfolio management, investor risk assessment, and financial policy formulation but also contributes to the behavioral finance literature by offering a validated framework for understanding investor decision-making under systemic risk conditions in China’s dynamic financial market.

This research investigates the divergence between subjective risk perception and objective actuarial risk during the "Iron Swords" war in Israel. As life expectancy continues to rise and demographic structures shift, the financial resilience of individuals and the stability of national social security systems depend heavily on accurate risk assessment and adequate insurance coverage.

The study utilizes a quantitative analysis of survey data (N=442) categorized into three cohorts: combat reservists, their immediate family circles, and the general public. We employed logistic regression models to identify predictors of life insurance purchase intentions and evaluated financial literacy levels across these groups.

The results reveal a significant "Knowledge Illusion" among high-risk populations. Combat reservists demonstrated a profound gap between their perceived and actual insurance literacy (P < 0.01), suggesting that those at the highest objective risk are the most prone to behavioral biases. Furthermore, the findings identify a "Neighbor Effect," where purchase intentions are driven more by social proximity to risk than by individual exposure.

Significance: From an actuarial and policy perspective, these behavioral gaps represent a systemic risk to the National Insurance Institute (NII) and the broader pension system. When high-risk cohorts remain under-insured due to cognitive biases, the long-term financial burden shifts to state-funded safety nets. We propose that addressing these gaps requires the integration of Automatic Adjustment Mechanisms (AAMs) and default enrollment schemes. Such interventions are essential to ensure the actuarial stability of the social security framework, especially as the system faces the dual challenges of prolonged geopolitical instability and an aging population.

In the transition to a low-carbon economy, the difference in greenium between twin bonds with different maturities is expected to vanish, or at least to decline in both level and volatility. This motivates the introduction of stochastic models for the greenium term-structure slope subject to a terminal transition constraint. Empirically, this slope exhibits mean-reverting behavior together with substantial changes across volatility regimes. To capture these features, we introduce two related models.

The first one is a Regulatory Deadline-Constrained Model (RDCM), namely a linear mean-reverting diffusion with a deterministic terminal date at which the greenium difference is forced to vanish. The second is a Switching Regulatory Deadline-Constrained Model (SRDCM), where the perceived transition deadline is regime-dependent and evolves according to a discrete-time latent Markov structure. In both cases, the model is formulated so that the transition date affects not only the terminal condition, but also the pre-terminal behavior of the drift target and diffusion coefficient through the time remaining to transition.

For the RDCM, we derive the exact Gaussian bridge likelihood and study its calibration on fixed observation grids. We then show that, under a fixed-horizon infill asymptotic scheme, the diffusion block can be consistently identified on the part of the observation interval where it is asymptotically visible. This result yields a structurally grounded local temporal discrimination rule between competing transition labels in the switching framework. The models are calibrated using data from twin German government bonds. In the empirical analysis, the evidence suggests that, in the most recent period, market perception has shifted toward a slower transition path and a delayed convergence of the greenium term structure.

Generalised Linear Models are a core tool for modelling non-normal data and are routinely fitted via the Iteratively Reweighted Least Squares algorithm. While this algorithm is computationally attractive, its performance can be hindered by convergence issues, sensitivity to starting values, and substantial estimation error. This paper develops a practical enhancement of the Iteratively Reweighted Least Squares algorithm by replacing the standard least squares step with Stein-type shrinkage estimators. These estimators reduce the theoretical mean squared error by introducing a controlled bias that leads to a significant variance reduction, without requiring cross-validation or increasing computational complexity. As a result, the proposed approach provides a scalable and efficient alternative to commonly used penalised generalised linear modelling. In addition, we propose an optimisation-based strategy for selecting starting values, which improves the stability and convergence of both standard and shrinkage-based Iteratively Reweighted Least Squares implementations. The paper is not primarily methodological; instead, it focuses on the practical deployment of generalised linear modelling. Extensive numerical experiments, covering a wide range of settings relevant to practitioners, together with real-data applications, demonstrate consistent improvements in accuracy, stability, and computational efficiency over the industry standard benchmarks.

This study investigates how conflict-related public attention, measured through Google Trends search data, shaped equity market dynamics in Israel during the 2023–2025 Israel–Hamas war. Search activity is disaggregated along two dimensions, the language of the query (English vs. Hebrew) and the geographic origin of the search (worldwide vs. domestic), producing eight attention indicators constructed from the terms "War in Israel" and "Hamas". The analysis employs an EGARCH(1,1) framework to capture asymmetric volatility responses and Granger causality tests to assess the predictive content of search-based attention at varying time horizons. The results reveal a clear linguistic asymmetry in how conflict-related information reaches asset prices. Hebrew-language searches exert a statistically significant negative contemporaneous effect on TA-35 returns, consistent with the rapid pricing of localized anxiety among domestic investors. English-language searches, by contrast, produce no significant same-week market response but carry robust predictive power at a two-week horizon, particularly domestic English searches for "War in Israel", which reject the null hypothesis of no Granger causality at the 1% level. In every specification, causality runs strictly from public attention to market returns, with no evidence of the reverse. A robustness check based on PCA-derived composite indices confirms both findings: the contemporaneous Hebrew war-specific effect and the lagged English predictive effect survive the compression of each linguistic channel into a single index. These findings suggest that the same conflict generates two distinct information transmission paths, one fast and local, driven by immediate threat perception, and one slow and global, reflecting the gradual accumulation of geopolitical awareness among internationally oriented investors. The results highlight that what moves markets during geopolitical crises is not simply the volume of public attention, but its linguistic and geographic origin.