The Underwriting Challenge: Whose EBITDA Are You Lending Against?

A borrower walks into a credit process presenting $12 million in Adjusted EBITDA. Their investment banker’s CIM lists $2.4 million in addbacks: management fees being eliminated, a one-time legal settlement, owner perks, and projected cost savings from a new ERP system. At a 4.5x leverage multiple, the deal pencils to $54 million in debt. But what happens when you strip away the assumptions?

This is the defining tension in direct lending credit underwriting: the EBITDA a borrower presents, and the EBITDA a lender can defensibly underwrite, are frequently not the same number. The gap between the two (and how a lender navigates it) is where credit quality is won or lost.

In today’s direct lending environment, this discipline has never been more critical. Fitch Ratings reported the U.S. private credit default rate reached a record 6.0% in April 2026, while Proskauer’s Private Credit Default Index recorded a 2.73% default rate in Q1 2026, up from 1.84% just two quarters earlier. Many of those stressed credits trace their roots to underwriting built on EBITDA that never materialized. Getting normalization right isn’t a formality, it’s the foundation of the entire credit thesis.

The Analytical Framework: Building a Three-Layer EBITDA Bridge

Professional credit underwriters don’t accept a single EBITDA number. They build a structured bridge across three distinct layers, each demanding independent verification:

Layer 1: Reported EBITDA

Start with the audited or reviewed financials. Net income + Interest + Taxes + Depreciation + Amortization. This is your baseline. Make no adjustments at this stage. Document the source: audited statements, reviewed statements, or management-prepared financials, each carries a different reliability weight in your credit memo.

Layer 2: Adjusted EBITDA (Removing Non-Recurring Items)

Here you apply adjustments that improve comparability and remove genuine one-time distortions. The most commonly accepted addbacks include: excess owner compensation above market replacement cost, one-time legal settlements where the matter is fully resolved, transaction-related advisory fees directly tied to the sale process, non-recurring severance and restructuring costs, and related-party rent or management fees normalized to market rates. This is Adjusted EBITDA, a cleaning exercise, not a forecasting exercise.

Layer 3: Normalized (Underwriting) EBITDA

This is the number you actually lend against. Normalized EBITDA goes further than adjusted EBITDA: it reflects the earnings a buyer can sustainably underwrite after considering steady-state staffing, market-rate compensation, and necessary reinvestment costs under new ownership. Critically, normalized EBITDA can be lower than adjusted EBITDA when the business requires additional costs under new ownership that weren’t present historically: a replacement CFO, an upgraded IT infrastructure, a sales team to offset founder-driven revenue.

Walking Through the Calculations: DSCR, Leverage, and Scenario Stress

Once the underwriter has established a Normalized EBITDA, three core metrics govern the credit decision:

1. Leverage Ratio (Net Debt / Normalized EBITDA)

This is the primary structural metric. Using the example above with $54M in proposed debt and $9.23M in normalized EBITDA, the underwritten leverage ratio is 5.85x: materially higher than the headline 5.0x presented in the CIM. In direct lending, typical maximum leverage covenants for middle market transactions sit around 4.0x to 4.5x for lower middle market companies (sub-$20M EBITDA), and up to 4.6x for upper middle market borrowers. At 5.85x, this deal exceeds standard covenant thresholds and warrants either a debt reduction or a re-pricing.

2. Debt Service Coverage Ratio (DSCR)

DSCR measures how much normalized cash flow is available to cover total debt service (interest + scheduled principal). A lender target of at least 1.20x–1.25x is standard. Using $9.23M normalized EBITDA, less $1.2M in maintenance capex (estimated at ~13% of EBITDA for manufacturing), leaves approximately $8.03M in cash available for debt service. If annual debt service on the $54M facility is approximately $6.5M (at SOFR + 500bps on a 5-year term), DSCR is 1.24x, right at the minimum threshold. Any further erosion in EBITDA puts the deal below 1.0x coverage.

3. Stress Testing the Normalization

Run a downside scenario applying a 15%–20% revenue decline (equivalent to a moderate recession or key customer loss). In this case, a 15% revenue drop that flows through at a 60% incremental margin reduces normalized EBITDA to approximately $7.5M. Leverage spikes to 7.2x and DSCR falls to 1.01x, effectively a covenant breach at origination under stress. This analysis, done pre-close, is the difference between disciplined underwriting and performance-chasing.

Industry Standards: What the Data Says About EBITDA Discipline and Credit Performance

The consequences of poor EBITDA normalization discipline are now showing up clearly in market-wide data. Practitioners should understand the following benchmarks:

Top Red Flags Institutional Lenders Watch For

• Run-rate revenue from contracts not yet signed. Anticipated new business belongs in the investment thesis, not the EBITDA bridge.
• Cost savings contingent on post-closing actions. ERP synergies, headcount reductions, or vendor renegotiations that haven’t been implemented should be rejected as addbacks.
• Recurring expenses relabeled as one-time. A company that has restructuring charges in three consecutive years is not non-recurring. A Proskauer survey identified EBITDA addbacks as a top-ranked risk in direct lending, alongside covenant-lite structures.
• Owner compensation below market replacement cost. A founder paying himself $150,000 to run a $50M revenue business creates a hidden labor cost that will emerge under professional management.
• Related-party rent or management fees at below-market rates. These represent embedded subsidies that disappear at closing and must be added back as costs, not removed as addbacks.
• LTM EBITDA that is significantly above any prior trailing period. A sudden EBITDA spike in the trailing twelve months before a sale process should prompt deep investigation into revenue recognition timing, cost deferrals, or working capital manipulation.

The EBITDA Normalization Underwriting Checklist

Use this framework on every credit that involves adjusted EBITDA in the underwriting thesis:

1. Obtain and verify source financials. Audited > reviewed > compiled > management-prepared. Understand the quality of the accounting before accepting any number.
2. Build the EBITDA bridge independently: do not start from the seller’s schedule. Build from reported net income and reconstruct line by line.
3. Require documentation for every addback. Board resolutions, invoices, legal settlement agreements, comp benchmarking data, every adjustment must have a paper trail.
4. Apply the three-part addback test: (a) Is the item genuinely non-recurring? (b) Is it documented and traceable? (c) Would it truly disappear or change under new ownership?
5. Normalize owner and related-party economics to market rates. Replace actual owner compensation with the cost to hire a market-rate executive. Adjust related-party leases to arm’s-length comparable rents.
6. Reject forward-looking addbacks. Projected cost savings, anticipated contracts, and planned synergies are not in the run rate and should not be underwritten as if they are.
7. Calculate LTM, two-year average, and three-year average EBITDA. A widening gap between these periods is a signal that recent performance may not be sustainable.
8. Stress test normalized EBITDA at -15% and -25% revenue scenarios before finalizing leverage and covenant structures. If DSCR falls below 1.0x in a moderate stress scenario, the structure is not sized correctly.
9. Tie normalized EBITDA to free cash flow. Subtract maintenance capex (not growth capex) and working capital requirements to arrive at true cash available for debt service.
10. Document your EBITDA conclusion in the credit memo with an explicit reconciliation from seller-presented adjusted EBITDA to underwriter-normalized EBITDA, and explain every variance.

Key Documentation to Request

• Three years of audited or reviewed financial statements
• Trailing twelve-month management accounts with monthly detail
• Customer concentration schedule (top 10 customers by revenue %)
• Related-party transaction schedule and supporting contracts
• Executive compensation detail with market comparables
• Independent Quality of Earnings report (mandatory for deals above $10M EBITDA)
• Legal matter summary from borrower’s counsel
• Capex history (maintenance vs. growth) for trailing three years

Key Takeaways

• ● The gap between seller-presented Adjusted EBITDA and underwriter-normalized EBITDA is where credit quality is determined. Build your EBITDA bridge independently, never start from the borrower’s schedule.
• ● Normalized EBITDA can be lower than adjusted EBITDA. When new ownership requires incremental costs (replacement executives, infrastructure investment, market-rate leases), the underwriting number must reflect that reality.
• ● Reject post-closing addbacks categorically. Projected synergies, anticipated contracts, and unimplemented cost savings do not belong in the EBITDA used to size debt. This discipline directly distinguishes sound underwriting from credit deterioration risk.
• ● Market data confirms the stakes. With private credit default rates ranging from 1.6% to as high as 6.0% depending on the measurement methodology (and 65% of 2025 defaults driven by distressed restructurings rather than hard payment failures) aggressive EBITDA underwriting is a leading indicator of future workout activity.
• ● Stress testing is not optional. A deal that passes coverage tests at underwritten EBITDA but fails at a 15%–20% revenue decline is not a conservatively structured credit. Run the downside scenarios before you set covenant levels, not after the first breach.