With the rising environmental concerns among consumers all over the world, sustainability has received considerable attention, and numerous enterprises are adopting various practices such as investing in energy-saving to improve sustainability in supply chains. However, many previous researches always assume that decision makers are perfectly rational and neglect the behavioral concerns of decision makers. This paper considers a two-stage sustainable supply chain with behavioral concerns in order to develop more realistic models, and mainly focuses on the energy-saving and pricing decisions in the decentralized system, as well as how to improve energy-saving level and profits. We develop decentralized decision-making models under two types of behavioral concerns: fairness concern and risk aversion, and derive the optimal strategy for each member with a Stackelberg game in which the manufacturer acts as the leader. The effect of the behavioral concerns on the optimal decisions and corresponding profits is discussed in detail. Theoretical analysis verified by numerical experiments shows that the fairness behavior always causes a negative effect on the manufacturer, total supply chain, and energy conservation, while it could benefit the retailer in profits. The risk aversion behavior always benefits the manufacturer, total supply chain, and energy conservation, whereas it could make the retailer suffer. Note that both the optimal energy-saving level and corresponding profit of the total supply chain under two types of behavioral concerns are lower than that in the centralized system, thereby we propose a revenue-cost-sharing contract to coordinate the supply chain, under which both the manufacturer and the retailer can achieve a win-win outcome and the energy-saving level can be improved. In addition, some managerial implications through our analytical and numerical results are summarized in this paper.

Supply chain finance (SCF) operations require extensive activities and a high level of information transparency, making them vulnerable to operational issues that pose significant risks of financial loss for commercial banks. Accurately assessing operational risks is crucial for ensuring market stability. This research aims to provide a reliable operational risk assessment tool for commercial banks’ SCF businesses and to deeply examine the features of operational risk events. To achieve these goals, the study explores the dependency structure of risk cells and proposes a quantitative measurement framework for operational risk in SCF. The loss distribution analysis (LDA) is improved to align with the marginal loss distribution of segmented operational risks at both high and low frequencies. A tailored copula function is developed to capture the dependency structure between various risk cells, and the Monte Carlo algorithm is utilized to compute operational risk values. An empirical investigation is conducted using SCF loss data from commercial banks, creating a comprehensive database documenting over 400 entries of SCF loss events from 2012 to 2022. This database is analyzed to identify behaviors, trends, frequencies, and the severity of loss events. The results indicate that fraud risk and compliance risk are the primary sources of operational risks in SCF. The proposed approach is validated through backtesting, revealing a value at risk of CNY 179.3 million and an expected shortfall of CNY 204.9 million at the 99.9% significance level. This study pioneers the measurement of SCF operational risk, offering a comprehensive view of operational risks in SCF and providing an effective risk management tool for financial institutions and policymakers.

Customer churn poses a significant obstacle for rail freight, and accurately predicting railway freight customer churn is necessary to improve customer retention and enhance the competitiveness of the railway freight market. The imbalance in the data set of railway customer shipment behavior, however, affects the prediction performance. This study introduces a strategy for predicting customer churn in rail freight using the Stacking technique. The approach establishes clear and measurable standards for identifying customer churn. It merges customer churn labels and behavior features with a thorough Recency, Frequency, and Monetary (RFM) model that tracks railway customers’ shipment behavior into a unified dataset. The dataset is then divided and sampled using a cross-time window. Additionally, a Stacking ensemble model is constructed using the Random Forest, Light Gradient Boosting Machine, and Gradient Boosting Decision Tree algorithms as the base learners, with the Logistic Regression algorithm serving as the meta learner. This model is then used to conduct comparative experiments between the Stacking ensemble and the base learners across different time windows. Using the freight waybill data of a railway firm from 2015 to 2019 as a case study for experimental validation, the results demonstrate that the Stacking model outperforms all the base learners in terms of the area under the curve (AUC), accuracy, precision, recall, and F1 values. Among the samples, the AUC is 0.8461, whereas the AUC across time samples is 0.8853. The Stacking model demonstrates high efficacy in predicting the likelihood of a single customer’s churn, exhibits improved accuracy in predicting customer churn, and possesses robust generalization capabilities, thereby addressing the limitations associated with imbalanced datasets. The impact of customer shipping behavior features on the prediction model reveals that customer shipping density plays a crucial role in predicting the likelihood of rail freight customers discontinuing their services.