Abstract
Yes, they can! Machine learning models predict leverage better than linear models and identify a broader set of leverage determinants. They boost the out-of-sample R2 from 36% to 56% over OLS and LASSO. The best performing model (random forests) selects market-to-book, industry median leverage, cash and equivalents, Z-Score, profitability, stock returns, and firm size as reliable predictors of market leverage. More precise target estimation yields a 10%–33% faster speed of adjustment and improves prediction of financing actions relative to linear models. Machine learning identifies uncertainty, cash flow, and macroeconomic considerations among primary drivers of leverage adjustments.
•Machine learning models predict target leverage better than linear models.•Out-of-sample R-squared increases from 36% to 56% over OLS and LASSO.•Machine learning identifies a broader set of leverage determinants.•More precise target estimation yields 10%–33% faster speed of leverage adjustment.•Machine improves prediction of financing actions in debt and equity markets.