The Impact of Environmental Performance on Corporate Cash Holdings: Evidence from Financial Institutions

Abstract

Corporate liquidity management increasingly intersects with environmental performance in financial institutions, where cash-holding decisions must reconcile sustainability commitments, prudential liquidity requirements, and climate-related financial risk. Existing evidence remains limited for emerging-market financial institutions and provides little insight into whether the Environmental Pillar Score is associated with cash-holding behavior after accounting for dynamic adjustment, endogeneity, and macroeconomic conditions. This study examines the association between environmental performance and corporate cash holdings in listed South African financial institutions. The analysis uses a balanced panel of 18 banks and insurance firms observed over 2015–2024, comprising 180 firm-year observations in the full panel, of which 162 enter the dynamic estimation; firm-level financial and environmental data are obtained from Thomson Reuters Eikon and macroeconomic data from World Bank Open Data. The empirical strategy combines descriptive statistics, correlation analysis, panel diagnostics, and a two-step dynamic system generalized method of moments estimator with Windmeijer-corrected robust standard errors in Stata version 16.1, using 13 instruments and a collapsed instrument matrix. The dynamic estimates show a positive and statistically significant association between the Environmental Pillar Score and cash holdings (β = 0.000420, standard error = 0.000107, p < 0.01), while lagged cash holdings are negative and significant (β = −0.387, standard error = 0.0214, p < 0.01), indicating mean-reverting dynamics; Arellano–Bond first-order and second-order p-values are 0.444 and 0.303, while Sargan and Hansen p-values are 0.082 and 0.992. Cash flow and firm size are positively associated with cash holdings (β = 0.0888, standard error = 0.0162; β = 0.132, standard error = 0.0220; both p < 0.01), whereas the current ratio, leverage, and growth opportunities are negatively associated (β = −0.00225, standard error = 0.0000625; β = −0.157, standard error = 0.0131; β = −0.0328, standard error = 0.00388; all p < 0.01); the lending interest rate is positive and significant (β = 0.00443, standard error = 0.000557, p < 0.01), while real gross domestic product and the Consumer Price Index are non-significant. The Environmental Pillar Score averages 45.14, ranges from 0.00 to 96.44, and is positively correlated with cash holdings (r = 0.2372, p < 0.01); however, the fixed-effects coefficient is negative and non-significant (β = −0.0000659), showing that the sign of the environmental-performance coefficient is not stable across estimators. The findings extend sustainable finance and corporate liquidity research by quantifying this relationship in an emerging financial-sector setting, highlighting estimator sensitivity and supporting further research on climate-related financial risk, liquidity management, and financial resilience without inferring causality from observational data.