What drives the profitability of banking sectors in the European Union? The machine learning approach

被引:0
作者
Bernardelli, Michal [2 ]
Korzeb, Zbigniew [1 ]
Niedziolka, Pawel [3 ]
机构
[1] Bialystok Tech Univ, Dept Management Econ & Finance, Kleosin, Poland
[2] SGH Warsaw Sch Econ, Coll Econ Anal, Warsaw, Poland
[3] SGH Warsaw Sch Econ, Coll Socio Econ, Warsaw, Poland
关键词
banking sector; machine learning; profitability; random forest; SHAP; G01; G21; G30; Z10; DETERMINANTS;
D O I
10.2478/ijme-2024-0022
中图分类号
F [经济];
学科分类号
02 ;
摘要
The study aims to establish patterns of relations between the profitability of the European Union (EU) banking sectors between 2007 and 2021 and sets of variables appropriate for clusters of countries into which the 27 countries of the EU are divided. The random forest method is deployed to identify the factors influencing the value of the return on equity. Shapley additive explanations are exploited to add interpretability to the results. The results show that the sets of variables shaping the profitability of banking sectors in the EU grouped by use of sovereign rating criterion are different. However, there are variables common to all banking sectors. These include cost efficiency and default risk. The study's novelty lies in the reliance on a broad spectrum of explanatory variables assigned to three groups of factors, reference to all EU countries, and decomposition of the sample to identify similarities among the determinants of profitability.
引用
收藏
页码:272 / 284
页数:13
相关论文
共 48 条