LIQUIDITY STRESS TESTING OF BANKS: A BIBLIOMETRIC AND CONTENT ANALYSIS

The article systematizes scientific achievements in the field of liquidity stress testing of banks, using the methods of bibliometric analysis, particularly the VOSviewer toolkit. Unlike a traditional literature review, this approach eliminates subjectivity, considers the importance of scientific works by the number of citations, and more effectively reveals and analyzes research relationships using the co-occurrence approach, highlighting related words in the full texts of publications, abstracts, and titles. This made it possible to single out key trends in the study of liquidity stress testing in banks: 1) analysis of the impact of the regulatory landscape and requirements for liquidity stress testing on liquidity risk management practices aimed at reducing systemic liquidity risk and ensuring financial stability; 2) management of systemic liquidity risk in the system of macroprudential regulation and supervision, taking into account the interrelationships between the banking sector, financial markets and the real sector of the economy; 3) the use of liquidity stress testing as a tool for making strategic and operational decisions in the field of bank risk management to reduce the likelihood of liquidity crises and financial instability; 4) methodologies and models for formulating stress scenarios, calibrating liquidity risk indicators, estimating potential losses, and integrating liquidity risk assessment into its management system. Based on the temporal aspect of the bibliometric analysis, the following promising directions of research in the field of bank liquidity stress testing have been identified: assessment of the impact of changes in the regulatory landscape and requirements for liquidity stress testing on liquidity risk management practices and bank activity indicators; use of machine learning and artificial intelligence methods; development of tools for network analysis and system risk analysis; minimizing the risks of stress testing models by improving the toolkit for developing stress scenarios, automating the verification of stress testing systems and developing methods for managing the uncertainty of models.