Leverage and Cash Dynamics
Harry DeAngelo, Andrei S Gonçalves, René M Stulz
Review of Finance, Volume 26, Issue 5, September 2022, Pages 1101–1144, https://doi.org/10.1093/rof/rfac043
The empirical corporate finance literature contains two distinct (and enormous) research silos, one focused on capital structure and the other on cash balances, with interactions between leverage and cash largely ignored. This paper reports a broad set of new and empirically important dynamic interactions between leverage (both market and book leverage) and cash-balance ratios (i.e., cash as a fraction of total assets). The evidence indicates a need to treat debt and cash policies as jointly interdependent through the roles they play in corporate funding.
Time-series variation in leverage depends strongly on cash balances, and vice-versa for time-series variation in cash, with leverage and cash ratios interacting approximately as predicted by the internal-versus-external funding regimes in Myers and Majluf (1984). Market and book leverage ratios (defined as debt, divided either by total market value or total assets) are quite volatile when cash-balance ratios are stable and vice-versa, while net-debt ratios (defined as debt minus cash, divided by total assets) are almost always volatile. Most firms increase leverage sharply as cash balances (internal funds) become scarce, despite widespread equity issuances that violate the pecking-order rule for external financing.
We study episodes in which firms move from their historically highest Cash/TA ratio to a later trough, a transition that takes a median of three years in our full sample of Compustat firms. Median Cash/TA declines from 0.298 at the peak to 0.044 at the trough, while median market leverage increases from 0.056 to 0.206. Absent external financing, most of these firms would have run out of cash by the trough year. The median firm raises funds equal to 290% of trough-year cash balances, which means that most of the new funds are used to cover outlays that internal funds cannot cover.
This connection between changes in leverage and cash-balance squeezes (defined as movements from peak Cash/TA to trough) stands out in bold relief when we divide our sample into firms that would have run out of cash absent external financing and firms that did not need outside funds to have positive cash balances. Median market leverage increases markedly from 0.066 to 0.283 for the former group and only slightly from 0.033 to 0.046 for the latter group. In short, the firms that otherwise would have run out of cash tend to increase leverage by large amounts while the remaining firms do not. We also find that cash-balance squeezes are pervasively important when we reverse the sampling approach and analyze levering-up episodes in which a firm moves from its historically lowest leverage to subsequent peak leverage.
We gauge the ability of the models of Gamba and Triantis (2008) and DeAngelo, DeAngelo, and Whited (2011) to explain the within-firm relation between cash-balance squeezes and time-series variation in leverage. We find that these models explain the qualitative link between leverage increases and cash squeezes at real-world firms, but both fall short quantitatively by predicting squeeze-related increases in market-leverage ratios that are far more muted than we observe in the real data.