Determinants of Capital Structure Revisited

Abstract

We examine the determinants of capital structure for S&P 500 firms over the recent 25 years from 2000 to 2024. Using both traditional regression models and machine-learning approaches, we investigate how firm fundamentals and macroeconomic conditions jointly shape firms’ leverage decisions. Across several empirical methods, including OLS, Lasso, Ridge, Random Forest, and Gradient Boosting, we find that operating profitability, tax-shield benefits, and earnings volatility consistently emerge as the most important determinants of leverage. Firms with higher profitability rely less on external financing, while tax incentives or greater internal cash flow uncertainty can result in the use of more debt. Nonlinear models further highlight liquidity as an important predictor, suggesting that internal cashflow buffers reduce reliance on external financing. At the macro level, GDP growth is negatively associated with leverage, consistent with firms deleveraging during economic expansions. Overall, our findings suggest that firm-specific characteristics play a dominant role in shaping capital structure decisions, while macroeconomic conditions exert a secondary but meaningful influence.