University of Thessaly (UTH)
Evaluation of the financial and managerial role of stress tests in banking
Abstract
dc:descriptionThe present thesis examines the role of supervisory stress testing, focusing on EU-wide solvency stress tests conducted under the auspices of European Banking Authority (EBA) / European Central Bank (ECB) and on the ECB’s 2022 Climate Stress Test, which was an “unprecedented” supervisory exercise. In that spirit, the thesis highlights both the financial and managerial roles of stress tests. Although stress tests are often viewed mainly as supervisory tools for assessing bank resilience and capital adequacy under adverse scenarios, this thesis argues that their role extends beyond this traditional function. In particular, supervisory stress tests seem to produce information, reduce bank opacity, shape investor expectations, support market discipline, and may influence banks’ subsequent risk-management behaviour through supervisory scrutiny. In this sense, stress tests are examined not only as “supervisory tools”, but also as critical strategic assessments that affect how banks, investors, and supervisors interpret and respond to potential risks. The thesis develops the above perspective through four connected chapters and combines three main empirical approaches. The empirical approaches include bibliometric analysis, event-study methodology, and panel regression analysis, while robustness checks are also applied, where necessary, to further support the validity of the empirical findings. Chapter I provides the conceptual and bibliometric background of the current thesis. Leveraging 692 Scopus-indexed articles published between 1999 and 2024, it maps the evolution of the academic literature on stress tests in banking. The chapter incorporates descriptive bibliometrics, co-citation analysis, bibliographic coupling, keyword co-occurrence analysis, thematic mapping, and trend-topic analysis to identify the intellectual structure and thematic development of stress tests. The results show that stress testing research expanded substantially after the global financial crisis and after the institutionalisation of major supervisory frameworks, including the stress tests in Europe and the U.S. The literature has become increasingly mature, policy-oriented, and centered on themes such as financial stability, systemic risk, credit risk, capital adequacy, disclosure, market discipline, and macroprudential policy. At the same time, certain new areas such as climate stress testing, cyber resilience, machine learning, scenario analysis, and geopolitical risk have become increasingly visible. This chapter, therefore, positions the dissertation within a growing research field and identifies the informational and behavioural effects of stress testing as key areas for further empirical analysis. Chapter II investigates whether EU-wide solvency stress test disclosures convey value-relevant information to equity markets. The chapter focuses on the 2018 and 2021 EU-wide Stress Tests and treats each exercise as a sequence of supervisory disclosure events rather than as a single announcement. Specifically, it examines how market reacts to the publication of stress test methodologies, macroeconomic scenarios, and final results. As far as the methodological approach is concerned, the chapter applies an event-study framework using daily bank equity returns and national market indices for each stress tested bank of the sample. Abnormal returns are estimated through a market model over a 252-day estimation window, and the analysis focuses on cumulative abnormal returns and absolute cumulative abnormal returns around short event windows. The results show that solvency stress test disclosures are related to abnormal stock price movements, but the reactions vary across banks, disclosure events, and stress test cycles. The evidence indicates that investors respond not only to the publication of final results, but also to earlier supervisory disclosures, especially when these provide information on the structure, severity, or supervisory treatment of the exercise. This supports the view that solvency stress tests operate as “multidimension information mechanisms” that shape market expectations and contribute to market discipline. Chapter III extends the analysis of market reactions to the field of climate stress testing. In particular, it examines the ECB’s 2022 Climate Stress Test and related climate supervisory disclosures. These include the publication of the methodology, the release of climate-related macro-financial scenarios, the publication of the results, and the publication of the ECB’s Thematic Review on Climate and Environmental Risks. Regarding the methodology applied, the chapter also applies an event-study approach. It estimates abnormal returns, cumulative abnormal returns, and absolute cumulative abnormal returns for a sample of listed European banks that participated in the ECB climate exercise. This empirical design provides clarity whether climate-related supervisory disclosures contain information that investors incorporate into bank valuations. The findings show that climate stress test disclosures generated abnormal equity price movements, despite the exploratory nature of the exercise and the absence of direct capital implications. The evidence based on absolute cumulative abnormal returns is particularly relevant, as it shows that investors reacted to climate-related supervisory information even when average directional price effects were mixed. The chapter suggests that climate stress testing may serve as a disclosure tool that reduces information gaps around emerging environmental and transition risks. Chapter IV moves from market reactions to supervisory discipline and bank behaviour. It examines whether banks changed their balance-sheet risk profile after participating in the EU-wide 2023 Stress Test, using risk-weighted asset density as the main measure. Although Chapters II and III focus on public disclosure and market valuation, Chapter IV focuses on the internal supervisory discipline channel.The chapter applies a panel regression model with weighted least squares and a difference-in-differences approach. This compares the sampled banks’ risk profiles before and after the stress test reference date. The analysis uses actual bank-level data for 31 December 2022, 30 September 2023, and 31 December 2023, as well as analysts’ consensus forecasts for year-end 2023. The control variables include prudential and profitability indicators, such as the NPL ratio, debt ratio, Tier 1 capital ratio, liquidity ratio, cost-to-income ratio, return on equity, and interest income ratio.The findings show a statistically significant short-term reduction in RWA density after the stress test exercise. This result is consistent with a temporary disciplining effect of supervisory scrutiny on banks’ risk profiles. However, year-end 2023 data show an increase in RWA density which means that the effect weakens over the longer horizon. Also, analysts’ forecasts seem not to fully capture these realised adjustments, suggesting that market expectations may understate the behavioural effects of supervisory stress test exercises. All in all, the thesis provides three main areas of contribution. First, it provides a structured bibliometric overview of the evolution and thematic development of banking stress test research. Second, through event study analyses, it shows that both solvency and climate stress test disclosures contain value-relevant information for equity investors, although market reactions differ across banks, events, and supervisory cycles. Third, it provides evidence consistent with a supervisory discipline channel, showing that stress test participation is associated with short-term changes in banks’ risk profiles. The thesis highlights certain critical aspects of the role of supervisory stress tests for different stakeholders. As far as supervisors are concerned, stress tests enable them to identify vulnerabilities, communicate expectations, and improve the transparency of the banking sector. Regarding banking institutions, stress tests may affect capital planning, portfolio compositions, risk appetite, governance, and actions towards preparation for emerging risks. As for investors, stress test disclosures provide useful information for reassessing bank resilience and the risk management quality of the stress tested banks. The thesis concludes that supervisory stress tests are multidimensional tools of banking supervision. They assess resilience, produce information, support transparency, shape market reactions, and may influence banks’ strategy and risk-management behaviour.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Karakostas, Dimitrios
- Καρακώστας, Δήμητριος
Subjects
dc:subject × 16- Εποπτικές Ασκήσεις Προσομοίωσης Ακραίων Καταστάσεων
- Κλιματικές Ασκήσεις Προσομοίωσης Ακραίων Καταστάσεων
- Πειθαρχία της Αγοράς
- Μεθοδολογία Μελέτης Γεγονότων
- Εποπτική Πειθαρχία
- Supervisory Stress Tests
- Climate Stress Tests
- Market discipline
- Event-Study Methodology
- Supervisory Discipline
- Οικονομικά και Επιχειρήσεις
- Κοινωνικές Επιστήμες
- Χρηματοοικονομικά
- Economics and Business
- Social Sciences
- Finance
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/62042