The Underwriting Challenge

Picture a sponsor-backed manufacturing business presenting a $10M adjusted EBITDA figure in its Confidential Information Memorandum (CIM). Reported EBITDA from the audited financials is $6.8M. The $3.2M bridge includes: $1.1M in “one-time” ERP implementation costs that have recurred for three consecutive years; $800K of owner compensation the sponsor claims is above-market; $750K in projected but unimplemented cost savings; and $550K in COVID-era freight surcharges the management team insists will never return.

Every item in that bridge sounds reasonable in isolation. Collectively, they represent a 47% premium over reported earnings, and the entire credit thesis hinges on which version of EBITDA a lender is willing to underwrite. At a 5.0x leverage multiple, the difference between $6.8M and $10.0M in underwritten EBITDA means $16M in incremental debt capacity. That is not an accounting exercise; it is a risk decision.

This is the central tension in leveraged credit underwriting. As deal environments grow more competitive and sponsor influence over loan documentation expands, the distance between reported and adjusted EBITDA continues to widen. Moody’s research has flagged that the “true” leverage in private credit portfolios could be closer to 7x debt-to-EBITDA once the effect of aggressive adjustments is accounted for, materially higher than headline figures suggest. A lender who accepts sponsor-produced EBITDA at face value is not underwriting; they are originating.

The Framework: How Institutional Lenders Normalize EBITDA

EBITDA normalization is distinct from adjusted EBITDA. The goal of normalization is to answer a specific underwriting question: What is the maintainable, arm’s-length earnings this business will generate under new ownership, in a steady-state operating environment, without the benefit of non-recurring tailwinds?

Sound underwriting builds a normalization bridge in four discrete layers:

Layer 1: Reported EBITDA (the Anchor)

Begin with GAAP-audited or reviewed financials for the trailing twelve months (TTM) and prior two full fiscal years. TTM is the primary measurement period; the multi-year view reveals trend and volatility. Run-rate figures that are not yet in the trailing period should carry a significant evidentiary burden before receiving credit.

Layer 2: Standard Add-Backs (the High-Confidence Tier)

These are adjustments that virtually all institutional lenders accept without material friction: transaction-related legal and advisory fees tied to the current sale process, non-cash stock-based compensation, one-time litigation settlements with documented resolution, and non-recurring insurance recoveries. Each must be traced to source documents (invoices, settlement agreements, board minutes) and must be non-recurring by legal or factual definition, not just management assertion.

Layer 3: Normalized Items (the Judgment Tier)

Owner and related-party compensation requires a market replacement cost analysis. If the founder earns $2.5M but the role can be replaced at $600K, only the excess $1.9M is addable, and only after confirming the replacement is not already planned and unfunded. Related-party lease arrangements that are below market should generate a downward normalization (an add-down) reflecting the arm’s-length cost a buyer would actually incur.

Layer 4: Run-Rate and Synergy Add-Backs (the High-Skepticism Tier)

Projected cost savings, announced price increases not yet fully realized, and integration synergies are the most frequently inflated inputs in sponsor-presented EBITDA. Best institutional practice requires that pro-forma adjustments forecasted to materialize beyond 12 months receive no credit without independent verification. The IOSCO good-practice framework for leveraged loan underwriting explicitly calls for “additional scrutiny to be applied to add-backs associated with cost-savings and synergies that are forecast to occur beyond 12 months.”

The Analysis: Walking Through Calculations and Stress Testing

Once a defensible normalized EBITDA is established, the underwriter builds the credit case around three core ratios and then stress-tests each.

Key Metrics

• Total Leverage (Debt / Normalized EBITDA): The primary sizing constraint. Middle-market first-lien debt averaged 4.5x leverage as of September 2024, according to PGIM, compared to 5.8x in the broadly syndicated loan market. For a lender targeting first-lien senior secured positions, underwriting to below 4.5x normalized EBITDA provides a meaningful cushion above market averages.
• Debt Service Coverage Ratio (DSCR): Normalized EBITDA minus capex and cash taxes, divided by total annual debt service. A DSCR below 1.20x is a standard threshold for covenant scrutiny; below 1.05x signals distress territory. Critically, DSCR calculations should use lender-underwritten normalized EBITDA, not management’s adjusted figure, because overly aggressive normalization can produce a DSCR that appears compliant on paper but fails under independent review.
• Interest Coverage (EBITDA / Cash Interest): S&P Global Ratings data showed that 20% of U.S. middle-market companies had interest coverage below 1.0x in 2024, meaning cash flow was barely covering interest payments. This figure jumped sharply as floating-rate borrowers absorbed the full impact of rate hikes on older vintage loans.

Stress Scenario: 10% EBITDA Compression

S&P’s mild stress scenario, a 10% EBITDA decline combined with a 50-basis-point rate increase, pushes over 25% of middle-market firms below 1.0x interest coverage. Applied to our example company:

• Lender-normalized EBITDA of $7.45M, stressed 10%: $6.71M
• Total debt at 4.6x normalized EBITDA: $34.3M (at SOFR + 500bps, ~9.3% all-in)
• Annual cash interest: ~$3.19M
• Stressed EBITDA / Cash Interest: 2.10x: still serviceable, but any further revenue deterioration eliminates the margin quickly

If the underwriter had accepted the sponsor’s $10.0M EBITDA and structured to 5.0x, stressed EBITDA of $9.0M would need to service $50M in debt at ~$4.65M annually, compressed coverage of just 1.94x, with the real business generating only $6.71M in EBITDA.

Industry Standards: What Institutional Lenders Require

Institutional credit committees do not accept EBITDA schedules from sponsors or management teams as conclusions. They treat them as starting points. The architecture of disciplined underwriting requires that any adjustment be traceable, reasonable, and consistent with how the business will operate after closing.

Common Red Flags

1. Recurring “non-recurring” costs: If a cost has appeared in each of the last three years, it is by definition recurring. ERP upgrades, legal disputes, and “pandemic-related” disruptions that persist into 2025 and 2026 are high-frequency examples.
2. Add-backs that exceed 20-40% of reported EBITDA: Some CLO structures have implemented explicit EBITDA add-back caps, one recent facility capped non-GAAP add-backs at 40% of reported EBITDA for borrowers with EBITDA below $50M. A bridge that stretches beyond this range demands exceptional documentation.
3. Unimplemented cost savings receiving full credit: Academic research on leveraged loan markets found that EBITDA add-backs, particularly unsubstantiated ones, increase the likelihood of loan delinquency and default, and increase the probability of borrower ratings downgrades.
4. Understated related-party costs: Below-market rent from an affiliated entity, underpaid founder labor, or discounted management fees from a parent company all create a need for downward normalization. A credible EBITDA bridge includes both add-backs and add-downs.
5. PIK income in the cap structure: Payment-in-kind arrangements in co-investment or subordinated positions inflate reported EBITDA without generating cash. Moody’s has flagged that the use of PIK arrangements is positively correlated with borrower stress, absent a strong sponsor backstop.

EBITDA Normalization: The Underwriting Checklist

✅ Documentation Requirements

• Three years of audited or CPA-reviewed financial statements plus TTM management accounts
• General ledger support for every line-item add-back exceeding $100K
• Third-party Quality of Earnings (QoE) report from an independent accounting firm
• Organizational chart and compensation register for all owner-related parties
• Executed contracts or invoices supporting any “one-time” cost claim

✅ Questions to Ask Management

• Has this cost appeared in any prior period? If so, why is it classified as non-recurring?
• What is the documented market replacement cost for any owner or family member on payroll?
• For projected savings: What implementation steps remain? Who is responsible? What is the documented timeline?
• Are any related-party lease, service, or supply agreements at below-market rates?
• What is the normalized capex requirement to maintain current revenue levels? (Capex often understated in EBITDA-focused presentations)

✅ Underwriter’s Independent Steps

1. Reconstruct reported EBITDA independently from financial statements before reviewing management’s bridge
2. Apply conservative market replacement costs to owner compensation using comparable job postings or industry surveys
3. Cap unverified cost savings at 50% of claimed benefit pending documentation; reject savings dependent on post-close actions entirely
4. Model base, stress (−10% EBITDA), and downside (−20% EBITDA) scenarios against debt service
5. Verify that the QoE firm operated independently of management and the deal sponsor
6. Document the rationale for every accepted and every rejected add-back in the credit approval memo