Predicting corporate bankruptcy is a key task in financial risk management, and selecting a machine learning model with superior generalization performance is crucial for prediction accuracy. This study evaluates the effectiveness of k-fold cross-validation as a model selection strategy for random forest and XGBoost classifiers using a publicly available dataset of Taiwanese listed companies. We employ a nested cross-validation framework to assess the relationship between cross-validation (CV) and out-of-sample (OOS) performance on 40 different train/test data partitions. On average, we find k-fold cross-validation to be a valid selection technique when applied within a model class; however, k-fold cross-validation may fail for specific train/test splits. We find that 67% of model selection regret variability is explained by the particular train/test split, highlighting an irreducible uncertainty real world practitioners must contend with. Our study extensively explores hyperparameter tuning for both classifiers and highlights key insights. Additionally, we investigate practical implementation choices in k-fold cross-validation—such as the value of k or prediction strategies. We conclude that k-fold cross-validation is effective for model selection within a model class and on average, but it can be unreliable in specific cases or when comparing models from different classes—this latter issue warranting further investigation.

Following the financial crisis of 2007 and the sovereign debt crisis in 2010 that affected the soundness and reduced the strength of public finance in European countries, there has been a growing interest in developing methodologies to the help assess and signal the vulnerability of fiscal policy. Therefore, the aim of this study is to develop a new framework (V-L-D) to assess fiscal vulnerability. V-L-D represents a new methodology on the measurement of fiscal vulnerability that relies on the assumption that vulnerability can occur even during calm times. In comparison with previous methodologies that studied fiscal vulnerability around crisis and fiscal distress times, our framework investigates fiscal vulnerability near fiscal adjustments episodes. Our methodology relies on two distinct indicators: one showing the vulnerabilities indicated by the level of the cyclically adjusted budget balance and distance-to-stability, and one showing the vulnerabilities pointed out through the changes of the cyclically adjusted budget balance and public debt. V-L-D is able to classify fiscal vulnerability into five distinct categories having scores from 0 (no fiscal vulnerability) to 4 (extreme fiscal vulnerability). Using annual data ranging over 1990–2013 for 28 European Union countries, we evidenced 310 episodes of fiscal vulnerability, out of which 128 episodes of low vulnerability, 94 of moderate, 62 of strong, and 26 of extreme fiscal vulnerability. We also found that over 2004–2013, Greece, Portugal, Romania, United Kingdom, Ireland, Spain, and Slovenia were the most fiscally vulnerable countries in the Union. United Kingdom and Greece went through the longest episodes of fiscal vulnerability, counting 12 and 11 consecutive years, respectively. We tested our framework’s effectiveness against the Excessive Deficit Procedure. We found that the overall performance is good: V-L-D assessed moderate fiscal vulnerability during the procedure, strong fiscal vulnerability in the first year when procedure was initiated, and extreme vulnerability one year before the initiation.

As we are witnessing a shift of paradigm from traditional economic growth towards sustainable development, backed by United Nation’s Sustainable Development Goals call to action, the concept of Environmental, Social and Governance (ESG) has been integrated in almost all aspects of the economy. Firms are probably the category of stakeholders towards the most expectations have been made for them to do better in terms of ESG. Significant scrutiny is being put on firms activating in traditionally emission-intensive sectors however, one cannot deny the aspect that the polluting sectors of economy have been the main pillar of economic growth since the industrial revolution. The current research aims at studying in what way the ESG performance of polluting firms is backed by improved financial performance and whether greening these firms leads to a trade-off or a win-win situation. By doing so, the paper aims at filling a gap in the existing literature as there is significant evidence concerning the relationship between ESG and financial performance in general, while a limited number of studies focus on firms activating in emission-intensive sectors. We perform the analysis using panel data regression models for a set of 138 firms registered across the European Union and over a period of 14 years. We look at financial performance from both a profitability and as well as market value perspective and the findings indicate that improvements in governance and social performance positively influence return on assets (ROA) and return on equity (ROE), while environmental performance, as well as the aggregated ESG score, has a limited impact, particularly in the case of Tobin’s Q. The study brings valuable insights to both business actors and policymakers by highlighting that the social and governance goals can be attained without sacrificing financial performance, while greening polluting firms may be perceived as burdensome and could imply financial trade-offs.

Identifying the reasons behind the value added tax gap is important for policymakers because it helps them understand areas in which tax compliance can be improved, and revenue losses decreased. The purpose of this paper is to examine the determinants of value added tax gap among a panel of 26 European Union countries during the period from 2000 to 2022. Through the analysis of the relationship between macroeconomic variables and the value added tax gap, the present work intends to provide useful information that could help making decisions that could help increase compliance rates and reduce the loss of revenues. The complicated relationship between income inequality and the value added tax gap requires further investigation into its mechanisms. The value added tax gap is not a universal concept but rather a result of complex interactions of various economic, institutional, human capital, and social factors that occur differently depending on the type of compliance. In high- value added tax gap countries, value added tax evasion occurs because of income inequality, underinvestment, and informal economy. On the other hand, in low- value added tax gap countries, compliance is achieved through educational systems, stability of the labor market, and proper exemptions. On the whole, the results show that policies which are context-sensitive and take into consideration structural inequalities in high-gap member states and institutional strengths in low-gap member states are crucial for successfully decreasing the value added tax gap within the European Union.