The U.S. leveraged (high-yield) loan market has more than doubled since the Great Financial Crisis (GFC), with nearly $1.2 trillion in outstanding debt in 2019 (Leveraged Commentary and Analysis, LCD). The rise in high-yield corporate debt (bonds included) in the decade following the GFC has prompted concern in the U.S. and Europe, with central bankers and other policy makers frequently pointing to its potential to amplify any economic shock.
The problem with high corporate leverage is that it can lead to financial insolvency, which can trigger a contraction in demand like what we saw in 2020. Firms can be left with an oversized debt burden, whose consequences can in various ways deepen the impact of the initial shock. Several papers have pointed to the importance of loan financial-covenant violations in amplifying the 2008-09 economic downturn (e.g., Chodorow-Reich and Falato, 2020). These papers argue that a shift in control rights to creditors was critical in spreading lenders’ distress to the non-financial sector.
The current high-yield debt environment, however, is different, because the financial covenant structure has changed significantly since the GFC. Historically, loans have had maintenance covenants that require a borrower’s continuous compliance with the covenants’ thresholds (e.g., covenants prohibit the leverage ratio from exceeding a certain threshold). Failure to comply constitutes an event of default, effectively shifting control rights to lenders. In contrast, high-yield debt is characterized by incurrence covenants that restrict some actions of the borrower, such as distributions to shareholders or investments, after the covenant threshold is crossed. Such covenants earned high-yield loans the moniker “cov-lite,” the share of which in the leveraged loan market has increased from just above 10 percent in 2007 to over 80 percent in 2020.
Despite the prevalence of incurrence covenants in high-yield debt, there is little empirical evidence showing how they affect borrowers. In a new study, we show that “latent” violations of incurrence financial covenants in the leveraged loan market – that is, a trigger of restricted actions in the loan contact – have sizable effects long before any default or bankruptcy. Overall, our evidence supports the significance of transmission of economic shocks through their impact on corporate balance sheets, but our insight is different from the previous studies. We isolate a novel mechanism relevant for understanding the consequences of high corporate leverage. Importantly, we speak to the current debt environment, which is dominated by high-yield loans with provisions that are more characteristic of high-yield bonds.
Central to the mechanism is that incurrence-covenant triggers activate a set of contractual constraints on a firm’s actions. As we show, these triggers have a strong effect on a firm’s investment policy. While not all restricted actions directly limit investment, they tend to be costly for equity holders and, as a result, they indirectly influence the firm’s capital expenditures. For example, consider a borrower that exceeds a cap on leverage (Net Debt/EBITDA), which is, by far, the most common financial covenant in leveraged loans. Therefore, to lift the restrictions, the borrower has to lower Net Debt/EBITDA, perhaps by cutting costs to boost EBITDA. Another approach is to reduce Net Debt by selling some assets or constraining capital expenditures that require financing.
Our empirical analysis of these effects exploits novel, hand-collected, loan-level data on covenant information and builds on empirical methodology in Chava and Roberts (2008). Our key empirical results are as follows:
- The investment rate drops about 1.8 percentage points (pp) in the case of a trigger of an incurrence covenant, as compared with 0.9 pp in case of a violation of a maintenance covenant.
- After triggering either a maintenance or an incurrence covenant, firms significantly deleverage. In our sample, we find that the debt-to-assets ratio decreases by about 1.6 pp when the firm violates a maintenance covenant. However, we also find that the trigger of an incurrence covenant leads to a reduction in the debt-to-assets ratio of about 2.7 pp.
- Triggers of restricted actions under the incurrence covenants and maintenance covenant violations both lead to a significant decline in equity values (about 6 percent), consistent with the role of covenants preserving creditor’s value. Moreover, the timing of these effects is as sudden as that of the maintenance covenants, pointing to the fact that the propagation of the economic shocks in an economy with highly levered corporate sector is quick.
Overall, the economic impact of the restricted actions tied to incurrence covenants appears to be more severe than the economic impact of maintenance covenant violations, after controlling for firm characteristics to address potential selection across the two groups. A plausible explanation is that a firm needs to be in a worse condition to trigger a maintenance covenant (as compared with incurrence covenants) because, as we show, the maintenance covenant thresholds are substantially more forgiving. Another aspect to consider is that, whereas maintenance covenants are typically used in loans funded and held by banks, incurrence covenants characterize loans funded by institutional investors – collateralized debt obligations, for example – and mutual funds, a group for which coordination cost among creditors is much higher. It is plausible that restricted actions, being a blunt contractual tool intended to protect creditors, end up being more restrictive to borrowers.
Overall, the mechanism identified in our study is essential to understanding the propagation of demand shocks, such as the 2020 pandemic, in a highly leveraged corporate sector and is independent of whether the firms eventually file for bankruptcy.
This post comes to us from Falk Bräuning at the Federal Reserve Bank of Boston, Victoria Ivashina at Harvard University and the National Bureau of Economic Research, and Ali K. Ozdagli at the Federal Reserve Bank of Dallas. It is based on their recent paper, “High-Yield Debt Covenants and Their Real Effects,” available here. The views expressed in this post do not necessarily reflect those of the Federal Reserve Banks of Boston or Dallas, the Federal Open Market Committee, or the Federal Reserve System.