A dataset of European banks in performance evaluation under uncertainty

The dataset contains financial indicators from the financial statements of 250 banks operating in Europe which are collated for the 2015 accounting year. First, the dataset is split into input and outputs measures. Then the preferred number of inputs and outputs in relation to the total number of data is selected according to the rule of thumb in data envelopment analysis (DEA). The dataset is related to the research article entitled “Robust optimization with nonnegative decision variables: A DEA approach” (Toloo and Mensah, 2018) [1]. The dataset can be used to evaluate the performance of banks and bank efficiency under uncertainty.


a b s t r a c t
The dataset contains financial indicators from the financial statements of 250 banks operating in Europe which are collated for the 2015 accounting year. First, the dataset is split into input and outputs measures. Then the preferred number of inputs and outputs in relation to the total number of data is selected according to the rule of thumb in data envelopment analysis (DEA). The dataset is related to the research article entitled "Robust optimization with nonnegative decision variables: A DEA approach" (Toloo and Mensah, 2018) [1]. The dataset can be used to evaluate the performance of banks and bank efficiency under uncertainty.
& 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Subject area
Operations research and management science More specific subject area Data envelopment analysis Type of data .

Value of the data
The raw data contains key financial statements indicators of 250 banks in Europe which were taken from the individual bank's financial statement in 2015.
The data are arranged in order of the largest bank to the smallest bank in terms of assets The data is useful for measuring the performance of banks in Europe and for comparative analysis of sub-regional performances and beyond.
The data can be used by researchers to evaluate a wide range of efficiency measures for the countries under consideration.

Data
The data comprises financial indicators in the financial statements of 250 public and private banks operating in Europe. Table 1 shows the distribution of these banks according to the sub-region. Including data on indicators such as assets, employees, personnel expenses, equity, loans, net interest income, deposit from banks, operating income and net fees and commission, the detailed financial statements of the banks were obtained from the Bureau van Dick -Bankscope database for the 2015 accounting year. The summary of descriptive statistics of these indicators for each subregion is provided in Tables 2-5. All the financial indicators were measured in millions of Euros with the exception of employees which is measured in actual figures. The total number of employees is defined as the number of banking professionals and the non-banking staff is given employed in the accounting year.

Experimental design, materials and methods
Financial statements of banks were first downloaded from the Bankscope database [2]. Then data on the financial indicators mentioned above were compiled from 250 banks financial statements individually and collated. These banks are arranged in descending order of their assets size. Subsequently, for the performance analysis of the banks using the data envelopment analysis (DEA) tool, the financial indicators are split into two samples. The first is the input measures and the second is the output measures. The separation of the financial indicators into inputs and outputs measures was done based on the selective measures described in [3, , 4]. The approach adopted for selecting inputs and outputs is the intermediary approach of banking studies, which is shown in Tables 2-4. With the exception of deposit and loans refereed mostly in literature as dual role factors, it is unarguable the selection of the measures as input and inputs. In this paper, the selection of deposit specifically as output corresponds  All the raw data are scaled for uniformity and to reduce round-off errors from excessively large values prior to analysis.