Systemic Risk Research Articles

Climate change risk pricing in the European stock market

ABSTRACT Climate risk is the potential risk for societies or ecosystems from the impacts of climate change. Whether financial markets price climate risks is critical to ensuring that the necessary funding flows into environmentally sound projects and that stranded assets risk is adequately managed along a ‘green transition path’. In this paper, we assess climate risks for the European stock market. We show that measures of returns spread of green vs. brown investment reflect climate risk and assets’ exposition to systematic macro-financial risk factors, which should be filtered out to measure climate risks accurately. We show that climate risks are priced by focusing on aggregate, industry, and company-level data. We propose a market-based green rating procedure to evaluate non-transparent and non-disclosing companies for which ESG information is unavailable. We illustrate its implementation using a sample of over 800 non-transparent firms.

IAmyP: A Multi-view Learning for Amyloidogenic Hexapeptides Identification Based on Sequence Least Squares Programming.

The development of peptide drug is hindered by the risk of amyloidogenic aggregation; if peptides tend to aggregate in this manner, they may be unsuitable for drug design. Computational methods aimed at predicting amyloidogenic sequences often face challenges in extracting high-quality features, and their predictive performance can be enchanced. To surmount these challenges, iAmyP was introduced as a specialized computational tool designed for predicting amyloidogenic hexapeptides. Utilizing multi-view learning, iAmyP incorporated sequence, structural, and evolutionary features, performing feature selection and feature fusion through recursive feature elimination and attention mechanisms. This amalgamation of features and subsequent feature selection and fusion lead to optimal performance facilitated by an optimization algorithm based on sequence least squares programming. Notably, iAmyP exhibited robust generalization for peptides with lengths of 7-10 amino acids. The role of hydrophobic amino acids in the aggregation process is critical, and a thorough analysis have significantly enhanced our insight into their significance in amyloidogenic hexapeptides. This tool represented an advancement in the development of peptide therapeutics by providing an understanding of amyloidogenic aggregation, establishing itself as a valuable framework for assessing amyloidogenic sequences. The data and code can be freely accessed at https://github.com/xialab-ahu/iAmyP .

Optimal Monetary and Fiscal Policies to Maximise Non-Parallel Risk Premia in Sovereign Bond Markets

In this paper, we analysed several emerging market (EM) and developed market (DM) sovereign yield curves to identify the proportion of parallel and non-parallel shifts over time. We found that non-parallel shifts are more prevalent in EM due to higher political and economic risks. Key drivers include systemic risk events like wars, debt distress, and pandemics. By backtesting a long butterfly strategy to extract non-parallel risk premia from June 2007 to March 2024, we observed that steeper slopes and greater curvature result in higher returns. We also quantified monetary and fiscal regimes to determine what types of policies are required to extract non-parallel risk premia from these sovereign yield curves. Our research suggests that countries with opposing monetary and fiscal policies possess higher return opportunities whilst countries with complementing policies require tactical butterfly strategies to optimise returns.

Financial stability at risk: evidence from market overreaction and herding behaviour in developed and emerging markets

PurposeThis study examines the relationships between herding behaviour, market overreaction and financial stability in developed and Brazil, Russia, India and China (BRICS) markets from 1 January 2017 to 31 December 2023. It identifies the significant differences in these phenomena across different market types and their implications for financial stability.Design/methodology/approachThis study employs panel data regression, quantile regression, Granger causality tests and the Baron and Kenny mediation model to analyse the data. These methods are used to explore the extent to which herding behaviour exacerbates market overreaction and affects financial stability.FindingsThe results reveal that herding behaviour exacerbates short-term market overreaction, leading to increased financial instability, particularly in BRICS markets. In contrast, herding behaviour does not significantly impact intermediate-term overreactions in developed markets. The study also finds that market overreaction significantly mediates the relationship between herding behaviour and financial stability.Practical implicationsThese findings have practical implications for policymakers. Understanding how herding behaviour and market overreaction impact financial stability can help formulate strategies to enhance market stability and mitigate systemic risks, particularly in more volatile BRICS markets.Social implicationsEnhanced financial stability has broad social implications, including improved investor confidence and economic growth. Policymakers can use these insights to create more stable financial environments, which can lead to more robust economic development and reduced vulnerability to financial crises.Originality/valueThis study provides new insights into the differential impact of herding behaviour and market overreaction on financial stability in developed and BRICS markets. By confirming the mediating role of market overreaction, this study enhances our understanding of financial market anomalies and contributes to the literature on financial stability.

The Fairness of Credit Scoring Models

In credit markets, screening algorithms aim to discriminate between good-type and bad-type borrowers. However, when doing so, they can also discriminate between individuals sharing a protected attribute (e.g., gender, age, racial origin) and the rest of the population. This can be unintentional and originate from the training data set or from the model itself. We show how to formally test the algorithmic fairness of scoring models and how to identify the variables responsible for any lack of fairness. We then use these variables to optimize the fairness-performance tradeoff. Our framework provides guidance on how algorithmic fairness can be monitored by lenders, controlled by their regulators, improved for the benefit of protected groups, while still maintaining a high level of forecasting accuracy. This paper was accepted by Will Cong, finance. Funding: This work was supported by the Autorité de Contrôle Prudentiel et de Résolution (ACPR) Chair in Regulation and Systemic Risk, the Fintech Chair at Dauphine-PSL University, and the French National Research Agency (ANR) [MLEforRisk ANR-21-CE26-0007, Ecodec ANR-11-LABX-0047, and F-STAR ANR-17-CE26-0007-01]. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.03888 .

The Impact of Regulatory Stress Tests on Banks' Portfolio Similarity and Implications for Systemic Risk

The Impact of Regulatory Stress Tests on Banks' Portfolio Similarity and Implications for Systemic Risk

Quantifying Uncertainty Costs in Renewable Energy Systems Considering Probability Function Behavior and CVaR at Low-Probability Generation Extremes using Deterministic Equations

The increase and integration of renewable energy sources in electrical power systems implies an increase in uncertainty variables, both in costs and production, of economic dispatch (ED) and currently have a significant influence on wholesale electricity markets (MEM). Uncertainty costs refer to the quantification of additional expenses or economic losses associated with the variability inherent in the generation of renewable energy, such as wind, solar, or hydroelectric. Therefore, this article presents deterministic equations related to cost overestimation and underestimation, as well as CVaR, to model and evaluate the stochasticity of risks associated with the integration of renewable sources, allowing system operators and planners to make informed decisions. To mitigate or use said risks in energy systems with high penetration of elements, mainly smart networks. In this study, a mathematical analysis is carried out using the histogram spectrum formed by the power generated by the probability density function (PDF) for solar generation, although it is possible to consider other types of functions to determine energy generation. The objective of the proposed model is to provide another tool to the system operator for energy management and planning, which relieves a little of the weight of the computational load and at the same time presents more precision in the results by being able to work with a database. Historical data if these values are available. Commonly, for this type of analysis, values are estimated using probabilistic calculations by density functions when integrating these functions, or in other recent cases by estimating them by analytical methods of the same functions. A validation of the model is presented by comparing the result with the Monte Carlo simulation, developing the total cost of uncertainty only from "low probability generation extremes". Furthermore, the results are presented through analytical uncertainty cost functions (AUCF). This analysis includes the calculation of uncertainty costs for low and high-probability energy generation, determined by the Conditional Value at Risk (CVaR), using deterministic equations.

AI Tools in the Digital Transformation Programmes of Industrial Enterprises

This article is devoted to the study of the prospects for the development and implementation of AI in industry in the context of digital transformation. Autonomous factories, integrated supply chains, and autonomous vehicles is a real proof of technological advances that AI brings to life. The article highlights a number of problems and challenges for Russian industry based on the results of applied research conducted in November-December 2023 at 18 medium and large industrial enterprises. The main conclusions include not only the identified systemic problems and risks (shortcomings in digitalisation methodology, ambiguity in calculating economic parameters, lack of technical expertise), but also the most promising areas for the development of AI technologies. The study finds significant potential for enhancing the technologies of the Industry 3.0 and Industry 4.0 paradigms using AI tools. Also, it describe the necessary changes in enterprise management and regulatory government activities aimed at realising the identified potential. In conclusion, the article emphasises the need to optimise parallel imports, develop human capital and adequately analyse the economic parameters of industrial enterprises’ digital transformation project. The authors emphasise the relevance of studying the prospects for the development of AI in industry for Russian economic science in the period 2024–2026.

UNDER THE LOAN: AN ANALYSIS OF ARMENIAN BANKING SECTOR AND CREDIT TRENDS

This study examines recent trends in Armenia's banking sector and export patterns, revealing underlying economic vulnerabilities despite apparent growth. Using data from the Central Bank of Armenia and official trade statistics, this paper analyzes shifts in lending patterns across various economic sectors and changes in export composition. Our findings indicate a concerning decline in industrial sector lending, coupled with a significant increase in mortgage and consumer loans. Additionally, we observe a heavy reliance on precious metal re-exports for economic growth, raising questions about the sustainability of Armenia's current economic trajectory. This research contributes to the understanding of potential risks in Armenia's financial system and highlights the need for diversification in its export base.

Performance Analysis and Novel Cross-Flow Scheme of Low-Temperature Multi-Effect Distillation for Treating High-Mineralized Mine Water

Low-temperature multi-effect distillation (LT-MED) can be used to desalinate and demineralize highly mineralized mine water, facilitating the recycling and reuse of water resources. This study conducted a simulation of LT-MED technology for treating highly mineralized mine water using Aspen Plus and proposed a novel cross-flow optimization scheme. Initially, the impact of operational parameters such as process configuration, number of evaporation effects, and steam input on the gained output ratio (GOR) and scaling risk of a conventional LT-MED system was analyzed. It was found that the number of effects and heating steam flow rate had the most significant influence on GOR, while different processes exhibited limitations regarding GOR and scaling trends. To address these issues, this paper introduced a cross-flow operation process that combined forward, backward, and parallel flow. The simulation results indicated that, under conditions of eight effects, a maximum evaporation temperature of 70 °C, a temperature difference between adjacent effects of 4 °C, and a feed temperature of 45 °C, the cross-flow process—where the feed was introduced from the sixth effect—achieved the highest GOR and significantly reduced scaling risks compared to parallel and backward flow configurations. Finally, to further utilize low-pressure exhaust steam from the final effect of the cross-flow LT-MED system, mechanical vapor compression (MVC) and thermal vapor compression (TVC) were integrated into the LT-MED process. The thermodynamic performance of the coupled system was analyzed, and the simulations demonstrated that the coupled system outperformed the standalone use of either TVC or MVC, with the LT-MED-MVC-TVC system showing superior performance overall.

Do sociodemographic and health predictors affect the non-insulin-based insulin resistance index? A cross-sectional study.

Insulin resistance (IR) is considered a risk factor for cardiovascular diseases (CVD). Therefore, early diagnosis of IR is clinically significant for primary and secondary CVD prevention initiatives. In addition, non-insulin metabolic indices may be useful for diagnosing IR. The first objective was to estimate the triglyceride and glucose (TyG) index and the metabolic score for insulin resistance (METS-IR) index values in a local community with high social deprivation and increased cardiovascular risk according to the Systematic Coronary Risk Evaluation scale. The second objective was to identify significant sociodemographic and health predictors for the TyG index and METS-IR index. This cross-sectional study was conducted in the local community of Janów district in eastern Poland and consisted of 2 stages. The 1st stage involved basic research (n = 4,040), while the 2nd stage involved enhanced diagnostics (n = 2,657). Data from the 2nd stage was used for the analyses. Anthropometric and physiological measurements were taken, blood was drawn for laboratory tests, selected sociodemographic and health variables were evaluated, and the TyG index and METS-IR index were calculated. The mean TyG index score in the study group was 8.65 (±0.58), and the mean METS-IR index score was 41.45 (±9.02). Both indices were significantly associated with age, male sex, smoking, and systolic blood pressure (SBP) in a multivariable model. In addition, alcohol consumption and body mass index (BMI) were significantly correlated with the TyG index, whereas education was significantly associated with the METS-IR index. Our results show the association between IR and sociodemographic and health variables in a group with a high social deprivation rate and increased cardiovascular risk. Early detection of cardiometabolic risk is important for both primary and secondary CVD prevention. In primary healthcare, this can be accomplished through surrogate markers of IR.

Abstract 4138519: A Scoping Review Exploring Cardiovascular Risk and Health Metrics and Cancer Prediction

Introduction: Cardiovascular disease (CVD) and cancer are leading causes of morbidity and mortality. Since they share risk factors, it is plausible that CVD risk scores and cardiovascular health (CVH) metrics could also be used to predict cancer. Research Question: What is the relationship between established CVD risk scores and CVH metrics and the incidence of cancer? Methods: A librarian searched four databases (Embase, PubMed, Scopus, Web of Science: Core) in February 2024 without language restriction. Using Covidence, two reviewers independently screened the titles, abstracts, and full texts of the retrieved articles based on the inclusion criteria (prospective observational studies, adults aged ≥18 years, reported incidence of any type of cancer as an outcome, and used at least one CVD risk score or CVH metric as an exposure). A third reviewer resolved conflicts. The characteristics of the studies retained after screening and the summary estimates of the associations were extracted using Excel. Results: Of the 4,165 records screened, 13 studies (reporting 14 CVD or CVH metrics) were included. The heterogeneity in the scales (comparison group) precluded a meta-analysis. Four studies evaluated CVD risk scores (Atherosclerotic Cardiovascular Disease [ASCVD], Framingham Risk Score, Systematic Coronary Risk Evaluation), and 10 reported CVH metrics (American Heart Association [AHA] Life’s Simple 7, AHA Life’s Essential 8). Sample size ranged from 1,880 to 342,226, with median follow-up from 8.1 to 29.6 years. Most of the studies included more women than men (61.5%) and mean age ranged from 44.8 to 72 years. The majority of studies included all cancer types (71.4%), including breast, lung, colorectal, and prostate cancer subtypes, and the total number of cancer events ranged from 387 to 11,643. Studies utilizing CVD risk scores consistently reported that individuals with a higher CVD risk, developed any type of cancer, with hazard ratios (HRs) ranging from 1.02 to 3.71, and studies employing CVH metrics generally indicated an ideal CVH is associated with a lower cancer risk, with HRs ranging from 0.49 to 0.95. Conclusion: Despite heterogeneity in CVD risk metrics and cancer subtypes, most studies reported that higher CVD risk scores or worse CVH metrics may increase future cancer risk. More rigorous studies are needed focusing on specific CVD risk metrics and cancer types of cancer to produce evidence suitable for a meta-analysis.

MICROPRUDENTIAL ASPECT BEYOND RULE OF THUMB IN RURAL BANK INDUSTRY IN INDONESIA

Objective: This research aims to examine the microprudential aspects beyond traditional rule of thumb measures in the regulation of Rural Banks in Indonesia, particularly focusing on the implications of recent regulatory changes that have prompted Rural Bank mergers. Theoretical Framework: This research utilizes a framework that incorporates financial ratios as critical indicators of bank performance and health. It considers the influence of regulatory frameworks—especially those outlined in POJK regulations—on the operational effectiveness and stability of Rural Banks. Related to the forced corporate action, reason that administrative reform may not have increased merger activity as expected is that larger, combined activities are likely to be more readily able to absorb the significant costs of complying with administrative reform than smaller Rural Bank. Method: This research methodology used is a qualitative descriptive method. In conducting this study, researchers used primary data and secondary data. Primary data obtained from the results practitioners' perspective of interviews (in-depth interviews) with experts and practitioners, who have an understanding of the issues discussed in Rural Bank Industry so that we can presumably determine the real current situation using a Qualitative Descriptive method with a Phenomenology study approach. While secondary data was complemented by data collected from literature studies, from the Financial Service Authority (OJK) official website, Several Rural Bank Performance Report and Financial Report during 2019 - 2023, scientific articles, books, information from the mass media and other relevant sources of information. Result and Discussion: This study highlights that reliance solely on the rule of thumb can be misleading in assessing bank health. It shows that a detailed analysis of various financial ratios—such as capital adequacy, asset quality, earnings, and liquidity—provides a more comprehensive understanding of a bank's performance and risks associated with mergers. The banking industry those had time to familiarize itself with the subject and to put legislation into concrete action. Other findings indicate that the enforcement of regulations aimed at increasing capital adequacy and reducing systemic risk has led to forced mergers among Rural Banks. Research Implication: The study underscores the importance of integrating microprudential regulations into the operational frameworks of Rural Banks. It suggests that regulators should prioritize comprehensive assessments based on financial ratios rather than simplistic heuristic methods, especially in the context of forced mergers driven by regulatory pressures. Effective and appropriate use of the merger strategies forced by regulator (OJK) have not facilitated Rural Banks’ efforts exactly to earn above-average returns. However, even when pursued for value-creating reason, merger strategies are not problem-free. Reason for the use of merger strategies and potential problem with such strategies are suboptimum goals congruence because of management control problem – include lack of direction, motivational problem, and personal limitation (Merchant et al, 2017). Recommendation: The research recommends that Rural Banks systematically evaluate their financial health using detailed financial ratios and adapt to the new regulatory landscape that encourages mergers. It also urges regulators to refine their guidelines to include these metrics as essential tools for ongoing performance assessment. Enhanced training for bank management on interpreting and applying these financial ratios can improve overall operational effectiveness, particularly in the context of mergers. Merger is not just about consolidating financial aspects, but also harmonizing the combination of tangible (Numbers on Balance Sheet) and intangible assets (Human capital aspect).

Exploring the Impact of Trade Credit and Debt on Systemic Risk in the Presence of Financial Distress and Constraints

Based on the extreme value and network contagion theories, this study intends to explore the effects of trade credit, and short/long-term debt on the systemic risk of 205 Pakistani non-financial firms over the period from 2005 to 2021. To evaluate the firms’ systemic contribution and vulnerability, we apply two firm-specific measures, the change in conditional value-at-risk (DCOVAR) and marginal expected shortfall (MES). Our findings reveal a significantly positive impact of trade credit, and short/long-term debt on the systemic contribution and vulnerability of these firms. The findings also indicate that the underlying relationship is significantly moderated by financial distress and financial constraints. Understanding how different forms of debt influence systemic risk can guide policymakers in crafting regulations or interventions to promote a healthier financial environment. Regulators may reevaluate the balance between short/long-term debt usages or develop guidelines for managing trade credit effectively to mitigate systemic risk.

The EU's Artificial Intelligence Act and copyright

AbstractThe European Union's (EU's) Artificial Intelligence Act (AI Act), published on 12 July 2024, seeks to establish a consistent legal framework for AI systems within the EU, promoting trustworthy and human‐centric AI while safeguarding various fundamental rights. The Act classifies AI applications into three risk categories: unacceptable risk, high risk, general purpose AI models with systemic risk and low or no risk, each with corresponding regulatory measures. Although initially not focused on copyright issues, the rise of generative AI led to specific provisions addressing general purpose AI models. These provisions include transparency obligations, particularly regarding the technical documentation and content used for training AI models, and policies to respect EU copyright laws. The Act aims to balance the interests of copyright holders and AI developers, ensuring compliance while fostering innovation and protecting rights.

Mitigating Risk and Emissions in Power Systems: A Two-Stage Robust Dispatch Model with Carbon Trading

The large-scale integration of renewable energy sources is crucial for reducing carbon emissions. Integrating carbon trading mechanisms into electricity markets can further maximize this potential. However, the inherent uncertainty in renewable power generation poses significant challenges to effective decarbonization, renewable energy accommodation, and the security and cost efficiency of power system operations. In response to these challenges, this paper proposes a two-stage robust power dispatch model that incorporates carbon trading. This model is designed to minimize system operating costs, risk costs, and carbon trading costs while fully accounting for uncertainties in renewable energy output and the effects of carbon trading mechanisms. This model is solved using the column-and-constraint generation algorithm. Validation of an improved IEEE 39-bus system demonstrates its effectiveness, ensuring that dispatch decisions are both robust and cost-efficient. Compared to traditional dispatch models, the proposed model significantly reduces system risk costs, enhances the utilization of renewable energy, and, through the introduction of a ladder carbon trading mechanism, achieves substantial reductions in carbon emissions during system operation.

Modelling the impacts generated by reclaimed wastewater reuse in agriculture: From literature gaps to an integrated risk assessment in a One Health perspective

The reuse of reclaimed wastewater is increasingly recognized as a viable alternative water source for irrigation. Its application, whether direct or indirect, impacts several interconnected compartments, including groundwater, surface water, soil, crops, and humans. Reclaimed wastewater provides essential resources for crops, like water and nutrients. However, it also introduces pathogens, and contaminants of emerging concern (CECs), defined as chemicals that may pose risks to human health and ecosystems but are not yet fully regulated, such as pharmaceuticals and personal care products, among others. Additionally, reclaimed wastewater may contain antibiotic-resistant bacteria (ARBs) and disinfection by-products (DBPs), all of which present potential health and environmental risks. Therefore, regulatory bodies stress the need for preventive risk assessments to ensure safe reuse.This paper critically reviews available models for assessing the impacts of reclaimed wastewater reuse in agriculture. It identifies gaps in current modelling approaches and outlines future research directions. Key areas requiring further investigation include the fate and transfer of CECs, ARBs and DBPs, and the co-occurrence of multiple risks in such interconnected systems, especially in the indirect reuse. To address these gaps, we proposed a simplified approach to integrate three types of risk associated with CECs in indirect reuse, focusing on risks posed by antibiotics and other pharmaceuticals: human health risk, environmental risk and risk from antibiotic resistance development. This approach aids in identifying the most critical endpoints within the One Health approach, supporting (i) CECs prioritization in regulations based on their critical endpoints and (ii) the adoption of CEC-specific mitigation measures.

Identification and control of risks in the payment system: Russian and foreign experience

The purpose of the study is to analyse the modern mechanism of identification and risk control in the payment system both in the Russian Federation (hereinafter referred to as RF, Russia) and abroad. The objectives of the study are to analyse a standardised approach to identification and risk control in the payment system of Russia and foreign countries, as well as a review of the estimated indicators of the identification and control system in the payment system of the RF. To achieve this goal, general scientific methods (analysis, synthesis, generalisation, etc.) and a number of private scientific methods (systemic, comparative legal, statistical, etc.) are used in this work, necessary for a deeper consideration of individual issues on the research topic. The paper analysed a standardised approach to identification and risk control in the payment system of Russia, the United Kingdom and the United States of America. The review of the estimated indicators of the identification and control system in the Russian payment system is conducted. The theoretical significance of the research lies in expanding knowledge and understanding of modern mechanisms for identification and risk control in the Russian payment system, as well as international experience in identifying and controlling risks in the payment system. The practical significance of the study is to identify the features of a standardised approach to identification and risk control in the payment system in Russia and foreign countries, as well as to identify the estimated indicators of the identification and control system in the Russian payment system.

Market-level tenant default risk on residential REIT operational efficiency

ABSTRACT Selecting which geographical markets a REIT serves is an important managerial decision that is expected to significantly impact firm performance. This paper investigates whether holding properties in areas with greater incidence of evictions impacts REIT operational efficiency. Using a comprehensive database of eviction lawsuits in the United States in combination with a time-series of residential REIT portfolio holdings, we construct REIT-year eviction exposure scores to analyse whether more exposure to areas with high aggregate eviction rates is a risk factor affecting REIT efficiency. We posit that market-level tenant default risk captures a locational risk dimension that should be considered in REIT portfolio management. Our results show that REITs with more exposure to markets with high tenant default risk observe higher operational inefficiencies after controlling for relevant firm characteristics and local market factors. These results are robust to multiple model specifications that control for potential endogeneity and omitted variable biases. Our findings have significant implications for REIT managers and investors, suggesting that the choice of property locations should consider aggregate tenant default risk as a determinant factor in residential REIT performance.

Personal Loan Default Prediction Analysis based on TabNet and Logistic Regression

Abstract. With the continuous development of the global economy, the demand for loans from individuals and enterprises is growing. However, loan defaults have gradually become a major challenge facing the financial industry. Loan defaults not only directly affect the profitability of financial institutions, but may also trigger systemic risks and pose a potential threat to the entire economic system. Therefore, improving the accuracy of loan default prediction is crucial for financial institutions to effectively manage credit risks. This study built a system for personal loan default prediction by combining TabNet with the Logistic regression model. Through feature engineering, this study extracts potential credit risk features by utilizing datasets from personal and Internet loans. The accuracy of default prediction is improved by combining TabNet's feature learning capabilities with Logistic Regression's interpretability. An AUC value of 0.89 was achieved by the integrated model, which achieved a notable performance. The results indicate that a system that relies on machine learning to predict defaults can significantly enhance the quality of credit approval decisions and lower the likelihood of bad debts for financial institutions. Future studies could aim to optimize the feature selection process, experiment with more advanced machine learning algorithms, or apply the model to diverse loan datasets, thereby enhancing both its generalization and accuracy. In conclusion, this study offers a novel approach to loan default prediction, demonstrating significant practical value and providing substantial support for the risk management and decision-making processes of financial institutions.