The Underwriting Problem at the Center of Every Deal
Imagine you are evaluating a senior secured term loan for a mid-market specialty distribution business. The sponsor-prepared confidential information memorandum reports $12.4 million in LTM (last twelve months) Adjusted EBITDA: enough, at 4.5x total leverage, to justify a $55 million credit facility. The deal looks clean. The borrower has strong customer concentration in a resilient end market, and the sponsor has meaningful equity behind the debt.
Then you open the add-back schedule.
Inside, you find $3.1 million in claimed adjustments: $800K in one-time management consulting fees, $600K in “non-recurring” IT migration costs (for the third consecutive year), $450K in above-market owner compensation, $750K in projected cost savings from a vendor renegotiation not yet executed, and $500K for a resolved legal settlement. Strip out the unsubstantiated items and you are looking at a business generating roughly $10.0 to $10.5 million in defensible, recurring EBITDA, and total leverage of 5.2x to 5.5x. That gap changes your credit decision entirely.
This is the EBITDA normalization problem. It sits at the core of every leveraged credit underwrite, and getting it wrong, in either direction, is one of the most consequential errors a credit team can make. This guide walks through the full analytical framework, the key calculations, the red flags that matter most, and how institutional lenders are approaching the problem in 2025–2026.
The Normalization Framework: Building the EBITDA Bridge
The starting point for any underwrite is understanding that reported EBITDA, adjusted EBITDA, and normalized EBITDA are distinct, and intentionally so. As practitioners define it, normalized EBITDA is a defensible estimate of a company’s sustainable, recurring operating earnings, adjusted to remove non-recurring, non-operating, and owner-specific items while reflecting a supportable run-rate for underwriting. Adjusted EBITDA, by contrast, is often used more broadly and can include deal-specific pro forma assumptions that normalized EBITDA would not accept.
The full EBITDA bridge moves through four stages:
- Reported EBITDA: GAAP-basis operating earnings plus D&A, as stated in audited or reviewed financials.
- Adjusted EBITDA: Reported EBITDA plus all sponsor- or borrower-proposed add-backs, as presented in the CIM.
- Normalized EBITDA: Adjusted EBITDA after the lender’s independent Quality of Earnings (QoE) review removes unsupported or non-recurring items. This is the number lenders underwrite.
- Pro Forma EBITDA: Normalized EBITDA plus credibly documented forward drivers (completed acquisitions, signed contract revenue, verifiable cost savings) that a lender may give partial or full credit for.
A Quality of Earnings analysis is the primary tool for testing adjustments across this bridge, reconciling the EBITDA narrative with the general ledger, cash receipts, and operating trends. The QoE validates the bridge from reported results to normalized, tests revenue quality and recurring patterns, and stress-tests working capital behavior, all of which directly affect debt sizing and covenant architecture.
Institutionally, lenders apply a simple but effective classification system to each add-back:
- Green-light items: Non-market owner compensation, personal expenses run through the business, one-time legal settlements with documented resolution, discontinued business lines with verifiable cost structure, duplicate vendor fees from a completed integration.
- Yellow-light items: Temporary cost cuts without contracts in place, vendor renegotiations claimed but not yet executed, growth hires treated as add-backs rather than pro forma drivers.
- Red-light items: Evergreen consulting arrangements that recur annually, recurring “one-time” restructuring charges, structural overhead that will return post-close, aggressive revenue add-backs for business interruption without insurance documentation.
The distinction matters because lenders use the most conservative defensible EBITDA definition to drive debt sizing, covenant levels, and repayment capacity modeling. An adjustment that cannot be traced directly to the general ledger and substantiated with documentation (invoices, board resolutions, contracts, compensation surveys) should not survive underwriting.
Walking Through the Calculations: A $12M EBITDA Manufacturing Transaction
Return to the distribution business from our opening scenario. Here is how a rigorous underwriting team works through the EBITDA bridge:
EBITDA Bridge: Illustrative Example
| Item | Amount ($000s) | Underwriting Status |
|---|---|---|
| Reported LTM EBITDA | $9,300 | Baseline |
| + Above-market owner compensation | $450 | ✔ Green-light (comp survey documented) |
| + Personal expenses (owner vehicle, travel) | $120 | ✔ Green-light (receipts verified) |
| + One-time legal settlement (resolved) | $500 | ✔ Green-light (court order on file) |
| + “Non-recurring” IT migration costs | $600 | ⚠ Yellow-light (third consecutive year) |
| + Management consulting (recurring vendor) | $800 | ✖ Red-light (evergreen contract confirmed) |
| + Projected vendor savings (not contracted) | $750 | ✖ Red-light (no executed agreement) |
| Lender-Normalized EBITDA | ~$10,370 | Underwriting Basis |
Illustrative example for educational purposes. Numbers are hypothetical.
With normalized EBITDA of approximately $10.4 million and total debt of $55 million, leverage re-underwriters to 5.3x: materially above a typical first-lien threshold of 4.0x–5.0x for this credit profile. Interest coverage (EBITDA / cash interest expense), assuming a floating-rate all-in cost of roughly 9.0% on the full facility, computes to approximately 2.1x, a level that warrants scrutiny given the limited margin for earnings compression.
The stress test is equally important. Apply a 15% EBITDA haircut (a common base-case downside scenario for cyclical sectors): normalized EBITDA falls to approximately $8.8 million, leverage climbs to 6.2x, and interest coverage drops to approximately 1.8x. At that level, the business is covering interest, but debt service coverage becomes dangerously thin. This is the scenario framework that determines whether to pass on the deal, resize the credit, or require a tighter structure.
What Institutional Lenders Are Seeing in 2025–2026
The current credit environment underscores why rigorous EBITDA normalization is not a back-office exercise; it is a front-line risk management discipline. Several data points frame the stakes:
📊 Credit Performance Data Box
- Leveraged Loan Default Rate (incl. distressed exchanges): The U.S. Leveraged Loan Index headline default rate stood at approximately 1.5% in Q3 2025, but including distressed Liability Management Exercises (LMEs), the combined rate rose to 4.3% as of Q3 2025. (Source: PitchBook LCD / Northleaf Capital, Q3 2025)
- High-Yield Projected Default Rate: Moody’s projected a high-yield bond issuer realized default rate of 3.2% for calendar year 2025, rising above 4% by Q1 2026, still below the 4.5% historical average since 1996, but elevated relative to recent tightening cycles. (Source: Moody’s Credit Strategy, July 2025)
- First-Lien Recovery Estimates: S&P Global Ratings reported that the average recovery estimate for first-lien new issues held at approximately 64% in Q4 2024, though the rising prevalence of PIK toggle instruments and covenant-lite structures introduces downside recovery risk. (Source: S&P Global Ratings, Q4 2024 Leveraged Finance Update)
- Private Credit / Direct Lending Loss Rates: The Cliffwater Direct Lending Index (CDLI) for senior direct lending produced an estimated annualized credit loss of approximately 1.47% through the study period, reflecting the structural advantage of maintenance covenants and proactive lender engagement. (Source: Cliffwater / Morgan Stanley Investment Management, 2025)
- Covenant Protections Deteriorating in BSL: The Loan Syndications and Trading Association (LSTA) noted in its Q4 2024 covenant trend report that loan documentation for large-sponsor-backed transactions trended looser, with more expansive EBITDA adjustment definitions, more generous grower baskets, and expanded incurrence-based carve-outs to debt caps. (Source: LSTA Loan Market Covenant Trends, Q4 2024)
- Middle Market Leverage & Coverage: MidCap Financial Investment Corporation reported a weighted average borrower net leverage of 5.29x and interest coverage of 2.3x across its core middle market portfolio as of fiscal year-end 2025, consistent with broader BDC sector norms. (Source: MidCap Financial Investment Corp. Form ARS, FY2025, SEC EDGAR)
One structural trend deserves particular attention: the dramatic rise in distressed exchanges as a share of all defaults. According to Moody’s, distressed exchanges accounted for approximately 64% of all defaults in the first half of 2025, roughly the same pace as all of 2024, which itself was the highest annual share on record. This matters for EBITDA underwriting because inflated add-backs that obscure true cash generation capacity are frequently identified as a contributing factor when borrowers enter liability management processes. The risk is not just default at origination; it is the slow deterioration that occurs when a company operates for 12–24 months at leverage ratios its real earnings cannot sustainably support.
For leveraged loan issuers specifically, Moody’s data showed that the loan market exhibited the most distress, particularly borrowers with loan-only capital structures backed by private equity firms, making rigorous first-lien underwriting a non-negotiable discipline. Meanwhile, 2021-vintage deals have shown weaker performance across private credit portfolios, likely attributable to post-COVID earnings inflation that was embedded into EBITDA bases at origination without sufficient normalization scrutiny.
The EBITDA Normalization Underwriting Checklist
Below is a step-by-step diligence framework for evaluating every EBITDA add-back in a credit package. This checklist applies to both buy-side underwriting and sell-side CIM preparation.
Step 1: Obtain the Raw Inputs
- Request 3 years of audited or reviewed financial statements (not just the CIM summary)
- Obtain monthly management accounts for the trailing 24 months
- Request the full GL-level add-back schedule with specific account codes for each adjustment
- Compare the CIM EBITDA bridge line-by-line to the audited statements, identify any gaps
Step 2: Classify and Interrogate Each Add-Back
- Is the item truly non-recurring? Pull three years of history, does it appear in multiple periods?
- Is the item truly non-operational? If it is removed, will the business still generate the claimed revenue and gross margin?
- Is there third-party documentation? Invoices, contracts, board minutes, compensation surveys, legal agreements
- For projected cost savings: Is there an executed agreement or signed contract? What is the implementation timeline and risk?
- For owner compensation adjustments: Is the replacement salary benchmarked to an independent compensation survey?
Step 3: Reconcile to Cash
- Compare normalized EBITDA to operating cash flow from the cash flow statement, large divergences are a warning sign
- Analyze the cash conversion cycle (DSO, DPO, DIO) for working capital distortions
- Identify any revenue timing or deferred revenue recognition issues under ASC 606
- Test whether gross margin is consistent with the revenue base, margin expansion without explanation warrants investigation
Step 4: Run the Leverage and Coverage Test
- Total Leverage: Total Debt / Normalized EBITDA, target threshold varies by sector, typically 3.5x–5.5x for first-lien middle market
- Interest Coverage: Normalized EBITDA / Cash Interest Expense, minimum threshold typically 2.0x–2.5x; below 1.5x is a structural credit concern
- Fixed Charge Coverage: (EBITDA – Capex – Taxes) / (Interest + Required Amortization), tests true debt service capacity
- Stress Test: Run a 15–20% EBITDA haircut and recompute all ratios; identify the EBITDA level at which covenant breach occurs
Step 5: Anchor the Covenant Package
- Tie maintenance covenant levels to the stress-tested (not base case) EBITDA to ensure meaningful lender protection
- Impose caps on aggregate add-backs, typically expressed as a percentage of reported EBITDA, to limit definition creep over the life of the facility
- Require quarterly compliance certificates within 45–60 days of period end, with full supporting add-back documentation
- Negotiate springing financial maintenance tests if covenant-lite terms are required for competitive reasons
Key Takeaways for Credit Practitioners
- 🔵 Normalized EBITDA, not adjusted EBITDA, is the underwriting basis. The distinction between the two can span hundreds of basis points of leverage, and a materially different credit decision. Lenders must build their own EBITDA bridge independently, not accept the sponsor’s number at face value.
- 🔵 Every add-back must clear a documentation standard. If an adjustment cannot be traced to the general ledger and supported by third-party evidence, it should not survive underwriting. The standard is not whether the add-back is commercially reasonable; it is whether it is operationally real and non-recurring.
- 🔵 The cash reconciliation test is the most powerful fraud and error detector in EBITDA analysis. When normalized EBITDA diverges materially from reported operating cash flow, the lender must understand why before sizing the facility. EBITDA that does not convert to cash is not underwritable EBITDA.
- 🔵 Stress testing is mandatory, not optional. With Moody’s projecting leveraged loan default rates near 7.5%–7.9% and distressed exchanges at record shares of all defaults, underwriting to base-case EBITDA without a downside scenario is a structural risk management failure. A 15–20% EBITDA haircut is a standard starting point, not a conservative outlier.
- 🔵 Covenant architecture must follow from normalized, not adjusted, EBITDA. Expansive EBITDA definitions in credit agreements (observed by the LSTA and Haynes Boone as a growing market trend) erode the protective value of financial maintenance covenants. Lenders should impose aggregate add-back caps and tighter ratio tests to preserve the lender protections that differentiate private credit from covenant-lite broadly syndicated loans.
