Business, Economics and Management

Abstract: Theory predicts that higher trade policy uncertainty (TPU) makes firms delay costly trade decisions, and that rises in uncertainty matter more than falls. This study tests whether rises and falls in US TPU have effects of different size on real US imports from Bangladesh, measured in US customs data and deflated by an import price index. It estimates a nonlinear autoregressive distributed lag (NARDL) model on 253 monthly observations from 2005:07 to 2026:07, with log TPU split into partial sums of increases and decreases and US industrial production as the demand proxy. Bounds tests find a long-run level relationship, but TPU’s long-run effects are small. A 1% rise in TPU lowers real imports by an insignificant 0.035%. A 1% fall raises them by 0.075%, significant only at the 10% level with conventional standard errors. Long-run symmetry is rejected in the main model (p < 0.001) but not once a linear trend is added (p = 0.632), and short-run symmetry is not rejected. The data do not establish an asymmetric effect, and the larger point estimate is for falls in TPU rather than rises. NARDL studies of a trending uncertainty index should test long-run asymmetry with and without a trend.

Abstract: Clothing rental is a circular consumption model in the sharing economy and has attracted growing scholarly attention. This study retrieves 174 English-language articles on clothing rental, online clothing rental and subscription-based clothing rental from the Web of Science Core Collection. Using the Bibliometrix package in R, it analyses publication trends, journals, authors, countries and keywords. Output grew slowly from 2005 to 2012, steadily from 2013 to 2020 and rapidly after 2021. Sustainability and the Journal of Cleaner Production are the most prolific outlets, and the United States, China and the United Kingdom are the leading contributors. Research hotspots include consumer rental intention, perceived value and risk, ownership attitudes, business models, sustainability and digital technology. Together these findings offer a panoramic picture of clothing rental research and a reference point for future scholarship and industry practice.

Abstract: The Belt and Road Initiative (BRI) has strengthened economic cooperation between China and its partner countries. This study examines its estimated policy effect on China’s traditional Chinese medicine (TCM) trade and the conditions shaping that effect. Using balanced panel data on China’s TCM trade with 50 partners from 2005 to 2023, we applied a difference-in-differences model and multiple robustness tests. The BRI had a significant positive estimated effect on TCM trade, which remained robust to parallel-trends, placebo, and PSM-DID tests. The trade-promoting effect was stronger at higher levels of technological innovation, logistics infrastructure, and Internet penetration. It was also strongest for crude medicinal materials, more pronounced among politically stable partners, and greater during 2013–2018 than during 2019–2023. The results identify technological innovation, logistics, and digital infrastructure as relevant facilitating factors. The estimated gains were concentrated in less-processed products, particularly Chinese medicinal materials and herbal extracts. These findings quantify the response of medical-product trade to a large-scale international cooperation policy and provide mechanism-consistent evidence relevant to sustainable resource allocation and international health cooperation.

Abstract: Does digital adoption raise regional prosperity in the region where it occurs, or does it work through neighbouring regions? We assemble a georeferenced panel of 234 EU NUTS-2 regions observed from 2017 to 2024, linking Eurostat data on GDP per capita in purchasing power standards, household broadband and internet use, employment and social exclusion to NUTS 2024 boundaries. Global Moran's I is positive and significant for every variable and year (0.39–0.78), and LISA maps reveal persistent high-income and low-income clusters. Non-spatial estimators give a positive digital coefficient: 0.21% higher GDP per capita per percentage point of adoption under two-way fixed effects, and 1.83% when lagged adoption instruments a noisy, survey-based regressor. However, fixed-effects residuals remain spatially autocorrelated, and a Spatial Durbin Model (ρ = 0.70) shows that the own-region (direct) effect is not significant, while the spillover (indirect) effect is positive (0.20% per point, p = 0.039) and robust across three spatial weight matrices. Of six clustering algorithms compared on eleven validity indices, Ward hierarchical clustering performs best and identifies seven spatially contiguous regional profiles. Of six machine-learning regressors, a random forest validated by region-grouped cross-validation performs best (out-of-sample R² = 0.57). Its SHAP decomposition reproduces the econometric signs: own adoption dominates in levels, while neighbours' adoption and employment dominate within regions. Digital policy in the EU should therefore be targeted at the level of functional macro-regions, not individual regions.

Abstract: Improving resource allocation efficiency and fostering innovation are important dimensions of sustainable urban economic development. Public data trading platforms may contribute to these objectives by facilitating data circulation, reducing information frictions, and improving the coordination of productive resources. However, empirical evidence on their implications for urban productivity remains limited. This study examines the effects of public data trading platform establishment on urban total factor productivity (TFP), drawing on prefecture-level city panel data from China covering 2010–2024. By exploiting the staggered establishment of these platforms as a quasi-natural experiment, we employ a staggered difference-in-differences (DID) framework to estimate their productivity effects and investigate potential transmission mechanisms. The results indicate that public data trading platform establishment is positively associated with urban TFP after controlling for city and year fixed effects. The findings also reveal a positive association between platform establishment and technological innovation. Further mechanism analyses suggest that high-level human-capital agglomeration and the spatial agglomeration of scientific and technological resources may represent additional pathways linking data-market institutions to urban productivity. These findings highlight the potential role of institutional infrastructure for data exchange in supporting innovation-driven growth and more efficient resource utilization. From a sustainability perspective, the study provides evidence relevant to the socioeconomic and institutional dimensions of sustainable urban development and suggests that policies promoting data accessibility, effective data governance, and complementary innovation capabilities may help strengthen the foundations of long-term urban productivity. The results do not directly measure environmental sustainability and should therefore be interpreted primarily in terms of economic and institutional sustainability.

Abstract: Does access to transaction-level records affect which forecasting method public-finance analysts should choose? To answer this question, we compare four forecasting strategies applied to a short, highly seasonal time series of child-benefit payments: top-down, bottom-up, data-enrichment, and minimum-trace reconciliation. Our application concerns monthly child-benefit expenditure in Republika Srpska (Bosnia and Herzegovina), where we have approximately a hundred observations of total payments, but more than a million observations of individual payments. We also consider four different ways to deal with the recurring near-zero observations typical of administrative payment data. Under an expanding-window design, where each model can only use information available at the time of each forecast origin, exponential smoothing performs best across every treatment, with minimum-trace reconciliation yielding statistically equivalent accuracy. Enrichment with features from microdata offers no advantage when those features must be known in advance, while machine-learning techniques fail to outperform the benchmarks. The choice of treatment for near-zero observations proves as important as the forecasting method itself. Overall, in the presence of strong seasonality and short time-series, widely-used forecasting techniques remain hard to improve upon, and more data only helps if it provides genuinely forward-looking information.

Abstract: This paper asks which type of political behaviour drives spatial interdependence in municipal spending. Under yardstick competition, voters benchmark their incumbent against neighbouring jurisdictions, but the literature offers two opposite predictions about who mimics most: the swing voter hypothesis predicts that electorally vulnerable incumbents mimic more, while the core voter hypothesis predicts that politically strong incumbents mimic more. I extend the yardstick competition model to allow the spatial autoregressive parameter to vary across regimes defined by two dimensions of political strength, horizontal electoral concentration and vertical alignment with the federal president and the state governor, and take the predictions to a panel of 857-940 Mexican municipalities over 2001-2021 using regime-switching spatial Durbin models estimated by quasi-maximum likelihood with time-varying Jacobians. When political strength is measured by electoral concentration, politically weak mayors mimic around 60 percent more than strong ones, supporting the swing voter hypothesis. When it is measured by vertical alignment, aligned mayors mimic significantly more, supporting the core voter hypothesis. Four-regime specifications reveal an asymmetry: national alignment combined with concentration reinforces the core voter pattern, regional alignment combined with concentration reverses the concentration effect, and the joint combination of both vertical alignments restores the swing voter pattern. The two mechanisms are not mutually exclusive but operate along different dimensions of political strength.

Abstract: Research on artificial intelligence (AI) and organizational change is growing fast but fragmented. Two perspectives coexist: one uses AI, defined here as machine learning, natural language processing, generative AI and related intelligent systems, as a tool to diagnose, design, implement and monitor change; the other studies AI as the object whose adoption triggers change. We map the field and how the two perspectives relate. Of 1,402 Scopus records, a large language model applying a criterion calibrated on the author's judgment retained 525 documents (2000–2026). We combined performance analysis, co-citation, bibliographic coupling, keyword and topic analysis, and a coding of change phases. The field has grown by almost 70% a year since 2018, mostly through research on AI as an object; the share of studies using AI as a tool fell from 28% up to 2020 to 11% in 2024–2026. The two perspectives rest on largely separate references. Research on AI as an object forms eight research fronts; research on AI as a tool forms none, and almost half of its documents are disconnected from the main coupling network. Only 32 documents propose or evaluate AI as an instrument for managing change. Diagnosis dominates and studies that use AI across several phases of change are rare. These findings hold under alternative inclusion criteria, except the absence of a tool-based front, which reflects the narrow definition of AI adopted. We argue that the field lacks integration and validation rather than new algorithms, and propose an integrative framework with six propositions.

Abstract: Long-term energy forecasting in a dynamically changing environment requires combining information contained in historical data with expert knowledge about structural, technological and policy changes. However, expert scenarios often disagree, and their relative credibility is difficult to assess. We propose a framework for combining a statistical forecast with expert scenarios while explicitly retaining uncertainty about both scenario outcomes and their relative importance. We examine this problem using gas demand in Poland up to 2040. A statistical forecast based on historical data provides a transparent reference, while expert scenarios developed by three organizations represent alternative structural developments that may not be captured by historical extrapolation alone. Uncertainty around the statistical forecast is represented by a probability distribution, while expert deviations from it are represented by triangular fuzzy numbers. We fix the total weight assigned to expert alternatives but leave its allocation among them unspecified. The combined result is a probability box, which gives lower and upper probability bounds. In the current case study, all expert scenarios imply higher demand than the statistical reference in 2030, while their demand levels differ strongly in 2040. In 2035, the midpoint of the combined expected-value interval remains close to the statistical reference despite disagreement among scenarios. Giving more weight to expert alternatives raises this midpoint in 2030 but lowers it in 2040. The results show that combining statistical forecasts with expert knowledge can reveal the dependence of long-term conclusions on structural assumptions and scenario credibility, particularly when historical relationships may not remain stable. The proposed framework therefore supports a transparent synthesis of statistical evidence and expert judgement without requiring a single precise weighting of alternative future scenarios.

Abstract: A researcher who backtests many trading strategies and keeps the best one faces two questions: how much to believe the winner's backtest, and whether searching more candidates leads to a better strategy. We show that, when strategies are sampled at random from a fixed pool, the second question has an exact answer: the expected out-of-sample payoff at any search width is a known weighted average of the leaderboard, the out-of-sample payoff of each in-sample rank. We measure that leaderboard for 652 mechanical trading signals on U.S. mega-cap stocks (2010-2025) and on industry and size/value portfolios (1970-2025), trading one day after the signal and paying realistic spreads. Three results stand out. The backtest always overstates, and the overstatement grows with search width. Net of costs, searching more candidates picks a less costly strategy, mainly slow momentum, so the chosen strategy improves relative to a narrow search on the diversified panels. But at no width does any rule, including shrinkage and combination, find a strategy with a reliably positive return after costs.

Abstract: This study examines the distributional heterogeneity of circular economy (CE) energy transitions and their relationship with economic growth across six BRICS economies—Brazil, China, India, South Africa, Egypt, and Indonesia—from 2000 to 2025. Using the Method of Moments Quantile Regression (MMQR) as the primary estimator, with fixed effects as a robustness check, we assess effects across the growth distribution while accounting for cross-sectional dependence and panel integration properties. We also provide a direct estimate of the energy rebound effect for the BRICS panel. The results reveal substantial heterogeneity that mean-based estimators obscure. Renewable energy share and energy intensity are positive and statistically significant mainly at middle-to-high growth quantiles (Q50–Q90). In contrast, gross capital formation and FDI are robust positive drivers across all quantiles. Urbanization is negatively associated with growth at Q25–Q90, while trade openness and domestic credit are largely insignificant in the MMQR specification. The rebound-effect estimate indicates that approximately 25.7% of energy efficiency gains are offset by increased consumption, implying net savings of about 74.3%. These findings suggest that CE energy transitions do not affect all economies uniformly: investment and foreign capital remain broad-based growth engines, whereas energy transition dividends are concentrated in faster-growing economies. The study offers differentiated policy implications for low-, middle-, and high-growth BRICS economies and cautions against one-size-fits-all CE prescriptions.

Abstract: Accurate forecasting of systemic risk is crucial for financial stability and regulatory oversight. Traditional parametric models often fail to capture the complex, nonlinear dependence structures in financial markets, particularly during periods of distress. This paper proposes a semi-parametric dynamic conditional correlation (DCC) framework that integrates nonparametric estimation of marginal distributions with a parametric DCC specification to improve the forecasting of systemic risk measures—Conditional Value-at-Risk (CoVaR) and Marginal Expected Shortfall (MES). We employ a kernel-based quantile regression approach for marginal modeling and combine it with a DCC-GARCH structure for capturing time-varying conditional correlations. Empirical results using U.S. financial institutions’ equity returns during the 2007–2009 crisis and the post-crisis period demonstrate that the suggested semi-parametric DCC framework yields more accurate and robust forecasts of CoVaR and MES compared to fully parametric alternatives. Importantly, the SP-DCC is the only model that maintains correct unconditional and conditional coverage at the 1% VaR level during stress periods, while Gaussian-DCC significantly over-rejects. The model also exhibits superior performance in identifying systemically important institutions and tail interdependencies during stress episodes. Our findings underscore the importance of flexible marginal modeling in systemic risk measurement and provide policymakers with an enhanced toolkit for macroprudential surveillance.

Abstract: Classical cattle-cycle theory was developed mainly in temperate, capital-intensive livestock systems, leaving its external validity under tropical, extensive production conditions less well established. Using 851 monthly slaughter observations from Colombia (1954–2024), this study integrates seasonal ARIMA and spectral methods, Bai-Perron break tests, a five-variable VAR, external-shock ARIMAX tests, and a census-anchored Kalman reconstruction of herd inventory. The Christiano-Fitzgerald component yields an average complete-cycle duration of approximately 6.1 years, while multitaper evidence is multimodal, with concentrations around 4.4–5.0 and 7.1–7.8 years. Under the reference Bai-Perron specification, BIC selects seven breaks defining eight regimes; the first four dates are stable across nearby specifications, whereas later dates are more sensitive. Slaughter's own component explains 63.6–65.2% of its 24-month forecast-error variance across alternative Cholesky orderings. Conventional VAR Granger results do not survive Toda-Yamamoto robustness tests, and climate, interest rates, and the real exchange rate add no robust predictive content. The reconstructed herd shows a marked 2011–2019 contraction followed by recovery through 2024, although its magnitude is calibration-sensitive. The results extend evidence on retention-liquidation dynamics to a tropical, extensive, dual-purpose system while showing that individual channels remain context-dependent.

Abstract: Causal estimators and representation-learning methods implicitly define notions of similarity: covariate matching asks which units look alike, propensity methods ask which units share treatment-assignment probabilities, heterogeneous-effect learners group units with similar treatment effects, and forest methods induce adaptive neighborhoods through repeated partitioning. These geometries need not coincide, and a geometry useful for one causal task may be inappropriate for another. We develop a task-specific view of causal geometry and focus on the geometry required for explicit counterfactual substitution. We define counterfactual geometry through treatment-specific response surfaces, so that two units are close when their expected potential outcomes are similar under both treatment states. We further introduce Counterfactual Neighbor Regret (CNR), which measures the loss in counterfactual-neighborhood quality incurred by a learned geometry relative to the best opposite-treatment neighborhood available in the sample. A finite-sample approximation result shows that if a learned metric uniformly approximates the oracle counterfactual metric, its selected neighbor has counterfactual-distance regret bounded by twice the metric approximation error plus neighbor-search error; this yields a direct bound on conditional matching bias. We study covariate, propensity, prognostic, treatment-effect, forest-induced, outcome-informed, and counterfactual geometries in simulations designed to separate treatment-effect similarity from counterfactual substitutability. Oracle counterfactual geometry is consistently strongest, while treatment-effect and propensity geometries can select poor counterfactual substitutes despite being useful for their native tasks. Outcome-informed stretching substantially improves matching, and augmenting it with estimated counterfactual coordinates yields modest additional gains and lower CNR. Preliminary IHDP experiments show that no learned geometry dominates universally. These findings motivate a broader principle: causal representations should be evaluated against the causal task they are intended to support, not by generic representation quality alone. We further introduce arm-specific orthogonal counterfactual coordinates (OCC): cross-fitted AIPW pseudo-outcomes for each treatment-specific response surface. Because raw orthogonal scores are noisy individual labels, OCC is denoised before being used as an auxiliary geometry teacher; in a nuisance-misspecification stress test, the hybrid nuisance+OCC geometry substantially lowers Counterfactual Neighbor Regret relative to direct nuisance-response geometry.

Abstract: Hourly forecasting of electricity load is a very important problem from the practical point of view, due to the mostly big spread on one-day energy markets. Many models have been proposed in the literature. Our idea is to combine five individual models through multiple regression estimated on last 48 hours of the demand time series. The individual models are: multiple regression with dummy variables coding week days and hours, ARIMA, exponential smoothing, support vector regression, and artificial neural regression. Residuals of these models were later analysed by ARIMA. The basic harmonic in time series was equal to 168 (7 days x 24 hours). Individual models theoretical values in last 48 hours were used to estimate regression model combining individual models. The combined model fitted the data considerably better in-sample, and its 48-hour forecast error was slightly lower than for the individual models (MAPE 3.57% vs 3.65–4.44%). The paper proposed some kind of framework for practical use based on rolling data, each day adding the latest 24-hour data, and eliminating 24 oldest observations. The structure of individual models can be kept constant, at least for some months.

Abstract: Geopolitical conflicts and the energy transition complicate risk linkages among conventional energy, new energy, ESG assets and gold. Using quantile time–frequency connectedness and complex network analysis, this study examines an ESG index, coal and natural gas equity indices, five new energy subsector indices, an aggregate new energy index and a gold ETF across market states and horizons. Under normal conditions, wind power, photovoltaics, natural gas and new energy vehicles are net transmitters, while the remaining markets are net receivers. Total spillovers rise sharply under extreme conditions, and short-term spillovers dominate across market states. Static estimates show stable transmission roles for photovoltaics, wind power, natural gas and gold across quantiles and frequency bands, alongside conditional switching in other markets. Dynamic estimates reveal additional variation over time. The Russia–Ukraine conflict is associated with a sustained increase in long-term connectedness, whereas COVID-19 primarily coincides with short-term volatility pulses. Network analysis identifies coal as a central node and photovoltaics as a major transmission hub within new energy. These findings inform market-state- and horizon-specific risk monitoring, asset allocation and crisis responses.

Abstract: Italy grants registered innovative start-ups fee exemptions, investor tax relief and guaranteed credit, and removes that status sixty months after incorporation. This paper asks whether certification changes firms’ markups. Markups are estimated from a translog production function on 76,466 firm-years of certified start-ups, certified innovative SMEs and ordinary SMEs over 2015–2024. Certified firms price about 12 per cent below ordinary SMEs, but within firms the gap shrinks to 2 per cent and vanishes once persistence is modelled: registration selects firms rather than changing their pricing. Exploiting the statutory expiry, and a 2020 extension that shifted it by one year for adjacent cohorts, the loss of status leaves markups unchanged on average (2SLS −0.005, first-stage F = 750). An unsupervised partition shows this null combines falls of five to six per cent in capital- and materials-intensive firms with none elsewhere; regression learners and double machine learning confirm the central estimates.

Abstract: Studies of small-firm profitability estimate financial structure as the focal regressor and treat cost composition as an unexamined control. This paper reverses the emphasis and places both blocks in one equation, estimated on 91,756 firm-year observations covering 14,913 unlisted Italian firms across three regulatory regimes. Three methods interrogate different assumptions on the same sample: panel estimation across fourteen specifications, unsupervised partitioning selected on eleven validity criteria, and machine-learning regression applied to the same firm-demeaned data. The capitalisation coefficient is not identified. It moves from +0.177 under two-way fixed effects to between −0.12 and −0.28 under five instrument sets, all of which fail because profitability persists at 0.271, so any lagged financial ratio embeds past profit. The labour-share coefficient is stable across estimators, groups and functional forms, and the paper tests whether that stability is arithmetic, since return on assets and the labour share are consecutive lines of one statement. It is not: value added over total assets varies by a factor of 2.8 across the four operating archetypes recovered from the data, while the coefficient varies by 1.07. The accounting identity explains part of the magnitude and none of the stability.

Abstract: This paper investigates the macroeconomic determinants of renewable energy consumption in Saudi Arabia during 1990–2025 using a hybrid approach of ARDL bounds testing and Random Forest machine learning with SHAP (SHapley Additive ExPlanations) to improve model interpretability. We employ annual data from the World Development Indicators to examine the effect of high-technology exports, trade openness, foreign direct investment, inflation and GDP on renewable energy consumption. Our results indicate that renewable energy consumption is highly path dependent, with previous adoption playing a significant role in current consumption (β = 0.776, p < 0.01). Trade openness also shows a negative contemporaneous effect ( = 0.0013, p 0.05) in line with the carbon lock-in hypothesis that typifies hydrocarbon-dependent economies. High-technology exports as a proxy for technological innovation do not appear to be a significant driver, indicating that the innovation-led energy transition in Saudi Arabia is still maturing. The Random Forest model confirms the importance of persistence effects explaining about 73 % of the variance. The analysis of the different time periods indicates that, after 2016 (also known as Vision 2030), there has been a significant change in the structure of the analyzed data, with the statistical model of research becoming much stronger as its explanatory power has reached up to 83%. The analysis of the SHAP allows understanding of the obtained results by measuring the influence of each individual variable. The findings that were obtained pertain to Sustainable Development Goal 7 (which focuses on developing cheap and clean energy), Sustainable Development Goal 9 (which focuses on industry, innovation, and infrastructure), and Sustainable Development Goal 13 (which focuses on climate action). It is our proposal that Saudi Arabia adopt focused innovation strategies, careful management of its trade integration, and a commitment to the institutional framework in order to accelerate the transition to renewable energy. However, it is important to note that this recommendation is suggestive and not definitive.

Abstract: Water stress is read in the sustainability literature as either a physical or an institutional outcome, but the two readings have never been tested jointly within the sovereign ESG framework. This paper estimates the environmental, social and governance determinants of the freshwater withdrawal-to-availability ratio across up to 170 economies observed between 2002 and 2022, drawing on the World Bank Sovereign ESG Data Portal. Three equations share the same dependent variable and differ only in the pillar supplying the regressors. Each is developed along three tracks: panel econometrics, comprising pooled OLS, fixed and random effects, between and weighted least squares estimators and two-step difference and system GMM; a competition among six clustering algorithms assessed on eleven internal validity indices; and a competition among six supervised learners used to validate the specification rather than to forecast. Water stress proves overwhelmingly structural, with 99.1 per cent of its log variance lying between countries. The governance block explains 38 per cent of cross-sectional and 4 per cent of within-country variation. The level of institutional quality, measured as the first principal component of four Worldwide Governance Indicators and absorbing 91.9 per cent of their variance, is unrelated to water stress in every estimator; its composition is not, with accountability relative to state capacity carrying a coefficient of −1.10 between countries. The dominant predictor across all three tracks is the female-to-male labour force participation ratio, whose partial dependence reveals a discontinuity concentrated in a five-point interval. Under cross-validation with countries held out, no flexible learner outperforms ordinary least squares.