How Covenant-Lite Debt Waves Reveal Late-Cycle Credit Excess

What does it mean when the lender stops asking questions? That's the plain-English version of a covenant-lite loan — a credit agreement that drops the maintenance covenants (the quarterly leverage and coverage tests that let a lender step in before trouble compounds) and keeps only incurrence covenants, which trigger solely when a borrower tries to do something new, like raise more debt. The distinction sounds technical. It isn't.

I've watched enough cycles turn to trust one signal more than most headline credit-spread numbers: how much protection lenders are willing to give up to win the deal. The covenant-lite share of new leveraged loan issuance is that signal in close to its purest form, and unlike a lot of credit-market data, you can actually go verify it yourself — it's disclosed loan by loan, in the credit agreement exhibits companies file alongside their 8-Ks.

It's a leading indicator, not a coincident one. By the time defaults show up in the headline numbers, the covenant erosion that made those defaults harder to see coming — and costlier to recover from — happened years earlier, one credit agreement at a time.

The Covenant Package Is a Credit-Cycle Instrument, Not a Legal Footnote

Maintenance covenants force a conversation. If a borrower's leverage ratio drifts past an agreed threshold at quarter-end, the lender gets a seat at the table — a waiver fee, a repricing, sometimes board-level involvement — well before the business is actually insolvent. Incurrence covenants do none of that. They sit dormant unless the borrower voluntarily takes an action the agreement restricts, which means a company can deteriorate steadily, quarter after quarter, without ever tripping a wire.

Lenders don't give that protection away for free. They give it away when competition for the deal is fierce enough that demanding it means losing the mandate to somebody else. By 2021, covenant-lite loans accounted for roughly 90% of new institutional leveraged loan volume, according to S&P Global Market Intelligence's LCD — up from about a quarter of the market back in 2007, the last time this same dynamic played out at scale. Same mechanism, same cast of characters: private credit funds and CLOs flush with capital, bidding for a limited supply of syndicated paper, competing away the protections that used to be table stakes.

None of this is really about the covenants themselves. It's about who holds the pricing power in the lending relationship — and late in a credit cycle, it's almost never the lender. That's worth sitting with, because the leveraged loan market sets terms that eventually filter down to the high-yield bond market and, with a lag, to bank underwriting standards more broadly.

What the Last Two Peaks Looked Like

The historical analogue here is imperfect, but it rhymes closely enough to be useful. The first modern covenant-lite wave built through 2005 and 2006 and crested in 2007, right before the financial crisis froze the leveraged loan market almost entirely. Covenant-lite issuance didn't just slow in 2008 and 2009 — it nearly vanished, because the CLO bid that had been funding it disappeared overnight and nobody was left to underwrite loose paper. Conditions flipped from lender-indifferent to lender-hostile in about eighteen months.

The second wave took longer to build and went further. Starting around 2013, covenant-lite share climbed steadily through the rest of the decade, and by 2018 it had become the market standard rather than the aggressive exception. Along the way, one deal turned the abstraction into a widely cited cautionary tale. In 2016, J.Crew Group used loopholes in its own credit agreement to move roughly $250 million of trademark intellectual property into an unrestricted subsidiary — beyond the reach of the term-loan lenders whose collateral package was supposed to include it. Lenders who'd waived away the protections that would have blocked the move found out, after the fact, what covenant-lite actually costs in a downside scenario.

What's notable is what didn't happen next. You'd expect that episode to have chastened the market — lenders demanding tighter language, covenant-lite share retreating. It didn't, not durably. Credit agreements did pick up so-called "J.Crew blocker" provisions after that, closing that one specific loophole. But the broader covenant-lite share of issuance kept climbing anyway, because the underlying imbalance — too much capital chasing too few deals — never went away. A single cautionary tale doesn't reverse a structural bid, and a related episode a couple of years later (PetSmart's 2018 transfer of a stake in Chewy to a similarly insulated affiliate) showed the playbook had spread well beyond one retailer's law firm.

A Framework for Reading Covenant-Lite Waves in Real Time

Treating covenant-lite share as a single headline number is where most people stop, and it's also where most of the signal gets lost. Here's the practical checklist:

1. Track the covenant-lite share of new issuance, not just the level — the rate of change matters more than any absolute threshold, since 70% and rising late-cycle behaves very differently from 70% and stable mid-cycle.
2. Identify who's buying the paper. CLOs and private credit funds bidding for yield compress covenant protection faster than bank balance sheets, which still retain some hold-to-maturity discipline; a wave dominated by non-bank buyers is a faster-moving wave.
3. Cross-reference EBITDA addback aggressiveness in the same credit agreement exhibits — inflated pro forma EBITDA and loosened covenants tend to move together, and the combination compounds the eventual recovery-rate problem.
4. Check where the wave sits relative to the broader rate backdrop. A covenant-lite wave that builds while the [yield curve](/blog/yield-curve-recession-signal) is still upward-sloping behaves differently than one still expanding after the curve has already inverted.
5. Respect the lag. Covenant-lite issuance historically peaks something like twelve to twenty-four months before defaults meaningfully rise, which is exactly why it's useful as an early read rather than a confirming one.

None of these steps requires anything more exotic than reading what's already public. The credit agreement exhibits are sitting in the same 8-K filings companies file for other reasons — most investors just never open them. If you want the underlying mechanics of what a loan agreement is actually promising, our explainer on what a bond is covers the fixed-income vocabulary that carries over almost directly to the leveraged loan market, even though loans and bonds sit in different parts of the capital structure.

What Covenant-Lite Levels Don't Tell You

A high covenant-lite share doesn't tell you when the cycle turns. It can stay elevated for years — as it did through most of the 2010s — without a default wave materializing, because low rates and open capital markets let even weak borrowers refinance their way out of trouble. The signal tells you that protection has eroded, not that the trigger is imminent. Waiting for it to spike as a market-timing tool will leave you waiting a very long time.

It also doesn't distinguish sponsor quality. A covenant-lite loan backing a durable, cash-generative business is a very different risk than the identical covenant package backing a leveraged rollup with no pricing power of its own. The framework also gets noticeably weaker once you leave the syndicated market — direct lending through private credit funds discloses far less, because so much of it is bilaterally negotiated rather than filed anywhere, and I'm less confident extending this checklist cleanly into that corner of the market. The data there is thin, and it happens to be the fastest-growing part of the leveraged finance world, which is its own kind of warning about where the next disclosure gap will matter most. Note the difference in altitude, too: this is an aggregate, market-wide read, not a single-company one. For the security-by-security version — what a compliance footnote in one company's own credit agreement is actually telling you — that's how to read a debt covenant compliance footnote.

Where This Cycle's Wave Diverges

Entering 2026, covenant-lite share of outstanding institutional loans has stayed parked near the highs the market set earlier in the decade rather than retreating meaningfully — a plateau, not the sharp reversal you'd see after a real repricing of credit risk. That matters for how you read the current regime. A credit market still willing to extend loose terms is, functionally, a market pricing a compressed equity risk premium too; the two move together because they're both downstream of the same appetite for risk, and both eventually run through the same discount rate that gets applied to future cash flows.

Worth being precise about which rate does the work here. Nominal yields on leveraged loans have moved around plenty with the policy rate, since most of this paper floats. Real financing costs — nominal yield net of inflation — are the number that actually governs whether a borrower's cash flow can service the debt over time, and they're the number covenant terms are, in effect, insuring against. A borrower can look fine on a nominal-coverage basis while its real debt-service burden is quietly rising if inflation cools faster than the loan resets. That's a subtler version of the same story: loose covenants remove the mechanism that would otherwise force an early conversation about it.

There's an earnings-channel version of this too, separate from the valuation-multiple story. Loose covenants make debt-funded acquisitions and buybacks cheaper to execute and easier to justify to a board, which shows up later as margin pressure once the acquired growth needs servicing at a less forgiving rate — a dynamic distinct from organic earnings growth even though it can look identical on a headline EPS line for a few years. Distinguishing the two matters more, not less, in a stretch where issuance has stayed loose this long.

But the plateau shouldn't be read as complacency being vindicated. Loose covenants don't cause a credit cycle to turn — they just determine how much room lenders have to react once something else does. Rate policy, growth data, an earnings shock at a large enough borrower: any of those can still be the actual trigger. Covenant-lite share just tells you how thin the cushion will be when it happens, and how much of that cushion has already been spent chasing yield in a financially repressive stretch for savers — a topic worth its own look at financial repression in its own right, not just a footnote to this one.

Key Takeaways

• Covenant-lite share of leveraged loan issuance is a real-time, publicly disclosed gauge of how much competitive pressure lenders are under — and it has tracked the same pattern in both the 2007 and 2018 cycle peaks.
• The signal works best as a rate-of-change and cycle indicator, not a timing tool; it historically leads defaults by roughly twelve to twenty-four months and can stay elevated for years without triggering anything.
• J.Crew's 2016 collateral maneuver, and PetSmart's similar move two years later, are the clearest illustrations of what loose covenants cost in a downside case — and the market's failure to durably reprice after them shows how sticky a structural capital imbalance can be.
• Credit conditions this loose usually map onto a compressed equity risk premium elsewhere in the market — the same channel that connects a leveraged-loan footnote to what you should expect from equity valuations. For how that channel has behaved across Fed tightening and easing cycles specifically, see our look at [Fed policy and sector leadership](/blog/fed-policy-and-sector-leadership).