With zero capital structure rebalancing costs, dynamic trade-off theory predicts that firms stay at their leverage targets with more profitable firms staying at higher leverage. This prediction is rejected by the robustly negative correlation between leverage and profitability. When rebalancing costs are added to this theory, it predicts a positive leverage–profitability correlation only in periods where companies pay these costs and actively rebalance their capital structures. However, we show that the correlation is negative when firms issue debt and distribute the proceeds to shareholders—precisely the case where the theory predicts it should be positive. Our results thus resurrect the leverage–profitability puzzle.

We test whether high-frequency net-debt issuers (HFIs)—public industrial companies with relatively low issuance costs and high debt-financing benefits—manage leverage toward long-run targets. Our answer is they do not: (1) the leverage–profitability correlation is negative even in quarters with leverage rebalancing; (2) the speed-of-adjustment to target leverage deviations is no higher for HFIs than for low-frequency net-debt issuers; and (3) under-leveraged HFIs do not speed up rebalancing activity in significant investment periods. Thus, even in the subset of firms most likely to follow dynamic trade-off theory, the theory does not appear to hold.