Value-at-Risk (VaR), the most widely used measure of market risk, is typically evaluated through backtesting of point forecasts. Such procedures, however, say little about the uncertainty of the estimated quantile. Existing interval methods are each tied to a specific model class and fail when its underlying assumptions are violated. We propose Quantile Dynamically-Tuned Adaptive Conformal Inference (QDtACI), a model-agnostic conformal calibration layer that constructs finite-sample intervals around any VaR forecast, using only the return series and the forecast itself. QDtACI adapts dynamically-tuned adaptive conformal inference to the quantile setting through two components: a pinball-loss nonconformity score aligned with the quantile objective and an asymmetric interval construction, and a multi-speed expert-aggregation mechanism driven by a smoothed violation error and a composite loss on coverage, width, and stability. On synthetic GARCH data, where the true VaR is observable, QDtACI attains near-nominal coverage of the true VaR when the underlying forecast is well-specified, and its coverage degrades in a controlled way as the forecast is misspecified. Against the Delta method, a bootstrap, and the DtACI baseline, it achieves coverage closer to nominal at comparable or better interval quality (Winkler score). Applied to a portfolio of 24 fixed-income assets (2016–2024) with VaR forecasts from CAViaR, DCC-GARCH, and copula models, the intervals are stable in calm periods and widen sharply during stress, including the COVID-19 shock and the 2022–2023 monetary tightening. Because true coverage cannot be measured on real data, we further provide a return-only diagnostic that indicates when interval calibration can be trusted as a proxy for coverage of the true VaR. QDtACI thus offers a single, broadly applicable procedure for uncertainty quantification in VaR, whose reliability tracks the quality of the underlying forecast.
Financial distress prediction (FDP) can help enterprises prevent financial distress in advance and assist investors in reducing investment risks, and it is of great significance for improving the performance of FDP models. The paper constructs a new highly class-imbalanced dynamic Bagging-XGBoost FDP model which combines the two ensemble learning algorithms of Bagging and Boosting, and tests the model based on a dataset from Chinese listed companies during the period 2010–2020. The results indicate that the dynamic Bagging-XGBoost FDP model developed on the balanced training set constructed by random under-sampling achieves a G-value of 93.88
This study examines the impact of corporate digital transformation on stock price crash risk using panel data from Chinese listed firms. We construct a novel digital transformation index based on textual analysis and Word2Vec algorithms. Empirical results indicate that digital transformation exerts a significantly negative effect on crash risk. This effect is more pronounced in high-tech firms and less competitive markets, yet remains insensitive to digital infrastructure. This mitigating effect operates through reducing information asymmetry. Investor sentiment moderates this relationship, with the attenuating effect strengthening under high sentiment. Further analysis reveals that technology-oriented transformation generates a stronger effect than application-oriented transformation. These findings highlight the critical role of digital transformation in enhancing market transparency and financial stability.
This study examines how financial volatility risk affects foreign green energy investments across 38 African countries from 2003 to 2023, and whether digital economy moderates this relationship. The given study utilized fixed effect with Driscoll-Kraay (DK) standard errors, IV-2SLS with DK, and two-step system GMM to investigate these nexus. Drawing on real-option theory, we find that financial volatility risk has a negative influence on foreign green energy investment for African economies. Further, we find that this negative impact is more pronounced in economies with low levels of ESG (economic, social, and governance), institutional quality, economic development, high level of national cultural secrecy, and over-financialization. Leveraging information asymmetry and signaling theory, we find that digital economy development significantly moderates and weakens the negative impact of financial volatility risk on foreign green energy investments. Additionally, we find that physical climate risk has a significant negative effect on dependent variable of the study, and digital economy effectively weakens this relationship.
This paper proposes the concept of investor stickiness and uses empirical data of listed companies in China from 2007 to 2023 to test the impact of investor stickiness on stock liquidity. We find that: (1) Contrary to intuition, investor stickiness significantly reduces stock liquidity. (2) For state-owned enterprises, the negative correlation between investor stickiness and stock liquidity is weakened. These indicate that investor stickiness is an important factor affecting stock liquidity. Further research shows that assimilation mechanism and game mechanism are important mechanisms by which investor stickiness affects stock liquidity, that is, investors’ adaptability reduces stock liquidity. This paper not only enriches the application of complex adaptive system theory in stock market, but also reveals the adverse effects of investors’ adaptive evolution.
This paper examines how the disaggregated Environmental, Social, and Governance (ESG) pillars influence bank financial performance (FP) in the ASEAN-5 region through the mediating role of bank risk-taking (BRT). It aims to clarify whether ESG activities serve as risk-disciplining mechanisms that improve market-based performance outcomes, addressing the empirical ambiguity surrounding ESG-FP linkages in emerging banking systems. The analysis uses dynamic panel data of 62 listed banks across Indonesia, Malaysia, the Philippines, Singapore, and Thailand from 2015 to 2024. To control for endogeneity and unobserved heterogeneity, the study applies the Windmeijer-corrected two-step System Generalized Method of Moments (SYS-GMM) estimator, complemented with Least Squares Dummy Variable Corrected (LSDVC) estimation for bias correction. Mediation is tested using the Baron and Kenny (1986) procedural framework. Results show heterogeneous mediation patterns across ESG pillars. The Environmental pillar exhibits no significant direct or indirect impact on FP, indicating its weak financial materiality in ASEAN banking. The Social pillar provides marginal but consistent evidence of full mediation through BRT, where improved stakeholder relations and community engagement appear to reduce risk-taking and enhance performance indirectly. The Governance pillar exhibits partial mediation, suggesting that board quality and compliance controls influence FP both directly and through risk moderation. These results position BRT as an important behavioural channel linking ESG discipline to financial outcomes. This study is among the first to employ dynamic panel estimators to test the mediating role of BRT in the ESG-FP nexus for ASEAN banks. It contributes to sustainable finance literature by reframing ESG integration as a risk-discipline mechanism rather than a mere compliance tool. The findings offer practical insights for regulators, compliance officers, and bank executives seeking to embed ESG metrics within prudential supervision, credit risk models, and governance frameworks.
This study analyses the impact of exchange rate fluctuations on firm valuations, utilising monthly data from publicly traded Indian firms between 2015 and 2025. We utilise linear, nonlinear, asymmetric, currency-specific, and crisis-sensitive models to analyse sectoral forex exposure, revealing significant variation across sectors, currencies, and crises. Merely 16
Most of the research on institutional dual holdings confirms its positive impact on corporate governance, investment efficiency, and firm innovation in developed countries. In contrast, this paper uncovers the flip side of dual holdings in transitional economies. Using China's context, we find that dual holdings increase stock price crash risk, supporting the transient investor hypothesis. Transmission tests indicate that this effect is more significant when investors are transient; dual holdings increase agency costs of controlling shareholders and corporate upward earning management; dual holders exert more selling pressure when the firm experiences negative returns. Heterogeneous tests show that dual holdings' impact is more obvious for firms with higher sales growth, higher managerial agency costs, or those located in lower market-level regions. We also demonstrate how dual-holding information improves practical risk monitoring and capital allocation decisions.
Opening price gaps provide an observable measure of how information released outside regular trading hours is incorporated into the first tradable price of the next session. Their empirical distributions often display heavy tails, mild asymmetry, and clustered extreme movements, which are difficult to describe using benchmark distributions based on a single regime structure. This study introduces the t-stable power series (TSPS) distribution as a parametric framework for modelling opening gap rates, referred to as opening diffrates, and overnight tail risk. The model combines a Student-t equilibrium component with a stable power series (SPS) shock component, allowing regular price discovery movements to be separated from random sum information shocks. This structure yields an interpretable decomposition of overnight risk through three quantities: the probability of the shock regime, the tail thickness of individual shock impacts, and the latent frequency of material information arrivals. We apply the framework to opening diffrates of major equity indices, including the Shanghai Composite, S P 500, DAX, and Nikkei 225, and compare it with Normal, Laplace, Cauchy, Student-t, and α -Stable benchmarks using likelihood-based information criteria and tail diagnostic tools. The empirical results show that the performance of TSPS varies across markets and sample periods. In samples where opening gaps exhibit shock clustering behaviour, the TSPS model provides additional explanatory power and a more informative decomposition of overnight risk. These findings suggest that the TSPS framework is best understood as a diagnostic distributional tool rather than as a uniformly dominant specification. Its practical relevance lies in connecting numerical tail fitting with interpretable risk components that may support overnight VaR and ES assessment, stress testing, and pre-opening risk monitoring.
Digital retail platforms increasingly combine embedded seller financing with platform-managed fulfilment, making internal fee and logistics policies potential drivers of credit risk. This paper studies a major fulfilment-policy reform at a large e-commerce platform that raised per-order fees, introduced delay-handling charges and repriced returns. Using internal data on more than 1.2 million orders and around 9500 active sellers, and exploiting a staggered difference-in-differences and event-study design, we estimate the policy’s effect on seller default probabilities and expected credit loss (ECL). The policy worsens operational performance for treated sellers: on-time fulfilment declines, return rates increase and refund ratios rise. PD30 increases by about 0.4–0.5
Understanding how environmental risks interact with business activity and financial conditions is increasingly important under heightened global uncertainty. This study examines the role of climate-related fluctuations, proxied by changes in global precipitation, within global macroeconomic and financial dynamics relevant to trade, production, and business risk. Using monthly global data from 2000 to 2025, the analysis integrates wavelet coherence with an asymmetric external decomposition connectedness framework to capture time-varying, frequency-dependent, and regime-specific interactions. The results show that precipitation fluctuation exhibits strong co-movements with financial stress and trade-related indicators, particularly during crisis periods and at medium- to long-term frequencies, while its relationship with real economic activity remains weak and intermittent. Asymmetric connectedness findings further reveal that precipitation consistently acts as a net shock receiver, with slightly stronger asymmetries under negative shocks. Overall, the findings highlight that climate-related risks materialize through financial stress and trade channels that shape business environments and strategic decision-making.
Industry market risk, a critical systemic risk component, requires precise measurement and timely warnings. This study analyzes risk spillovers among ten Chinese industries post-2008, integrating historical patterns and future risk trends. Using elastic net combined with generalized variance decomposition, we quantify static and dynamic risk spillover networks and transmission under exogenous shocks. Advancing existing methods, we develop a CNN-based model that outperforms alternatives in accuracy and stability for early warnings. Key findings: 1) Consumer Discretionary, IT, and Raw Materials are primary risk sources, while Financial and Real Estate absorb most risks; 2) Domestic events amplify cross-industry spillovers; 3) The CNN system effectively predicts aggregate and sector-specific risks. This framework equips regulators to prioritize prevention and enables investors to adjust portfolios dynamically. Our work bridges systemic risk measurement and forecasting, offering actionable strategies for managing interconnected market risks.
This study examines the relationship between income diversification and the financial stability of Vietnamese commercial banks over the period 2015–2024, a decade marked by structural reforms, rapid digital transformation and increasing fintech competition. Income diversification is measured using a Herfindahl–Hirschman Index-based approach, while financial stability is proxied by the Z-score. Using panel data from 26 banks, the study employs fixed effects estimation and feasible generalized least squares to address heteroskedasticity and autocorrelation, complemented by an instrumental variable approach to account for potential endogeneity. The study addresses an important issue in contemporary banking risk management literature: whether expanding non-interest income activities enhances or weakens bank stability in emerging financial systems undergoing structural transformation. The results show that income diversification is positively and significantly associated with financial stability, suggesting that more balanced income structures are associated with lower insolvency risk. In addition, loan-to-total-assets ratio, net interest margin and bank size are positively associated with stability, whereas inflation is negatively associated with financial resilience. By focusing on a reform-driven emerging economy, the study extends recent literature on bank risk management and diversification by showing that the stabilizing role of diversification may vary across institutional transition, operational efficiency and macroeconomic conditions. These findings remain robust after controlling for endogeneity. The findings provide implications for bank managers and policymakers regarding revenue structure strategies and financial system resilience in emerging banking markets.
Geopolitical risk (GPR) has emerged as a critical source of systematic risk in global financial markets, yet its heterogeneous impact across industries in emerging economies remains underexplored. This study investigates how global and country-specific GPR influence industry-level volatility in South Africa using the GARCH-MIDAS framework, which decomposes volatility into short- and long-run components. Results reveal clear asymmetries across industries. In line with theory, financials and industrials exhibit elevated volatility in response to both global and domestic GPR shocks, reflecting their exposure to capital flows, global supply chains and macroeconomic uncertainty. Conversely, defensive industries such as consumer staples remain relatively resilient. Out-of-sample forecasts confirm the predictive relevance of GPR, particularly for the healthcare industry, where GPR-augmented models consistently outperform specifications based on realised volatility. Overall, the findings imply that policymakers should prioritise stabilisation in highly exposed industries, while investors should adopt allocation strategies tailored to industry-specific sensitivities that balance risk reduction with opportunities for strategic positioning during periods of heightened uncertainty.
Global supply chains during the post-COVID19 period faced with unprecedented volatility, characterized by structural regime shifts and greater inter-dependency among energy, shipping, and logistics sectors. Additionally, the traditional univariate forecasting models fail to capture these spatial spillovers across the economic network. Thus, this study proposes a novel spatio-temporal framework for predicting supply chain price dynamics. We first use Economic Network Analysis (ENA) to map the shifting topology of global interdependencies, identifying structural changes among the pre- and post-COVID19 periods. Further, these inputs are integrated into a Graph Attention Network (GAT) coupled with a Long Short-Term Memory (LSTM), producing a walk-forward validation mechanism to handle post-pandemic non-stationarity. Our results reveal a structural break in the post-COVID19 period, with Systemic Risk Index (SRI) for Crude oil increasing by over 341
This study examines the long-term effects of the European Central Bank’s (ECB) unconventional monetary policy (UMP) interventions on the yields of sovereign bonds in the Eurozone. Using a sample of 14 European countries from January 2009 to December 2023, our findings indicate that increases of 1 billion euros in the ECB’s balance sheet are associated with average long-term decreases of 48 basis points in the European sovereign yields; though these effects are larger on yields of the European periphery countries compared to those of the core countries. Our results are robust under a set of various endogenous panel data models including the dynamic Arellano-Bond estimations. This research makes a significant contribution to the understanding of debt markets pricing by elucidating the extent to which interventions in government debt securities foster price distortions in complex debt markets such as the European and the associated risk management implications.
Traditional stochastic dominance approaches are often too restrictive for practical portfolio selection, as strict dominance conditions rarely hold in real-world financial data. This study addresses this limitation by proposing a novel framework that integrates p-values into the Almost Stochastic Dominance (ASD) approach, interpreting them as preference-based tolerance parameters for flexible and risk-sensitive asset selection. Using U.S. energy sector equities classified into Mega Cap, Large Cap, and Mid Cap groups, we evaluate dominance relationships under first-, second-, and third-order stochastic dominance (FSD, SSD, and TSD) across different p-value thresholds. Unlike the strict non-parametric dominance rule, the proposed framework interprets p-value thresholds as tolerance parameters that control the acceptance of dominance under statistical uncertainty. Empirically, the analysis is between 2021–2025 period and further strengthened through regime-based robustness tests and a 30-day rolling-window Sharpe maximization framework. The results indicate that TSD-based portfolios generally achieve stronger risk-adjusted performance and exhibit improved downside-risk characteristics relative to FSD- and SSD-based portfolios. Tail-risk measures such as Value-at-Risk, Expected Shortfall, and Maximum Drawdown indicate that dominance-based screening becomes more informative when combined with dynamic optimization rather than equal-weighted allocation. The findings also suggest that the optimal p-value threshold is regime-dependent: stricter thresholds perform better in more turbulent periods, whereas intermediate thresholds may improve performance in more stable environments. Overall, the ASD–p-value framework provides a flexible and practically implementable decision rule for risk-sensitive portfolio selection.
This case study investigates the collapse of Credit Suisse (CS), a globally systemically important bank (G-SIB), culminating in its state-brokered acquisition by UBS in March 2023. Employing a qualitative case study methodology using public secondary data (regulatory reports, company disclosures, investigation findings, academic literature, and grey literature), the study addresses how and why this failure occurred despite the post-GFC regulatory framework. Guided by an integrated framework (Agency Theory, Organizational Decline, Systemic Risk/Regulatory Effectiveness), the analysis identifies critical internal deficiencies as primary drivers. Findings highlight systemic risk management failures (Archegos, Greensill), persistent corporate governance weaknesses (leadership instability, accountability deficits, misaligned incentives), and strategic inconsistencies eroding resilience, consistent with organizational decline theories. These internal agency problems interacted with external pressures, including market contagion (SVB failure) and a fatal confidence loss, triggering a digital bank run. The study details the emergency Swiss intervention involving unprecedented liquidity support, bypassing standard protocols and controversially writing down AT1 capital while preserving equity. This deviation from the too big to fail (TBTF) playbook challenges resolution framework effectiveness for G-SIBs in crises, underscores addressing deep organizational pathologies beyond metrics, and reveals moral hazard and wealth transfer implications of the state-engineered rescue. The case offers key lessons for bank regulation, risk management, corporate governance, and AT1 bond treatment. Its main contribution is an integrated multi-theory framework that identifies a failure cascade and supervisory paradox overlooked by single-theory analyses.
The ever-changing nature of financial markets underscores the need for early warning mechanisms to prevent and mitigate systemic financial risks. This paper proposes a novel approach to measure and predict the higher-order moment risk spillovers, offering early warning risk detection signals in commodity markets by integrating machine learning and traditional quantitative modeling. We employ a combination of the Autoregressive Conditional Density (ARCD) model, the Time-Varying Parameter Vector Autoregression Extended Joint Connectedness (TVP-VAR-EJC) model, and the improved Graph Convolutional Network (IGCN) model with an edge-deletion method. Our results show significant heterogeneity between volatility and higher-order moment risk spillover. We document that energy and precious metals are the main net risk transmitter and receiver of the moment-based spillovers. The pairwise net spillover between energy and precious metals contributes the most to the total commodity risk spillover prediction. Our results show that the proposed IGCN model outperforms alternative deep learning models such as LSTM, GRU, and Transformer.
The study examines the impact of biodiversity risk on default probability. The study shows that biodiversity risk increases the likelihood of default. The impacts are significant for firms with lower access to credit and cash reserves. Firms with leverage jumps to meet operating needs exhibit significantly higher 5-year-ahead default risk. As firms reserve sufficient liquidity, leverage jumps decrease default risk. Local risk climate factors moderate the impact of biodiversity on default risk. Treated firms that are prone to transitional and physical risk factors experience a significantly increased default probability. There is a bright side for firms with closer attention from management boards and analysts to climate change exposure. Local climate action maintains a firm’s growth prospects through the mitigated effects of biodiversity risk on default probability. Given the increasing awareness of stakeholders, the empirical findings validate hypothesized theoretical predictions on related risk factors and corporate behaviors, along with potential mechanisms that remain for future studies.
