Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

A multicriteria decision framework for the assessment of banks' capital needs

Tsagkarakis Minas-Polyvios

Full record


URI: http://purl.tuc.gr/dl/dias/056B2DF2-83FD-4619-807C-E84A17D1DD38
Year 2021
Type of Item Doctoral Dissertation
License
Details
Bibliographic Citation Minas-Polyvios Tsagkarakis, "A multicriteria decision framework for the assessment of banks' capital needs", Doctoral Dissertation, School of Production Engineering and Management, Technical University of Crete, Chania, Greece, 2021 https://doi.org/10.26233/heallink.tuc.91275
Appears in Collections

Summary

The global financial crisis affected significantly the soundness of individual banks and the health of the U.S. and the European banking system as a whole. Building on the outcomes of past regulatory exercises and decisions to capitalized weak banks, this thesis propose the development of an early-warning system that could serve in the future as an automated decision support system for the continuous monitoring and timely identification of weak banks, subsequently guiding regulatory decisions as for the capitalization needs of banking institutions. At the same time, bank managers could use the model to know in advance if their bank is developing a profile that is close to the one that would trigger supervisory actions.Within this context, the proposed approach is based on a multicriteria decision aid (MCDA) technique, the UTADIS method (UTilitè Additive DIScriminantes), which enables the development of additive models for decision making and prediction purposes in a classification setting. The additive form of the models facilitates their interpretability, which is an important feature for their use in a regulatory context. For comparison purposes we benchmark the UTADIS models against logistic regression, as well as with two widely used measures the SRISK, and the Texas Ratio.Using a sample of 76 large U.S. and European financial institutions and a set of 22 criteria across different dimensions related to bank-level risk factors, bank-level microeconomic criteria, and banking and financial market aggregate conditions we developed various multi-attribute models to distinguish between banks with capital needs and well-capitalized ones. This allows us to build a decision support framework that captures vulnerabilities from both a microprudential and a macroprudential perspective.

Available Files

Services

Statistics