Bank-Level Financial Panel: Profitability of Individual Ukrainian Banks, 2016-2025

This dataset provides a bank-level (not system-aggregate) panel of financial performance for individual Ukrainian banks from 2016 to 2025: total assets, total capital, net profit, income components, and both officially reported and independently calculated return on assets (ROA) and return on equity (ROE), for every bank licensed in Ukraine in each year, sourced directly from NBU individual bank financial reporting. The panel contains 728 bank-year observations across 60-113 banks per year (the count falls over time due to the sector consolidation documented in this author's companion dataset on bank count). This is intended primarily to support panel regression analysis of bank profitability determinants, of the kind used in the literature on post-Soviet and transition-economy banking (e.g. Yuksel et al., 2018, MDPI Economies), where firm/bank-level panels give far more statistical power than aggregate annual time series. A pooled OLS regression of bank-level ROA on the NBU discount rate and the system-wide government bond (OVDP) share of assets, using this dataset, yields a statistically significant result (n=633, F-test p=0.0003) that is not obtainable from the 10-observation annual aggregate series alone - notably, the bank-level discount-rate coefficient is negative, in contrast to the positive association seen in system-aggregate data, suggesting the aggregate "profitability paradox" during the war may be concentrated among a subset of large banks with substantial government bond holdings rather than reflecting the typical bank. CONTACT INFORMATION Illia Kyselov, PhD in Law Associate Professor, Department of Law, Bohdan Khmelnytsky Melitopol State Pedagogical University, Zaporizhzhia, Ukraine CMSA (R), FPWMP (R), and BIDA (R) (Corporate Finance Institute) Email: [e-mail] The author is open to collaboration on Scopus-indexed publications and further research using Ukrainian economic and financial statistics. Inquiries regarding this dataset, potential co-authorship, or data requests are welcome.本数据集提供了2016年至2025年间乌克兰各单体银行的银行层面（而非系统整体层面）财务表现面板数据，覆盖乌克兰每一家持牌银行每年的相关指标，包括总资产、总资本、净利润、收入构成，以及官方报告与独立测算的总资产收益率（Return on Assets, ROA）和净资产收益率（Return on Equity, ROE），数据直接源自乌克兰国家银行（National Bank of Ukraine, NBU）的个体银行财务报告。 该面板数据集共包含728组银行-年度观测样本，每年覆盖60至113家银行（样本数量随时间递减，原因在于本文作者关于银行数量的配套数据集所记录的银行业整合进程）。 本数据集主要用于开展银行盈利能力影响因素的面板回归分析，这类分析方法常见于后苏联与转型经济体银行业的相关研究（例如Yuksel等人2018年发表于MDPI期刊《Economies》的论文），相较于年度总体时间序列，企业/银行层面的面板数据能提供更强的统计效能。 利用本数据集开展的针对银行层面总资产收益率的混合普通最小二乘回归，将乌克兰国家银行贴现率与全系统政府债券（OVDP）占资产比例作为解释变量，得到了统计显著的结果（样本量n=633，F检验p值=0.0003），而仅通过10组观测值的年度总体时间序列无法得到该结果。值得注意的是，银行层面的贴现率系数为负值，与系统整体数据中观测到的正相关关系形成鲜明对比，这表明战争期间的总体"盈利能力悖论"可能仅集中于少数持有大量政府债券的大型银行，而非代表行业内多数普通银行的经营状况。 联系方式 伊利亚·基塞洛夫，法学博士 扎波罗热州鲍哈丹·赫梅利尼茨基梅利托波尔国立师范大学法学院副教授 持有企业金融研究所（Corporate Finance Institute）认证的CMSA(R)、FPWMP(R)与BIDA(R)资质 邮箱：[e-mail] 作者欢迎针对斯高帕斯（Scopus）收录学术论文的合作，以及使用乌克兰经济与金融统计数据开展的后续研究。垂询本数据集相关问题、潜在合著机会或数据索取请求均欢迎。