Factors Determining the Financial Performance of Public Sector Banks in India

The study addresses how bank-specific and macroeconomic factors influence the profitability of Indian public sector banks (PSBs). Against the backdrop of India’s banking reforms since the 1990s and recent shocks (COVID-19, Russia–Ukraine conflict) that pressured banks’ profitability and capital, the paper aims to identify internal (bank-level) and external (macroeconomic) determinants that drive ROA, ROE, and NIM. Evaluating these determinants is important for regulatory supervision and for distinguishing stronger from weaker banks in a competitive, technology-driven environment. The research focuses on PSBs over 2010-11 to 2021-22 to provide updated evidence for policy and managerial decision-making.

The paper reviews extensive global and country-specific evidence on bank profitability determinants. International studies (e.g., Short, Bourke, Demirgüç-Kunt & Huizinga; Athanasoglou et al.; European panels) highlight roles of capitalization, efficiency, asset quality, and macro conditions (GDP, inflation). Single-country studies from the US, Europe, and Asia report mixed effects for size, capital adequacy, operating efficiency, asset quality, liquidity, and macro variables. Indian evidence shows profitability linked to interest cost, deposit/credit structure, non-interest income, spread, operating expenses, NPAs, capital-to-assets, diversification, and macro factors. Some studies find negative impacts of risk and inflation; others note positive effects of bank size and GDP depending on context. The review motivates hypotheses on expected signs: size (ambiguous), CAR (+), CTI (−), NPA (−), credit risk (−), CDR (+), GDP (+), inflation (−).

Design: Quantitative panel analysis of 12 Indian public sector banks selected by highest market capitalization in 2021-22. Period: FY 2010-11 to 2021-22 (12 years). Data sources: Moneycontrol.com, Reserve Bank of India reports, Indian Banks’ Association publications, academic journals, working papers, newspapers, and World Bank data for macro variables. Variables: - Dependent (profitability): ROA (Net Income/Total Assets), ROE (Net Income/Shareholders’ Equity), NIM (Net Interest Income/Total Assets). - Internal (bank-specific): Size (ln Total Assets), CAR (Total Equity/Total Assets), CTI (Total Cost/Total Income), NPA (Net NPAs/Advances or to Total Assets as described), CrR (Loan loss provisions/Total Assets), CDR (Total Advances/Total Deposits). - External (macroeconomic): GDP growth rate (%), Inflation (average annual CPI growth rate). Model and tools: Multiple Linear Regression for each profitability measure separately: PRO = β0 + β1(Size) + β2(CAR) + β3(CTI) + β4(NPA) + β5(CrR) + β6(CDR) + β7(GDP) + β8(Infl) + ε, where PRO ∈ {ROA, ROE, NIM}. Statistical analyses include descriptive statistics, correlation matrix, multicollinearity diagnosis, t-tests, F-tests, ANOVA, using SPSS v28. Reported model fit includes R, R², Adjusted R², F-statistic, and p-value.

Descriptive statistics (selected): ROA mean 1.16% (SD 8.59%); ROE mean −5.49% (SD 158.93%); NIM mean 16.74% (SD 29.47%). Size mean (index) 153.67; CAR mean 148.06; CTI mean 447.50; NPA mean 53.40; CrR mean 3.67; CDR mean 842.53; GDP mean 4.63%; Inflation mean 6.52%. High coefficients of variation indicate considerable variability. Correlation highlights: Size, CTI, and NPA are significantly negatively correlated with ROA, ROE, and NIM (e.g., Size with ROA −0.73, with ROE −0.83, with NIM −0.82). CDR shows positive correlation with profitability measures (e.g., with ROA 0.67), while CAR, CrR, GDP, and Inflation exhibit positive but often statistically insignificant correlations (except some for CDR and CrR). Model fit (ANOVA/model summary): - ROA: R=0.969, R²=0.939, Adj R²=0.697, F=3.881, p=0.221 (overall model not statistically significant). - ROE: R=0.994, R²=0.987, Adj R²=0.936, F=19.306, p=0.05 (significant at 5%). - NIM: R=0.995, R²=0.990, Adj R²=0.952, F=25.648, p=0.038 (significant at 5%). Regression coefficients (signs): - Size: negative for ROA (−0.285), ROE (−0.186), NIM (−0.300); not significant. - CAR: negative for ROA (−0.742) and NIM (−0.119), positive for ROE (+0.328); not significant. - CTI: negative for ROA (−0.087) and ROE (−0.050); positive for NIM (+0.019); not significant. - NPA: negative for ROA (−1.493), ROE (−0.575), NIM (−0.941); not significant. - CrR: positive for ROA (+0.521), ROE (+0.128), NIM (+0.448); not significant. - CDR: negative for ROA (−0.998) and NIM (−0.351), positive for ROE (+0.327); weak/insignificant. - GDP: positive for ROA (+0.227), ROE (+0.056), NIM (+0.094); not significant. - Inflation: negative for ROA (−0.072), ROE (−0.313), NIM (−0.164); not significant. Summary of hypotheses vs. results: Many expected signs hold (e.g., negative effects of CTI and NPA; positive ROE–CAR), but most coefficients are statistically insignificant. Overall factor sets significantly explain ROE and NIM (5% level), not ROA. The sector benefits weakly from GDP growth and inflation.

The analysis indicates that internal efficiency and asset quality (CTI, NPA) are aligned with theoretical expectations in their negative association with profitability, underscoring the importance of operating efficiency and credit risk management for PSBs. Larger size correlates negatively with profitability measures, suggesting possible diseconomies of scale or structural inefficiencies within PSBs over the study period. CAR’s mixed effect—negative for ROA/NIM yet positive for ROE—suggests capitalization may enhance returns to equity while compressing margins and asset returns, possibly through reduced leverage-driven spreads. Macroeconomic variables show broadly positive (GDP) or weakly beneficial (inflation) relationships with profitability, but lack robust statistical significance at the coefficient level. Despite high R² for ROE and NIM models and significant F-tests, individual predictors often lack significance, which may reflect multicollinearity, sample size constraints, or heterogeneity across banks. Policy and managerial implications include prioritizing NPA reduction and cost optimization, cautious growth strategies to avoid scale diseconomies, and prudent management of credit and liquidity risks to stabilize profitability across cycles.

The study concludes that PSBs’ profitability is primarily shaped by internal factors—especially operating efficiency and asset quality—while capitalization and macro conditions exert mixed, generally weaker effects. ROE and NIM are well explained by the combined determinants, whereas ROA is not statistically well-explained in the model. Managerially, banks should emphasize operational efficiency, rigorous credit appraisal and recovery to reduce NPAs, and careful balance-sheet management to optimize margins without incurring excessive risk. Policy suggestions include stronger supervision of credit and liquidity risks and fostering competition. Future research should extend the time horizon, consider broader samples and comparative groups, and further address potential multicollinearity and specification issues.

The study focuses on 12 Indian public sector banks over 2010-11 to 2021-22 and relies on secondary data, which may limit generalizability. Many coefficient estimates are statistically insignificant (despite high model R² for ROE and NIM), and the ROA model lacks overall significance, potentially reflecting sample size, variable measurement differences, or multicollinearity. The single-country, PSB-only scope and the study period may constrain external validity; the authors suggest extending the period and splitting samples in future work.