Back to search

Institutional Repository of Vilnius University

Kreditų rizikos ir stresinių testų modelių lyginamoji analizė finansų sektoriuje /

Abstract

dc:description

Credit risk management and the application of stress testing are critically important tools in the financial sector, helping to ensure the stability and resilience of institutions under dynamic economic conditions. In modern banking, credit risk models have become not only computational tools sbut also strategic decision-making instruments that directly influence capital adequacy, credit pricing, and client evaluation. Stress tests serve both preventive and educational purposes – they promote the reduction of systemic vulnerability, the strengthening of capital reserves, and the optimization of internal processes. Over the past decade, risk management has undergone a fundamental methodological shift. Traditional models based on statistical relationships have evolved into advanced artificial intelligence solutions, the integration of ESG (Environmental, Social, and Governance) factors, and the application of XAI (Explainable Artificial Intelligence) systems. This transformation fundamentally changes the understanding of risk – model quality is now assessed not only by predictive accuracy but also by explainability, data provenance, and sensitivity to ESG changes. The aim of this thesis is to conduct a comparative analysis of credit risk models and stress testing methodologies applied in commercial banks operating in Lithuania, assessing the differences between model types, the level of ESG and XAI integration, and decision quality parameters. It evaluates the differences between applied models, the level of ESG and XAI integration, and the parameters of decision quality. Quantitative methods were applied in the research, including variance analysis, regression, and SHAP value interpretation. Data from the three largest commercial banks in Lithuania – AB „SEB bankas , AB „Swedbank and AB „Luminor – were analyzed for the period 2020–2023, covering PD, LGD, ECL, ESG indicators, and stress test results. The paper provides practical recommendations for ESG data integration, model calibration, and ensuring decision transparency in small open economies such as Lithuania. The analysis revealed that advanced models demonstrate high predictive accuracy but often pose interpretability challenges. Meanwhile, hybrid models that combine XAI and ESG corrective layers offer greater clarity in decision-making and regulatory compliance. This study is relevant both to the academic discussion on the evolution of risk models and to banking practice in the search for sustainable, transparent, and regulation-compliant solutions.

Degree

thesis:*
Grantor dc:publisher
Institutional Repository of Vilnius University
Year dc:date
2025

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Pritulskis, Rytis,

Rights

dc:rights
Statement dc:rights
  • info:eu-repo/semantics/openAccess
Language dc:language
lit

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
Vilnius University
Base URL
epublications.vu.lt/oai
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Pritulskis, Rytis,. Kreditų rizikos ir stresinių testų modelių lyginamoji analizė finansų sektoriuje /. Institutional Repository of Vilnius University, 2025. https://repository.vu.lt/VU:ELABAETD239021634&prefLang=en_US