Business, Economics and Management

Abstract: This study examines risk interconnectedness and shock transmission across eight European stock markets over the period 2005–2025. It focuses on how market relationships change during financial and geopolitical crises and in response to European Central Bank (ECB) monetary policy. It also investigates whether dynamic spillover networks can improve forecasts of market connectedness. The empirical analysis combines econometric methods with a network-based deep learning approach. GARCH-family, VAR, and Markov-switching models are used to examine market volatility and shock transmission, while the Diebold–Yilmaz framework measures total and directional connectedness. An event study evaluates market reactions to ECB interest-rate decisions. The analysis covers major stress periods, including the Global Financial Crisis, the European sovereign debt crisis, Brexit, COVID-19, the Russia–Ukraine war, and the European energy crisis. The forecasting analysis uses a Graph Attention Network combined with a Gated Recurrent Unit (GAT–GRU) to predict future changes in the Total Connectedness Index (ΔTCI). The model combines dynamic spillover networks with market, ECB, and crisis information. Its performance is compared with simpler forecasting methods at 5-day and 22-day horizons, using historical volatility estimates and repeated tests with five random seeds. The results show a high level of financial connectedness across European stock markets, with a TCI of approximately 72.7%, indicating that a large share of market risk is linked to shocks originating in other markets. CAC 40 and DAX are the main net transmitters of shocks, while WIG20 and SOFIX are the main net receivers. Risk transmission is also stronger from euro-area markets to non-euro-area markets than in the opposite direction, while overall connectedness increases during major financial crises. Markets differ in their roles as shock transmitters and receivers, while their responses to ECB monetary policy are not uniform. Separate four-market models show much stronger internal connectedness in the euro-area group (static TCI: 69.69%) than in the non-euro-area group (41.35%), while the eight-market GFEVD results indicate stronger average cross-group spillovers from the euro area to the other markets. At both forecast horizons, GAT–GRU shows only small RMSE differences relative to Persistence, without consistent statistically significant improvements.

Abstract: Artificial intelligence in financial services raises questions about whether willingness to engage reflects confidence in technology, expectations of consumer protection or confidence attributed to institutional safeguards. These positions are frequently discussed together despite their different meanings. This study examines their distribution and association in a Bangladesh sample with substantial legal education and professional representation. A cross-sectional questionnaire database containing 385 consenting, AI-aware adults was analysed at item level. Five protection-related statements were assessed using endorsement distributions, paired contrasts, joint response patterns and exploratory adjusted regression. Low internal consistency discouraged combining the statements into a reflective trust scale. Clear-protection conditional willingness was endorsed by 87.8%, compared with 67.5% for oversight-related trust, 83.6% for disclosure, 83.6% for redress and 84.2% for accountability-related confidence. The oversight statement was endorsed 20.3 percentage points less often than conditional willingness (bootstrap 95% interval −26.0 to −14.8; Holm-adjusted p < 0.001). Only 53.0% endorsed conditional willingness, oversight-related trust and accountability-related confidence together. None of the four focal adjusted associations with willingness met the Holm-corrected threshold; ordinal and law-student-restricted analyses likewise retained none. The findings distinguish broad approval of protective conditions from a supported individual-level model of willingness. Institution-based trust and procedural justice provide interpretive lenses, while the measures do not establish current provider trust, actual financial-service use or regulatory effectiveness. Practical implications concern testing understandable disclosures and accessible review arrangements against operational evidence rather than treating favourable survey responses as evidence of implementation.

Abstract: This study re-examines the effect of digital finance on corporate ESG performance, arguing that the modest and specification-sensitive estimates in prior research stem from a pervasive omission of temporal dynamics. Using a panel of Chinese A-share listed firms and a distributed-lag framework, we show that the ESG payoff of digital finance is intrinsically dynamic: contemporaneous associations are statistically indistinguishable from zero, whereas lagged effects of usage depth are robustly positive, so static specifications can therefore understate the delayed cumulative association. The positive lagged associations persist in several alternative specifications. Transmission is quality-selective: digital finance stimulates green invention patents more strongly than utility-model patents, with formal cross-channel tests confirming that the invention-based pathway is significantly stronger. External audit assurance provides additional evidence that the translation of digital-finance capabilities into ESG outcomes depends partly on firms’ information and governance environment. The effect is also structurally uneven: environmental and governance pillars improve while the social pillar declines, and formal cross-pillar tests confirm a significant E/G-versus-S divide, with significant differences in the digital-finance–ESG association by state ownership and firm size. The findings reframe digital finance as a slow, selective, and structurally uneven driver of sustainable corporate transformation.

Abstract: We examine the robustness of the Jump Diffusion Model relative to the Geometric Brownian Motion in forecasting the Japanese Yen carry trade daily returns for the BRIC countries and three bellwether currency pairs GBP/JPY, EUR/JPY, and PLN/JPY during the January 2, 2012 - July 31, 2026, sample period. We include the interest rate return, specifically, the interest earned from the (long position) component of the FX Carry trade in arriving at the daily cumulative carry trade returns. Focusing on the uncovered interest rate parity (UIRP) condition, we find that Jump Diffusion outperforms Geometric Brownian Motion JPY carry trade. However, the Geometric Brownian Motion fits the daily time series data better on some FX Carry trading strategies. The anticipated accuracy of both models is better under quantitative easing relative to quantitative tightening monetary policies.

Abstract: The perceived accessibility and usefulness of financial technology (fintech) do not always translate into greater financial inclusion. This study aims to investigate the interplay between fintech perceptions and financial inclusion through the mediating roles of digital and financial literacy. Data were gathered from 202 female microentrepreneurs in Yogyakarta, Indonesia. Results from Partial Least Squares Structural Equation Modelling (PLS-SEM) suggest that both literacies considerably mediate the link between perceptions of fintech and financial inclusion. Direct tests indicate that fintech perceptions significantly contribute to enhanced digital and financial literacy. Both forms of literacy also are positively associated with financial inclusion. Conversely, the direct connection between fintech perception and financial inclusion was insignificant. These emphasize the vital role of individual capabilities in explaining the interaction between fintech perceptions and financial inclusion. This study expands the application of the capability approach in the fintech and financial inclusion literature by positioning digital and financial literacy as two distinct personal conversion factors. It also implies that the availability of fundamental access to financial services does not necessarily result in identical stages of the inclusion, especially when users' digital and financial capabilities differ.

Abstract: Banks and financial technology providers often train churn models on several historical snapshots of their customer base, a practice known as multi-slice training. Evidence for its benefits rests on few datasets, and widely used random row-wise validation can place copies of one customer record in training and test data. We test the practice under an analysis plan written before any model was trained on the bank data, using public transaction histories of 98,721 corporate clients of Sber (2021–2022) and 50,000 retail clients of T-Bank (2019–2020), with churn defined as three months without transactions. Models trained on the latest slice are compared with models trained on all slices, using out-of-time and customer-disjoint test sets, DeLong tests and the expected maximum profit criterion. Multi-slice training changed the area under the receiver operating characteristic curve by +0.0005 for corporate and +0.0011 for retail clients, within the pre-specified equivalence margin of ±0.005, and raised expected campaign profit by 1.2% for corporate clients only. At equal sample size, older slices performed worse than the latest slice. Random row-wise validation inflated performance markedly only when slices duplicated customer snapshots. In Uzbekistan, Central Bank and Global Findex data show registrations consistent with multi-banking and falling concentration.

Abstract: The low-carbon energy transition, stringent EU climate policies, and recurrent energy supply crises shape industrial operating costs and resilience in coal-dependent Central and Eastern European (CEE) economies. This study investigates the asymmetric and sectorally heterogeneous transmission of fossil fuel commodity and EU Emissions Trading System (EU ETS) CO₂ allowance prices to twelve companies listed on the Polish, Czech, and Hungarian stock exchanges over 2018–2026. To examine differences in exposure to energy and carbon price changes, we compare power generation and oil-refining companies with firms in food processing, telecommunications, and finan-cial services. Daily Brent crude oil, natural gas, Rotterdam coal, and CO₂ allowance prices are modeled using Nonlinear Autoregressive Distributed Lag (NARDL) specifi-cations and Wald symmetry tests, controlling for domestic benchmarks (WIG, PX, BUX). Results reveal distinct differences across the analyzed companies, consistent with their energy exposure profiles: oil-refining companies exhibit pronounced rela-tionships with crude oil prices, whereas the analyzed power generation company shows associations with long-run CO₂ allowance prices and short-run coal price dy-namics. While domestic benchmarks govern immediate fluctuations, nonlinear re-sponses to energy and carbon shocks manifest predominantly in the short-run adjust-ment lag structure rather than cumulative magnitude. Finally, residual outlier analysis and Friedman tests indicate significant differences in abnormal post-shock adjustment durations across shock types, but not across firms, sectors, or countries.

Abstract: Artificial intelligence is increasingly being incorporated into financial-sector risk management, including automated anomaly detection and fraud monitoring. Yet the legitimacy of AI-enabled fraud controls depends not only on technical security gains but also on whether users perceive monitoring as proportionate, privacy-protective, reviewable, and institutionally accountable. This study develops a security-privacy-continuance perspective for mobile financial services (MFS) in Bangladesh and examines how AI literacy, perceived regulatory protection, and procedural fairness relate to users’ acceptance of AI-enabled fraud controls and continued-use intentions. The study follows the Group 8 research design and uses the Q8 five-point Likert questionnaire. The supplied database contains 385 records; however, its Codebook identifies the respondent sequence and contextual comments as synthetic, so the empirical results are explicitly treated as demonstrative rather than field evidence. Composite reliability diagnostics reveal a major measurement problem: Cronbach’s alpha is low for AI literacy (α=.322), fairness (α=.193), regulatory protection (α=.250), and the full fraud/privacy/continuance block (α=.224). This finding is substantively important rather than merely technical: the instrument’s items appear to combine related but non-identical governance dimensions, making a single latent construct inappropriate without refinement. Item-level robust regressions nevertheless show that regulatory protection is positively associated with both continued-use indicators (FPC4: b=.335, p=.002; FPC5: b=.255, p=.009), while AI literacy is positively associated with perceived fraud-detection benefit (FPC1: b=.242, p=.019) and false-alert reviewability (FPC3: b=.267, p=.015). In the aggregate model, AI literacy, fairness, and regulatory protection jointly explain 15.8% of variance in the FPC composite. The study’s principal contribution is therefore a measurement-aware account of AI financial trust: users may support fraud detection while simultaneously requiring privacy safeguards, review mechanisms, and credible regulatory protection. For Bangladesh, the implication is that AI fraud governance should be designed as a socio-technical consumer-protection architecture rather than as surveillance technology alone. The manuscript identifies precise questionnaire revisions and a validation pathway required before substantive publication.

Abstract: Why did bad loans pile up so unevenly across Europe after the 2008 financial crisis? Using data for 28 European countries from 2008 to 2016, this paper finds that rising unemployment was the strongest driver of loan defaults: each additional percentage point of joblessness is associated with roughly 1.5 additional percentage points of non-performing loans. Faster economic growth eases the problem, although more modestly. Bank capital ratios show no reliable protective effect. Countries with denser bank branch networks tend to have fewer defaults, though this evidence is only suggestive. Past loan problems also appear to linger, but this persistence is imprecisely estimated. Taken together, the results suggest that protecting employment and speeding up the resolution of bad debts deserve a central place in banking-stability policy.

Abstract: Economic uncertainty shapes how analysts and managers forecast corporate performance, yet comparative evidence across crises is limited. This study compares the earnings and stock price forecasts of financial analysts and company managers during the 2008 financial crisis and the COVID-19 pandemic. Using annual data for five U.S. firms (Goldman Sachs, AIG, Ford, Wells Fargo and Starbucks) over 2009–2012 and 2020–2022, actual earnings and year-end stock prices are related to analyst and management forecasts through panel regressions with fixed and random effects, supported by descriptive, correlation, normality, robustness and Hausman tests. In the 2008 earnings model, neither forecast is a significant predictor of actual earnings, whereas in the stock price models and the COVID-19 earnings model analyst forecasts are significantly and positively related to actual outcomes and management forecasts are insignificant or negatively related. Both groups tended to underestimate the post-crisis recovery. The findings suggest that analysts and managers should adapt forecasts more quickly to changing economic conditions; recommendations on risk analysis, communication, digital transformation and diversification are derived.

Abstract: Investors value sustainability information that is financially material to a firm’s industry, but little is known about how much of it firms outside the United States and Europe report. We examine 7,278 statutory business reports of 808 non-financial KOSPI firms for fiscal years 2016–2025. A classifier trained on large language model annotations sorts every narrative sentence by environmental, social and governance (ESG) issue and specificity, and the industry standards of the Sustainability Accounting Standards Board separate material from immaterial content. The classifier outperforms dictionary and zero-shot methods; blind human coding confirms this advantage but shows that the classifier understates material content. Material ESG information is a small, slowly growing, mostly qualitative part of the reports. After Korea’s 2021 roadmap for mandatory sustainability disclosure, the largest firms expanded their ESG content more than other firms. They added specific material information but also generic and, less robustly, immaterial statements, so their disclosure did not shift significantly towards material issues; the estimates are sensitive to flexible trend controls. Material disclosure is not robustly related to firm value, and its weak association with better liquidity is shared by ESG content generally. Anticipated mandates can expand disclosure without clearly making it more useful to investors.

Abstract: After 17 years of conventional privatization, Egypt is changing course. Recently, the government proposed a new privatization scheme called the Egyptian Citizenship Ownership Program (ECOP), aimed at speeding up the privatization program, widening equity ownership, and improving firm performance. The proposed program is different¾in design, proposed implementation, institutional settings, and macroeconomic environment¾-than those implemented in transition economies during the early and mid-1990s. We argue that the ECOP is designed to solve a specific problem, the need to rapidly privatize a large number of firms in circumstances where only a few potential buyers had sufficient funds to purchase. The ECOP also provides a means of transferring these assets to the population at large, and hence it is politically acceptable and renders a practical solution to accelerating privatization of the entire enterprise sector. By giving firms to the public for free or at a nominal fee, the process is designed to give individual citizens a greater stake in the new economic system by making them owners of the assets of the country, so it also addresses the issue of economic justice.

Abstract: The banking sector serves as a critical transmission belt for capital allocation toward renewable energy (RE) projects, yet traditional equity valuation models fail to capture banks' exposure to energy transition risks. To bridge this gap, this paper develops a novel fundamental-environmental indicator—the "Green F-Score"—tested on 10 major commercial banks listed on the Warsaw Stock Exchange (WSE) between 2023 and 2025. By synthesizing classical financial efficiency criteria with EU Taxonomy disclosures (Green Asset Ratio), the portfolio sorting methodology reveals how capital markets price sustainable energy financing in Central and Eastern Europe. The empirical results demonstrate that portfolios with high environmental-fundamental readiness ("Green Leaders") generate a superior mean annual return of 60.43% and a substantial price-to-book (P/B) premium. Conversely, banks lagging in RE financing suffer significant carbon risk discounts. Correlation analysis confirms a paradigm shift: the market premium is primarily driven by the ESG sub-score, rendering traditional financial metrics a mere operational hygiene factor. Consequently, the Green F-Score emerges as a robust tool for institutional investors, proving that accelerated lending to the energy transition effectively lowers the cost of equity capital for commercial banks.

Abstract: Earnings management (EM) remains a significant concern in corporate financial re-porting, particularly among carbon-intensive companies facing increasing pressure to demonstrate environmental responsibility (ER) while maintaining financial reporting credibility. This study investigates the moderating role of board gender diversity (BGD) and board remuneration (BR) on carbon emission reduction (CER)- earnings manage-ment (EM) relationship among Johannesburg Stock Exchange (JSE)-listed car-bon-intensive companies. This study adopts a quantitative panel data approach, exam-ining 106 companies over the period 2015–2025. Fixed-effects model is employed to es-timate the main relationships, while two-step system generalised method of moments estimation for additional robustness testing. The findings reveal a statistically significant negative relationship between CER and EM, suggesting that stronger carbon emission reduction is associated with reduced earnings management. Furthermore, the findings indicate that BGD and BR shape the CER–EM relationship, highlighting the importance of corporate governance mechanisms in financial reporting behaviour. The study con-tributes to the sustainability reporting literature by showing the relevance of board composition and remuneration arrangements in understanding the environmental re-sponsibility-financial reporting quality nexus.

Abstract: This study examines the direct effects of fiscal condition indicators (fiscal balance and external debt) and political stability on sovereign credit ratings, and the moderating role of political stability, using panel data from 71 countries covering the period from 2002 to 2023, which was further decomposed based on economic development levels. I used the dynamic BCMM to estimate the baseline results and applied the DKSE estimator for robustness checks and sensitivity analysis. The BCMM results show that fiscal balance and political stability positively affect sovereign credit ratings, while external debt has a negative effect. The magnitude of these effects varies, and they are more pronounced in EMDEs. The empirical results also show that political stability makes fiscal balance have a net positive marginal effect on sovereign credit ratings, while it turns the negative effect of external debt positive, especially in EMDEs. However, the study finds no statistically meaningful non-linear thresholds for political stability, implying a monotonic relationship when political stability interacts with fiscal balance and external debt. Among the control variables, economic growth, FDI, trade openness, and financial market depth are significantly and positively associated with sovereign credit ratings, while the current account balance and inflation are significantly and negatively associated. In comparison, the study finds empirical evidence that advanced economies' more favorable ratings than EMDEs can be attributed to better debt management capacity, deeper financial markets, and greater global trade, along with lower risks from current account deficits and inflation. The findings have significant implications for policymakers and global rating agencies in formulating policies that ensure better, more objective sovereign credit ratings and competitiveness in international capital markets.

Abstract: The pandemic significantly disrupted Global Financial Markets, which had never experienced such unusual levels of volatility and uncertainty since the Covid-19 outbreak. The post-recovery phase saw continued price fluctuations due to inflation, geopolitical conflicts, and economic policy changes.This study examines the volatility and market behaviour of six major global stock market indices — S&P 500, FTSE 100, DAX, Nikkei 225, NIFTY 50, and Shanghai Composite — using daily data from January 2019 to April 2026. The data is divided into three sub-periods: pre-COVID (2019), COVID (2020), and post-COVID (2021–2026). A combination of statistical, econometric, clustering, and machine learning methods is applied, including descriptive statistics, correlation analysis, ADF test, ARCH LM test, GARCH (1,1), K-Means clustering, PCA, Random Forest, XGBoost, and Support Vector Regression (SVR) with an RBF kernel. The GARCH (1,1) model confirmed high volatility persistence across all markets. Among the machine learning models, Random Forest achieved the best overall prediction performance with a mean R² of 0.7678, mean RMSE of 0.003384, and mean MAE of 0.001973, followed by XGBoost (R²=0.7565) and SVR (R²=0.5851). The findings indicate that the COVID-19 pandemic was associated with persistent changes in global stock market volatility and behaviour that extended into the post-COVID period.

Abstract: Grid (averaging-down) trading models add larger orders as price moves against an open position and close the whole group, or basket, together. We study one automated gold (XAUUSD) grid model using every closed order from 2 January to 9 October 2026, rebuilt into baskets and matched to one-minute prices. A case is a basket at the moment its largest single order first reaches a lot level x; the outcome is whether the basket later opens an order of at least 2x ("runs on"), and the floating loss still ahead of it. The model changed its order-size steps on 12 August 2026, so the two settings are analysed separately. In the current setting (2,091 cases from 788 baskets), 29.4% of cases at 0.15–0.25 lots ran on, against 48.6% at 0.7 lots and above: a difference of 19.2 percentage points (95% basket-bootstrap interval 6 to 31; 14.5 points in a worst-case treatment of six baskets that could not be rebuilt). The share of baskets whose later floating loss exceeded 0.05% of the account balance rises from 3.7% at 0.15 lots to 50.9% at 0.7 lots. Time taken to reach a level is associated with running on in the earlier setting but not in the current one, and a 24-measure model ranks out-of-sample cases only slightly better than chance. All backtest results are exploratory. Hypotheses for a live test of the running model were published on 21 October 2026, before the first live case; this paper reports the design and the backtest only.

Abstract: We test whether round-number price levels predict reversals in gold, using 5,138 daily XAUUSD bars from 2005 to 2024. For every touch of a multiple of US $ 100 we record whether the close returns to the side of the level on which the day opened, and we compare the round grid with 80 placebo grids of identical spacing shifted away from round numbers. Inference is based on the round grid's rank among all 81 grids and on a month-block bootstrap. Among 232 eligible first touches (231 usable same-day observations), the same-day holding rate is 45.0% (104/231) against 49.2% for the placebo grids; the round grid ranks 72nd of 81 (rank-based p = 0.247), and no primary test survives a Holm correction. Results are similar for US $ 50 levels and for other first-touch windows. We find no robust evidence that round-number levels predict reversals under this daily-bar definition.

Abstract: AI-based credit scoring may broaden the information available to lenders while creating questions about explanation, unequal treatment and control over personal data. This article develops a Bangladesh-focused framework for perceived fairness, integrating the performance-expectancy component of UTAUT2, organisational justice and privacy-calculus reasoning. It also examines whether the supplied Group 8 Law-SDG database can support the proposed empirical model. The workbook contains 385 records and five credit-scoring statements, but it does not contain the proposed multi-item measures of usefulness, explainability, procedural justice, trust, privacy concern and overall perceived fairness. An item-level audit therefore replaces the earlier illustrative regression. Endorsement of disclosure about automated credit decisions is 82.3%, regular bias testing 83.6%, correction of inaccurate information 63.9%, relevant and lawful data use 63.6%, and the proposition that fairness affects acceptance 60.0%. These percentages describe supplied records rather than a representative population. The credit-scoring block has low internal consistency (Cronbach’s alpha = 0.193), precluding its use as a validated reflective fairness scale. The five proposed hypotheses remain untested. The article contributes a distinction between endorsement of safeguards, perceived fairness and acceptance, together with a measurement plan for evaluating their relationships. Recommendations concern understandable reasons, data correction, accessible review and documented model governance. Numeric-response provenance and digital-finance eligibility require confirmation; the codebook labels identifiers and comments synthetic. The study consequently offers a conceptual framework and a descriptive measurement audit, not a completed causal or structural test among verified borrowers.

Abstract: The Consumer Financial Protection Bureau (CFPB) publishes each complaint it forwards to a financial institution, the company's response category, and, until 2017, whether the consumer disputed it. The database records no company response date. We show that initial response timeliness is a compliance flag that varies for 2.4% of complaints, and that the only continuous timing field, the CFPB routing delay from receipt to transmission, is an administrative delay that precedes the company's involvement. In 737,010 complaints resolved by 4,092 institutions, with the dispute outcome recovered from an archived vintage, complaints closed with monetary relief are disputed 12.4 percentage points less often than similar complaints closed with an explanation (95% CI −13.4 to −11.4). The association survives fixed effects, narrative based severity controls, and sensitivity bounds, but the company chooses the response, so we report it as an association. Within firm, product, month, channel, and category cells, routing delay is associated with fewer disputes among complaints closed with an explanation (−1.1 points per log day, CI −1.4 to −0.8) or with no relief −1.8, CI −2.7 to −1.0), and with no change among complaints closed with monetary relief. Dispute indicators should be conditioned on routing delay, channel, and product.