The objective of this paper is to demonstrate that traditional statistical problems in credit scoring can be solved efficiently by implementing a quantum-enhanced version of the traditional Support Vector Machines algorithm. Three significant case studies are presented, involving regression, dichotomous and multi-class classification problems in the field of credit risk management. The analysis highlights the promise of the quantum kernel method as a competitive solution especially for multiclass problems with strongly structured features. Quantum-enhanced models tend to outperform classical approaches because they can capture more intricate non-linear relationships between variables, partly due to their ability to leverage entanglement. This advantage is particularly evident in credit risk assessment, where it can improve the credit evaluation process and help reduce credit losses for financial and insurance institutions
The level of environmental vulnerabilities is elevated in Sub-Saharan Africa. This study adopted an African perspective to evaluate the impact of capital market development on environmental sustainability, drawing on inclusive theoretical and empirical constructs of the nexus between the two. Beyond carbon emissions as a proxy for environmental quality, the study added proxies for natural resources and forest depletion to examine the relationship. The sample consisted of 10 countries, and data were collected between 1993 and 2022. Econometrically, we applied panel-estimated generalized least squares, cross-section seemingly unrelated regression with panel-corrected standard errors, dynamic ordinary least squares, and fully modified ordinary least squares to investigate the static and dynamic relationships between the variables. Our observation was that the impact of stock market development on environmental sustainability depends on the proxy used to measure environmental sustainability. By and large, we observed that, on the one hand, stock market development significantly increases carbon emissions and, on the other hand, reduces forest and natural resource depletion. This study recommends that stock exchange regulators strictly enforce mandatory environmental impact disclosures and financial reporting regulations to ensure compliance by listed firms. For instance, regulators should require listed extractive industry firms to provide environmental accountability as part of their annual financial reporting, ensuring continuous compliance and ongoing monitoring of their environmental impacts. The issuance of climate-backed financial securities, such as green and brown bonds, is recommended to encourage firms to fund their green projects and to help firms with high emission rates transition to environmentally sustainable production methods
This paper provides an empirical assessment of the implementation of the new IRRBB and CSRBB regulatory framework approximately one years after its introduction, based on evidence from an AIFIRM survey on 25 Italian banks. The analysis covers three reporting dates (31 December 2023, 31 December 2024 and 30 September 2025) and suggests that the evidence is consistent with a progressive adaptation of banks' IRRBB exposures following the introduction of the new regulatory framework. Furthermore, across the three reporting dates, exposure levels differ across banks depending on their size and the measurement approaches adopted. IRRBB governance practices appear to be progressively consolidating in some areas, including internal capital calibration and the definition of RAF thresholds, while remaining heterogeneous in others, notably model risk assessment. As regards CSRBB, the measurement perimeter in most cases includes HTC and HTCS portfolios and is accompanied by a widespread use of sensitivity-based measures, as well as limited integration within RAF frameworks. The integration of CSRBB in the risk management framework can be characterized as being at an intermediate stage, as further methodological developments by banks are expected in the near term, also taking into account the recent clarifications provided by the EBA in the IRRBB heatmap implementation published in January 2026.
This paper proposes C.A.R.E. (Critical Analysis & Review Engine), a framework for the structured epistemic analysis of complex textual constructs, built on a deterministic aggregation engine. The system decomposes each instance into four isolated cognitive archetypes: Procedural (I1), Substantive (I2), Analytical (I3), and Resolutive (I4); each produces a structured judgment comprising a Key Assumptions Check (KAC), fragility classification, an operational prescription, and adversarial steelmanning. These judgments, expressed on three ordered classes (A = Admissible, Rv = Revisable with reservation, E = Excluded, carrying veto effect), are aggregated through a positionally invariant decision engine based on arrangements with repetitions, D(3,4) = 3⁴ = 81 ordered arrangements. These are partitioned into four asymmetric zones under a cascading hierarchy: Hard Veto (45 arrangements, triggered by I2 = E or I3 = E), Soft Veto (20 arrangements, triggered by I1 = E or I4 = E), Fluctuation (15 arrangements, no E present but ≥1 Rv), and Consensus (1 arrangement: four A judgments). The system’s primary output is the CE Rating (Epistemic Class, CE-1 to CE-5), an ordinal scale that classifies the construct under examination into five traceable and auditable levels of epistemic quality, determined deterministically by the zone to which the judgment vector belongs. The combinatorial engine is verified by exhaustive enumeration across all 81 D(3,4) arrangements and across 256 extended combinations incorporating the Rv+ diagnostic class (Qualified Revisable). All heuristic parameters are defined through expert judgment; empirical validation constitutes the priority research agenda. The architecture is domain-agnostic; the default configuration is calibrated for credit risk assessment (EBA/GL/2020/06), adopted here as the applied case study. This work stands in methodological continuity with the author’s earlier publication, which developed the D(n,k) = n^k architecture in the context of multi-model credit rating for Italian SMEs, and transposes it here to multi-archetype epistemic assessment
The objective of the present study is to implement the alternative stochastic binomial trees for the evaluation and estimation of the main sensitivity measures of convertible bonds, thus filling a gap in scientific literature. The paper proposes the implementation of the Haahtela, Jarrow-Rudd and Tian numerical schemes and explores the characteristics, convergence properties and reliability of these evaluation tools. A comprehensive case-study considering the German market, which is an extremely active market in the issuance and trading of these hybrid instruments, is also illustrated.
