The Underwriting Question Every Lender Must Answer
Imagine you have just closed a $45 million senior secured term loan to a mid-market specialty manufacturer. EBITDA at close is $10 million. Leverage is 4.5x. Coverage looks comfortable at 2.1x. Six months later, raw material costs spike, a key customer delays orders, and EBITDA quietly slides to $7.2 million. Leverage has drifted to 6.25x. Coverage has fallen to 1.4x. And without a properly designed maintenance covenant in place — you may not find out until the borrower misses an interest payment.
This is the foundational problem that covenant design solves. Covenants are not administrative formalities. They are early-warning systems, negotiating levers, and structural protections that separate recoverable credit stress from undetected default. Getting them right — at origination — is one of the most consequential decisions in leveraged lending underwriting.
The rise of liability management exercises (LMEs) — where borrowers exploit loose documentation to pit lenders against each other, as seen in the J. Crew, Pluralsight, and Serta transactions — has made structuring expertise more critical than ever. According to Oak Hill Advisors (OHA), published February 2026, documentation and structuring expertise at the time of underwriting are key to protecting lenders against harmful borrower actions in an increasingly complex LME environment.
The Analytical Framework: Two Covenant Types, One Unified Purpose
Institutional credit underwriters draw a hard distinction between two covenant architectures. Understanding both — and when each applies — is foundational to any rigorous deal structure.
1. Maintenance Covenants
Maintenance covenants require the borrower to continuously satisfy a financial test, typically tested quarterly. They are the cornerstone of traditional bank lending and provide lenders with a proactive right to intervene before a default becomes a loss event. The most common maintenance covenants include:
- Maximum Leverage Ratio (Debt/EBITDA): Caps total net debt as a multiple of trailing twelve-month (TTM) EBITDA. Example: must not exceed 5.25x at any quarter-end.
- Minimum Fixed Charge Coverage Ratio (FCCR): Requires that EBITDA minus CapEx minus taxes covers all fixed charges (interest plus mandatory amortization) by at least 1.10x to 1.20x.
- Minimum Interest Coverage: EBITDA must exceed interest expense by a defined multiple, typically 2.0x or higher.
As documented in leveraged finance debt capacity analysis, lenders in the current rate environment typically require minimum DSCR of 1.25x, with 2.0x or higher preferred for quality credits.
2. Incurrence Covenants
Incurrence covenants are only tested when a borrower takes a specific action — issuing new debt, making an acquisition, paying a dividend, or disposing of assets. They are the dominant structure in high-yield bond documentation and have migrated aggressively into broadly syndicated leveraged loans through covenant-lite structures. Incurrence covenants provide no ongoing monitoring signal; they only restrict discrete events.
For direct lenders and mid-market underwriters, a hybrid structure — maintaining at least one springing maintenance covenant alongside incurrence-style baskets — represents current best practice. This preserves early-warning capability while allowing operational flexibility for the borrower.
Walking Through the Numbers: A $10M EBITDA Manufacturing Credit
Consider a specialty manufacturer with the following at-close profile:
- TTM EBITDA: $10.0M
- Total Debt: $45.0M (4.5x leverage)
- Annual Interest Expense: $4.5M (10.0% all-in rate, consistent with current mid-market pricing of 9%–11% for quality credits per current market data)
- Maintenance CapEx: $1.2M
- Mandatory Amortization: $0.9M
At Close — Covenant Headroom Calculation:
- FCCR = (EBITDA − CapEx − Taxes) ÷ (Interest + Amortization) = ($10.0M − $1.2M − $0.8M) ÷ ($4.5M + $0.9M) = $8.0M ÷ $5.4M = 1.48x
- Covenant threshold set at 1.10x — providing 38 basis points of headroom
- Leverage covenant set at 5.50x — providing 100 basis points of headroom on a 4.5x close multiple
Stress Scenario — EBITDA Declines 28% to $7.2M:
- FCCR = ($7.2M − $1.2M − $0.6M) ÷ $5.4M = $5.4M ÷ $5.4M = 1.00x — a covenant breach at the 1.10x threshold
- Leverage = $45.0M ÷ $7.2M = 6.25x — a breach of the 5.50x maximum
The breach triggers a technical default, giving the lender an immediate right to accelerate, renegotiate terms, require a cure, or extract additional structural protections — all while the business still has meaningful going-concern value. This is the covenant system working as designed.
Per OHA’s February 2026 Credit Investor’s Checklist, recovery rates are a primary driver of long-term credit performance. A portfolio with higher recoveries generates materially higher loss-adjusted spreads than a lower-recovery portfolio with identical nominal spreads — reinforcing why early intervention through covenant breach is economically critical.
What Institutional Lenders Look For — and Where Structures Break Down
Experienced credit managers evaluate covenant packages along three dimensions: tightness (how close the threshold is to current performance), definition quality (how EBITDA and debt are defined), and cure rights (what the borrower can do if a breach occurs).
📊 Credit Performance Data Points
- Borrowers with EBITDA ≥ $100M experienced a 32% lower default rate than smaller borrowers from 1995–2025 across multiple credit cycles, per Pitchbook LCD data cited by OHA (December 2025).
- Total leverage in most leveraged buyouts now ranges from 4.0x to 5.0x EBITDA, down from levels exceeding 6.0x common in 2018–2019, per current leveraged finance market data.
- Sponsor equity contributions have risen to 45%–55% of purchase price from 35%–40% in lower-rate environments, reflecting tighter lender scrutiny.
- Fitch’s methodology requires EBITDA-based coverage, leverage, and DSCR metrics as core quantitative inputs across all corporate credit ratings, per Fitch Ratings Procedures and Methodologies (2025).
- Moody’s Analytics framework uses probability of default (PD), loss given default (LGD), and correlation modeling to quantify portfolio-level credit risk, per Moody’s Analytics RiskFrontier Methodology.
Common Structural Mistakes That Undermine Protection
- Loose EBITDA definitions: Allowing excessive addbacks — synergies, one-time charges, cost-saving projections — inflates the denominator and artificially widens covenant headroom. Per OHA, loose EBITDA definitions are a known borrower tactic used to modify leverage calculations in their favor.
- Excessive equity cure rights: Permitting unlimited equity cures allows borrowers to paper over structural deterioration without addressing root causes. Experienced lenders limit cures to two per year and require that cured EBITDA not be double-counted.
- Uncapped investment and distribution baskets: Builder baskets that grow with EBITDA can permit significant asset stripping in a stress scenario. PGIM’s large-cap private credit strategy explicitly cites tighter security documentation as a key differentiator in its January 2026 underwriting case study.
- Absent cross-default provisions: Without cross-default language, a breach in a subsidiary facility may not trigger the senior credit agreement, leaving the lender blind to cascading deterioration.
Covenant Underwriting Checklist: 10 Diligence Points
- Define EBITDA precisely. Cap addbacks at 15%–20% of reported EBITDA. Explicitly exclude recurring charges, unrealized synergies, and non-cash revenue items.
- Set maintenance covenant thresholds with scenario analysis. Test covenant headroom against a base, downside (−20% EBITDA), and severe stress (−35% EBITDA) scenario before agreeing to any threshold.
- Require at least one quarterly-tested maintenance covenant even in covenant-lite structures. A springing leverage covenant triggered at 35% revolver utilization is a minimum acceptable standard.
- Restrict equity cure rights. Maximum two cures in any rolling twelve-month period; no double-counting of cured amounts in forward EBITDA calculations.
- Cap permitted debt and lien baskets. Limit incremental facilities (accordion features) to the greater of a fixed dollar amount or leverage-neutral incurrence — not open-ended EBITDA multiples.
- Include asset sale sweep provisions. Require 100% of net proceeds from non-ordinary-course asset sales to prepay debt above a de minimis threshold.
- Establish information rights with teeth. Require monthly management accounts within 30 days, quarterly compliance certificates, and annual audited financials. Include a lender call right upon covenant breach.
- Negotiate cross-default and cross-acceleration provisions across all material debt instruments above a defined threshold (typically $1M for mid-market credits).
- Build covenant step-downs. If the deal includes an acquisition or integration period, step covenants down quarterly toward a steady-state level rather than holding at a single static threshold throughout the term.
- Document LME protections explicitly. Include anti-J.Crew provisions (restricting IP transfers to unrestricted subsidiaries), anti-Serta provisions (limiting uptier exchange offers), and majority lender consent requirements for any debt subordination.
Key Takeaways
- Covenants are early-warning systems, not formalities. A well-designed maintenance covenant package gives lenders the right to intervene while the business still has going-concern value — dramatically improving recovery outcomes versus detecting distress at payment default.
- EBITDA definition quality determines covenant effectiveness. Loose addback language can inflate reported EBITDA by 20%–30%, creating illusory headroom that evaporates under stress. Underwriters must control the definition at origination.
- Larger borrowers default less — but require tighter documentation. Per Pitchbook LCD data (December 2025), borrowers above $100M EBITDA exhibit 32% lower default rates, but their scale gives them greater capacity to execute LMEs. Documentation rigor must scale with borrower sophistication.
- Recovery rates drive realized returns more than nominal spread. OHA’s February 2026 analysis illustrates that a higher-recovery portfolio generates materially superior loss-adjusted returns versus an identical-spread, lower-recovery portfolio — making covenant-enabled early intervention a direct source of investment alpha.
- Stress testing covenant headroom is a required underwriting step, not optional analysis. Every credit submission should include a covenant sensitivity table showing the EBITDA decline required to trigger a breach — ensuring that thresholds provide meaningful early warning across a realistic range of adverse outcomes.
