Alina Yefimenko

Bank capitalization is a key factor in the banking system’s resilience to economic shocks, as outlined in the Basel Accords on banking regulation. The paper aims to empirically evaluate the connection between bank capitalization and the socio-economic development level. The study uses World Bank data across 34 European countries of varying income levels (high, upper-middle, and lower-middle) for the years 2010, 2015, and 2020 as periods of financial and socio-economic turbulence. The study applies principal component analysis and correlation methods to identify relevant indicators. Countries were grouped using hierarchical clustering and K-means clustering methods. The results of the variance analysis revealed that only two indicators of bank capitalization – the share of non-performing loans and return on assets, and three socio-economic indicators – the Gini index, inflation, and unemployment, were statistically significant. As a result of both qualitative and quantitative changes between clusters, countries were reallocated from four clusters in 2010 and 2015 to three clusters by 2020: high-income countries, including Austria, Belgium, Denmark, Ireland, Iceland, Latvia, and others, consolidated into a single stable cluster (Cluster 3) in terms of both bank capitalization and socio-economic development. The primary drivers of inter-cluster movements were the positive dynamics of non-performing loans and bank assets, superior bank capitalization ratios, and improvements in inflation and unemployment levels. Based on the study’s findings, a matrix of transformational sustainability among European countries has been developed. The analysis confirms a relationship between the components of the chain: “level of bank capitalization – level of socio-economic development”.

Type of the article: Theoretical Article

Abstract
National cybersecurity opens new opportunities for digital transformation and helps counter the growing threat of cyberattacks. The purpose of this study is to develop a methodology and identify the most influential predictors of national cybersecurity in global competitiveness, utilizing the data mining approach. To assess the impact of cybersecurity characteristics, modern machine learning methods were implemented in the Salford Predictive Modeler software environment, including Classification and Regression Trees, Random Forest, Stochastic Gradient Boosting, Multivariate Adaptive Regression Splines, Generalized Path Seeker, and classical regression. Data mining was conducted using the World Competitiveness Ranking (WCR) indicators, constructed by the Institute for Management Development, and the National Cybersecurity Index (NCSI), constructed by the e-Governance Academy, for 64 countries as of 2024. Three groups of countries were identified with different predicted levels of competitiveness, underscoring the importance of balancing the development of both external and internal components of cyber defense. The empirical results demonstrate that countries with strong cybersecurity management capabilities achieve higher competitive positions, even with limited participation in global cyber initiatives. At the same time, active global involvement without sufficient internal readiness does not create sustainable competitive advantages. The regression tree provided a transparent allocation of key subfactors, with global cybersecurity contribution and cybersecurity management playing leading roles. The significance of the study lies in shaping methodology and state policy in cybersecurity, human resource development, regulatory framework enhancement, and digital ecosystem strengthening.

Acknowledgments
The paper is prepared within the scientific research projects “Digital transformations to ensure civil protection and post-war economic recovery in the face of environmental and social challenges” (№0124U000549) and “Drivers and barriers to human capital transformation for a circular and green economy” (No. 0126U001080).