The Underwriting Question: What Happens When the Business Deteriorates, and When Does the Lender Find Out?
Every credit underwriter faces a version of the same scenario: a borrower’s EBITDA begins eroding six months post-close. Revenue softens. Margins compress. Management blames “temporary headwinds.” The question isn’t whether the lender will eventually find out, it’s how early, and what contractual rights the lender has to act when they do.
That answer lives entirely in the covenant package.
In middle-market direct lending, where borrowers are often privately held, reporting is less frequent, and secondary liquidity is limited, financial covenants serve as the underwriter’s early-warning system. They are not mere legal boilerplate; they are structural instruments of risk management. Yet in an environment where covenant-lite structures have migrated from broadly syndicated loans into the direct lending market, lender protections have eroded in ways that are only fully visible when a credit begins to stress.
According to research published by Oak Hill Advisors (OHA) via T. Rowe Price in February 2026, documentation and structuring expertise at underwriting are key to protecting lenders against harmful borrower actions, including asset stripping tactics popularized in cases like J. Crew, Pluralsight, and Serta. This article builds the covenant framework that experienced lenders use to prevent exactly those outcomes.
The Analytical Framework: Maintenance Covenants vs. Incurrence Covenants
Before designing a covenant package, underwriters must distinguish between two fundamentally different covenant architectures:
- Maintenance Covenants require the borrower to satisfy defined financial ratios on a continuous, periodic basis, typically tested quarterly. They are the gold standard for middle-market lenders because they create regular checkpoints that trigger lender rights before a default, not after.
- Incurrence Covenants are tested only when the borrower takes a specific action: issuing new debt, making an acquisition, or paying a dividend. They are reactive by design and are the dominant structure in broadly syndicated, covenant-lite markets.
For a direct lender underwriting a $10M EBITDA manufacturing company at 4.5x total leverage, relying on incurrence covenants alone means you may not learn of a 30% EBITDA decline until the borrower has already drawn on its revolver, shifted collateral, or executed a liability management exercise (LME) that subordinates your position.
Maintenance covenants, tested quarterly, give the lender legal standing to engage well before that sequence unfolds. According to the CFA Institute’s credit analysis curriculum, the probability of default is a dynamic variable that changes with a borrower’s credit migration, and maintenance covenants provide the mechanism to capture that migration in real time.
The Core Maintenance Covenant Trio
Experienced middle-market underwriters typically build their covenant package around three primary tests:
- Total Net Leverage Ratio: Total Net Debt ÷ Trailing Twelve Month (TTM) EBITDA. Commonly set with a maximum threshold of 0.5x–1.0x above the underwritten entry leverage, with a step-down schedule aligned to the amortization profile.
- Interest Coverage Ratio (ICR): TTM EBITDA ÷ Cash Interest Expense. Minimum floors of 1.75x–2.25x are standard for middle-market borrowers, though the right floor depends on the cash flow profile and interest rate environment.
- Minimum Liquidity Test: Unrestricted cash plus revolver availability, tested monthly or at quarter-end, set at a minimum of 10–15% of total debt outstanding. This is the earliest-warning covenant, liquidity compression precedes EBITDA decline in most stress scenarios.
Walking Through the Calculations: A $10M EBITDA Manufacturing Credit
Consider a borrower with the following underwritten profile at close:
- TTM EBITDA (underwritten, normalized): $10.0M
- Total debt: $45.0M (first lien term loan: $40M, revolver: $5M drawn)
- Entry leverage: 4.5x Total Net Debt / EBITDA
- Annual cash interest expense: $3.6M (SOFR + 550bps on $40M)
- Interest Coverage Ratio at close: 2.78x ($10.0M ÷ $3.6M)
The covenant is set with a maximum leverage of 5.25x and a minimum ICR of 2.00x, tested quarterly. Now, model a stress scenario: six months post-close, TTM EBITDA declines 20% to $8.0M due to customer concentration risk and raw material inflation.
- Stressed Leverage: $45.0M ÷ $8.0M = 5.63x: a covenant breach (maximum: 5.25x)
- Stressed ICR: $8.0M ÷ $3.6M = 2.22x: still inside the minimum (2.00x floor)
The leverage covenant catches the deterioration. Without it, the lender has no contractual standing to demand a management meeting, require a business plan, block restricted payments, or accelerate the loan. With it, the lender can engage immediately: renegotiating terms, adding structural protections, or accelerating if cure rights are not exercised.
This is the operational value of maintenance covenants: they convert credit deterioration from an eventual discovery into an immediate, rights-generating event.
What Institutional Lenders Require: Standards, Red Flags, and Performance Data
Institutional direct lenders have developed specific covenant documentation requirements that go beyond ratio thresholds. Research published by PGIM in January 2026 highlights that tighter security documentation is a defining characteristic of large-cap private credit, and that superior covenant packages correlate with lower defaults and higher recovery rates across credit cycles.
Key institutional standards include:
- Defined EBITDA: Covenant EBITDA definitions must cap addbacks. Loose addback language (permitting pro forma synergies, run-rate cost savings, or open-ended management adjustments) inflates the denominator of every leverage and coverage test, artificially widening covenant headroom. OHA’s research specifically flags loose EBITDA definitions as a documented tactic used by borrowers to dilute lender protections.
- Equity Cure Rights: Borrowers are typically permitted to cure a covenant breach by injecting equity within a defined window (usually 10–15 business days post-quarter). Institutional lenders limit cure frequency (no more than two times in any rolling four-quarter period) and cap the cure amount to prevent repeated use as an operational substitute for actual performance improvement.
- Step-Down Schedules: Maximum leverage thresholds should tighten over time, aligned with the expected deleveraging path underwritten at close. A static covenant that never steps down provides diminishing protection as the business ages into its debt.
- Anti-LME Protections: Given the proliferation of out-of-court restructurings, underwriters must negotiate explicit restrictions on collateral transfers, priming debt incurrence, and uptiering transactions without unanimous lender consent.
📊 Credit Performance Data Points
- Borrowers with EBITDA ≥ $100M experienced a 32% lower default rate than smaller borrowers from 1995–2025, based on Pitchbook LCD data cited by OHA (February 2026).
- Portfolios with higher recovery rates generate materially higher loss-adjusted spreads even at identical nominal spreads and default rates, OHA’s illustrative analysis shows a 44% higher loss-adjusted spread for high-recovery vs. low-recovery portfolios.
- Moody’s Analytics’ RiskFrontier framework identifies Loss Given Default (LGD) and Probability of Default (PD) as the two primary drivers of credit portfolio risk, both of which are directly influenced by covenant strength and structural protections.
Underwriting Covenant Checklist: 10 Diligence Points Before Close
- Map the EBITDA definition: Identify every permitted addback category. Quantify the gap between reported EBITDA and covenant EBITDA. Flag any open-ended or uncapped addback baskets.
- Set leverage headroom intentionally: Covenant maximum should be no more than 0.75x–1.0x above entry leverage. Document the stress scenario that would breach the covenant.
- Calibrate the ICR floor to interest rate sensitivity: In floating rate structures, model the ICR under a +200bps rate shock. Ensure the covenant floor still provides meaningful protection.
- Build a minimum liquidity test: Do not rely solely on leverage and coverage. Liquidity is the leading indicator, cash compression precedes EBITDA compression in most stress paths.
- Negotiate restricted payment blockers: Confirm that dividends, distributions, and management fees are blocked during any covenant breach or cure period.
- Define the testing cadence: Quarterly testing for leverage and coverage; monthly for liquidity in credits with thin coverage ratios or seasonal cash flow profiles.
- Limit equity cure frequency and size: Maximum two cures per four quarters; cure amount capped at the minimum needed to remedy the breach, not to create headroom.
- Require financial reporting within 45 days of quarter-end: Late or incomplete reporting must itself trigger a covenant default after a defined grace period.
- Negotiate step-down schedules: Maximum leverage should decline 0.25x–0.50x per year, consistent with the base case deleveraging path modeled at underwriting.
- Review anti-LME provisions: Confirm that collateral transfers to unrestricted subsidiaries, priming debt incurrence, and amendment thresholds require supermajority or unanimous consent.
Key Takeaways
- 🔵 Maintenance covenants are not optional for middle-market lenders. Incurrence-only structures eliminate the lender’s ability to identify and respond to credit deterioration before it becomes irreversible.
- 🔵 EBITDA definition quality determines covenant quality. A leverage covenant built on inflated, addback-heavy EBITDA is functionally a covenant-lite structure, regardless of the ratio threshold on paper.
- 🔵 Liquidity testing is the leading indicator. Cash runway compression is the earliest observable signal of borrower stress and should trigger lender engagement before leverage or coverage covenants breach.
- 🔵 Recovery rates are as important as default rates. Structural protections (tight covenants, collateral perfection, anti-LME provisions) are the primary levers that determine recovery outcomes when a credit defaults.
- 🔵 Covenant monitoring is an active, ongoing discipline. The value of a well-designed covenant package is only realized through rigorous quarterly review, prompt breach identification, and a pre-defined escalation protocol, not just at the time of underwriting.
