Institutional Repository of Vilnius University
Credit union profitability estimation and prediction: the case of lithuania /
Abstract
dc:descriptionThe purpose of this master thesis is to investigate main determinants of Lithuanian credit unions profitability and to predict future credit union profitability values. The examination of main profitability factors has been done by panel autoregressive distributed lags model with the error correction term (ARDL). Estimated panel ARDL model, regression trees and more novel tree-based algorithms have been applied to predict future profitability ratios. The results indicate that country economic activity, average loan interest rates of banks and ratio of operating expenses to total income are significant long-term profitability determinants. Additionally, it was found that in short-term, changes in capital adequacy ratio and net interest income have positive impact to credit unions’ profitability, while growth of worsened loan portfolio quality, have negative consequences on credit unions profitability. Comparison of predictions for different forecasting horizons of the fitted statistical panel ARDL model and three treebased methods, indicate that boosted trees predicts credit union profitability with the lowest prediction error. Moreover, it was shown that in some cases, prediction errors of panel ARDL models were sufficiently close to the boosted trees. The findings of the study provide better understanding of internal and external credit union profitability determinants and the overall insights can be useful for supervisory institutions.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Griškevičiūtė, Agnė,
- Contributors dc:contributor
-
- Markevičiūtė, Jurgita
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD81591376&prefLang=en_US