The Underwriting Question Every Credit Officer Must Answer
You have just approved a $40 million senior secured term loan to a specialty manufacturer with $10 million in trailing twelve-month EBITDA. Leverage at close sits at 4.0x. The sponsor argues the business is stable, margins are expanding, and working capital is seasonal but manageable. Twelve months later, EBITDA has drifted to $7.8 million, a new competitor has entered the market, and the borrower’s revolver is nearly fully drawn.
The question is not whether the credit has deteriorated. It clearly has. The question is: did your covenant package give you the information, the leverage, and the timeline to act before value was permanently impaired?
This is the central discipline of covenant design. In a market where ainvest.com reports that “bad PIK” conversions, retroactive deferrals of cash interest, have surged from 36.7% in Q4 2021 to 57.2% in Q3 2025, structural weaknesses are no longer theoretical. They are appearing in real portfolios, in real losses, right now.
The Analytical Framework: Two Covenant Architectures
Institutional lenders distinguish between two fundamental covenant architectures, each serving a different protective purpose:
1. Maintenance Covenants (“Guard Rails”)
Maintenance covenants are tested periodically, typically every fiscal quarter, regardless of borrower action. They create an ongoing obligation to meet a defined financial threshold. If breached, the lender receives an immediate event of default (or a cure right period), creating negotiating leverage before the borrower reaches a liquidity crisis.
Common maintenance covenants include:
- Maximum Total Leverage Ratio: Total Debt ÷ EBITDA ≤ agreed threshold (e.g., 5.0x stepping down to 4.5x by year two)
- Minimum Interest Coverage Ratio (ICR): EBITDA ÷ Cash Interest Expense ≥ 2.0x–3.0x
- Minimum Fixed Charge Coverage Ratio (FCCR): (EBITDA − CapEx − Taxes − Distributions) ÷ Debt Service ≥ 1.10x–1.25x
- Maximum Capital Expenditure: Limits discretionary spending that could reduce free cash flow
2. Incurrence Covenants (“Permission Gates”)
Incurrence covenants are only tested when a borrower takes a specific action: issuing new debt, making an acquisition, paying a dividend. They are the dominant structure in broadly syndicated loans and increasingly in larger private credit transactions. As ibinterviewquestions.com notes, total leverage in most LBO deals now ranges from 4.0x to 5.0x EBITDA, down from over 6.0x, a shift that reflects the re-pricing of incurrence risk in a higher-rate environment.
For direct lenders in the middle market, maintenance covenants remain the gold standard. They are early-warning systems. Incurrence-only packages effectively remove the lender’s ability to intervene until a borrower voluntarily crosses a threshold, by which point, recovery value may already be compromised.
Walking Through the Calculations: A $10M EBITDA Manufacturer
Return to the manufacturing borrower. At close, the credit profile looks like this:
| Metric | At Close | 12-Month Drift | Covenant Threshold |
|---|---|---|---|
| EBITDA (TTM) | $10.0M | $7.8M | – |
| Total Debt | $40.0M | $41.2M (revolver drawn) | – |
| Leverage Ratio | 4.0x | 5.28x | ≤ 5.0x |
| Interest Coverage (ICR) | 2.9x | 2.25x | ≥ 2.50x |
| FCCR | 1.38x | 1.07x | ≥ 1.15x |
At month twelve, all three maintenance covenants are in breach. The lender now holds three simultaneous default notices: providing real negotiating leverage to demand a forbearance agreement, a management change, or an equity cure. Without maintenance covenants, none of these triggers would exist. The borrower could continue deferring issues until liquidity fully collapsed.
This is precisely the dynamic that drives PIK conversion abuse. According to ainvest.com, retroactive PIK conversions occur most often when borrowers face underperformance, effectively deferring interest while masking the true deterioration that well-structured covenants would have surfaced earlier.
What Institutional Lenders Require: Standards and Red Flags
The Basel Committee on Banking Supervision’s updated bis.org Principles for the Management of Credit Risk (April 2025) reaffirm four pillars that directly govern covenant design and monitoring: (i) establishing a sound credit risk environment, (ii) rigorous credit-granting processes, (iii) ongoing credit administration and measurement, and (iv) adequate controls over credit risk. Covenant structures must satisfy all four.
Against that framework, experienced institutional lenders apply the following standards:
- Covenant headroom of 20–30%: Set thresholds at least 20% more conservative than projected performance to absorb forecast error without immediate breach.
- Quarterly EBITDA definitions are negotiated, not accepted: Scrutinize add-backs. Aggressive EBITDA normalization inflates the denominator of every coverage ratio. An EBITDA figure loaded with non-recurring adjustments creates phantom headroom.
- Equity cure rights should be capped: Limit cure events to no more than two per year and four over the loan life. Unlimited cures convert maintenance covenants into incurrence covenants de facto.
- Cascade testing after cure: Require re-testing of all financial covenants after an equity cure is injected to confirm true compliance, not just technical compliance.
📊 Data Box: Credit Performance Benchmarks
- Private credit reported 8.6% net IRR with only 0.4% loss rates, compared to 1.1% loss rates for broadly syndicated leveraged loans, reflecting the protective benefit of tighter covenant packages in direct lending. (ainvest.com)
- The private credit market reached $1.67 trillion in 2025, projected to grow to $2.9 trillion by 2030 at an 11.62% CAGR, increasing competition pressure on underwriting discipline. (ainvest.com)
- Total LBO leverage has compressed to 4.0x–5.0x EBITDA in 2025–2026, down from 6.0x+ in prior cycles, directly affecting where maintenance covenant thresholds must be set. (ibinterviewquestions.com)
- “Bad PIK” conversions reached 57.2% of PIK activity in Q3 2025, up from 36.7% in Q4 2021, a structural warning sign of covenant erosion across the market. (ainvest.com)
Common Red Flags in Covenant Packages
- 🚩 EBITDA definitions exceeding 15–20% in add-backs relative to reported net income, a sign that the metric is being engineered rather than measured
- 🚩 Springing covenants tied to revolver utilization; these only activate when the borrower has already drawn heavily, eliminating early-warning value
- 🚩 “Grower” covenant baskets tied to a percentage of EBITDA; these automatically expand as EBITDA grows, reducing protective value over time
- 🚩 No cross-default provisions to affiliate or parent entities, allows structural deterioration to occur in adjacent entities without triggering lender rights
Covenant Underwriting Checklist: 10 Non-Negotiable Diligence Points
- Define EBITDA precisely. Enumerate every permitted add-back category, cap run-rate synergies, and require sign-off from the independent auditor on non-recurring classifications.
- Set leverage covenant at a 25% discount to base-case projections. If management forecasts 4.0x at year-end, set the covenant at 5.0x maximum, not 4.5x.
- Layer all three coverage tests simultaneously: leverage, ICR, and FCCR. No single ratio captures the full debt service burden.
- Request trailing twelve-month (TTM) financials tested quarterly, not annually. Annual testing creates a 12-month blind spot between measurements.
- Stress test covenant headroom under a 15–20% EBITDA decline scenario before finalizing thresholds. If the covenant is breached immediately under a mild stress, renegotiate at origination.
- Cap equity cure rights explicitly in the credit agreement: frequency, size, and form of cure (cash only, no new indebtedness).
- Include a minimum liquidity covenant (e.g., unrestricted cash + revolver availability ≥ $2M) as a floor, regardless of EBITDA performance.
- Negotiate real-time financial reporting obligations: monthly management accounts within 30 days of month-end for credits at or below 4.5x leverage.
- Document covenant calculation methodology in a compliance certificate template attached to the credit agreement. Ambiguity in definitions is the most common source of covenant disputes.
- Require lender consent for EBITDA definition amendments. Borrowers frequently attempt to renegotiate add-back definitions informally during amendment requests.
Key Takeaways
- 🔑 Maintenance covenants are early-warning systems, not formalities. A well-structured package surfaces deterioration before the borrower reaches a liquidity crisis and recovery value is impaired.
- 🔑 EBITDA definition engineering is the single greatest risk in covenant design. Inflated add-backs compress every coverage ratio simultaneously and create phantom headroom that disappears under stress.
- 🔑 The surge in bad PIK conversions to 57.2% in Q3 2025 is a market-wide signal that underwriting discipline has eroded in covenant-lite and covenant-loose structures across private credit.
- 🔑 Quarterly testing with 20–30% headroom is the institutional standard. Annual testing and thin headroom are sponsor-friendly concessions that transfer risk from equity to debt, without commensurate pricing.
- 🔑 Covenant monitoring is as critical as covenant design. A lender who cannot detect a breach within 30–45 days of quarter-end (or who waives breaches without renegotiating terms) has effectively surrendered structural protection.
