EBITDA Adjustments and Normalization: A Credit Underwriter’s Framework for Separating Sustainable Earnings from Financial Noise

The Underwriting Challenge: When Reported Earnings Mislead

A sponsor presents a compelling opportunity: a $10M EBITDA manufacturing business, asking price of 7.0x, senior debt of 4.5x. On paper, the coverage ratios hold. The leverage is manageable. The deal looks financeable.

Then the diligence team opens the EBITDA bridge.

Buried inside: $1.2M of “one-time” litigation costs that recur every other year. Another $400K in owner travel reclassified as non-recurring. A $600K “pandemic recovery adjustment” still sitting in the model three years later. Pro-forma savings from a vendor renegotiation that hasn't been executed. Strip those out, and the real EBITDA is closer to $7.8M, not $10M. Senior leverage jumps from 4.5x to nearly 5.8x. The deal no longer clears the credit box.

This scenario plays out across middle-market credit desks with uncomfortable regularity. EBITDA normalization is not a formality; it is the foundational analytical act of credit underwriting. Every downstream decision (leverage, coverage, covenant headroom, debt capacity) flows from the number a lender chooses to underwrite. Getting it wrong doesn't just affect one deal; it quietly corrupts the portfolio.

This article provides a practitioner-level framework for building a defensible, independently verified normalized EBITDA, and for spotting the adjustments that should never survive a credit committee.

The Framework: Four-Layer EBITDA Normalization

The underwriting objective is not to arrive at the highest possible EBITDA; it is to arrive at the most accurate, sustainable, and defensible one. Institutional lenders treat the seller's adjusted EBITDA schedule as a starting point, not a conclusion. The credit team must reconstruct the number independently.

A sound normalization process operates across four distinct layers:

Layer 1: Reported EBITDA (The Baseline)

Start with audited or reviewed financials. Pull three full fiscal years plus trailing twelve months (TTM). Calculate EBITDA from the income statement: Net Income + Interest + Taxes + Depreciation + Amortization. Do not accept a management-prepared schedule as your starting point until you have verified it ties back to the trial balance and general ledger line by line.

Key metric: Year-over-year EBITDA variance. If reported EBITDA fluctuates by more than 15–20% across the three-year period without a clear operational explanation, treat the volatility itself as a risk signal, not a normalization opportunity.

Layer 2: Addbacks (What Legitimately Improves the Picture)

Not all addbacks are equal. Institutional credit standards generally accept four categories of addbacks, each requiring documentation:

• Truly Non-Recurring Items: One-time litigation settlements with proof of resolution, terminated lease obligations, costs directly associated with the sale process itself. The test: will this cost appear again in the next 24 months? If there is any reasonable probability it will, it is not non-recurring.
• Owner/Operator Normalization: Excess compensation above fair market replacement cost for the owner's operational role. This requires a formal benchmarking analysis, not a rough guess. Critically, the replacement cost must be modeled in as a cost, not simply removed.
• Non-Cash Charges: Stock-based compensation, non-cash impairments, and certain accounting provisions are standard addbacks, provided they are isolated and verified against the financial statements.
• Implemented Cost Savings: Only those already reflected in the run rate: with documented invoices, signed contracts, or confirmed vendor terms. Anticipated or prospective savings are categorically not a normalization adjustment.

Layer 3: Adddowns (What Legitimately Reduces the Picture)

This is the layer most sellers omit and most junior analysts miss. A credible normalized EBITDA schedule includes both addbacks and adddowns. Adddowns are reductions applied when the reported financials understate ongoing costs:

• Below-market rent paid to a related party (the market rent must be substituted)
• Owner salary below the cost of a market replacement
• Deferred maintenance or capital expenditure that masks true operating costs
• Revenue recognized on a non-standard basis that inflates reported earnings

Failing to apply adddowns produces a materially overstated earnings figure that will not survive post-close reality.

Layer 4: Run-Rate Adjustments (Pro Forma, But Carefully)

Pro forma adjustments for completed acquisitions, full-period revenue from new contract wins, or pricing changes already reflected in invoiced revenue may be appropriate, but each requires a defined evidentiary standard. The question is always: is this already in the business, or is this an aspiration?

The Analysis: Walking Through a $10M EBITDA Manufacturing Credit

Consider a practical example. A sponsor presents a manufacturing business with the following EBITDA bridge:

The credit implications are significant. At a 4.5x senior leverage assumption:

• Sponsor's view: $49.5M senior debt capacity; leverage of 4.5x on $11M EBITDA
• Lender's view: $40.3M senior debt capacity; leverage rises to 5.5x on $8.95M EBITDA against the proposed $49M debt load

A $2.05M EBITDA discrepancy translates to a full turn of leverage misstatement, the difference between a fundable credit and one that breaches the investment policy from day one.

Stress Testing the Normalized EBITDA

After arriving at a lender-verified normalized EBITDA, apply a minimum two-scenario stress test before sizing debt:

• Base Case: Normalized EBITDA held flat. Test debt service coverage ratio (DSCR) at current interest rates. Minimum threshold: 1.25x DSCR.
• Downside Case: Apply a 15–20% EBITDA haircut to simulate a single bad year. At 20% stress on $8.95M: EBITDA falls to $7.16M. Model whether the business still covers interest at the proposed debt load. If it doesn't, the structure needs to be revised, not the EBITDA assumption.

Industry Standards: What Institutional Lenders Are Seeing in 2025–2026

The current credit environment has materially sharpened lender scrutiny of EBITDA adjustments. Several converging data points define the institutional baseline:

Common Red Flags That Fail Credit Committee

Experienced credit committees have developed pattern recognition for the adjustments that signal poor underwriting discipline. The following consistently trigger scrutiny or rejection:

• "One-time" costs that appear in multiple years. If litigation, restructuring, or consulting expenses appear in two of the last three fiscal years, they are recurring by definition.
• High percentage of EBITDA from run-rate or non-cash add-backs. Credit committees increasingly flag deals where a disproportionate share of EBITDA derives from unverified run-rate adjustments, treating that dependence as standalone credit risk rather than financial engineering.
• Pro forma savings from actions not yet taken. Anticipated cost savings that depend on post-closing actions are categorically not normalization adjustments.
• Undocumented owner compensation adjustments. The excess compensation addback only survives when the replacement cost analysis is rigorous and independently benchmarked.
• Missing documentation on key adjustments. Documentation gaps (including missing trial balances, inconsistent revenue recognition, and unexplained balance sheet changes) are consistent red flags in middle-market credit diligence.

Underwriting Checklist: EBITDA Normalization Diligence Points

The following checklist reflects institutional lender standards for EBITDA normalization diligence on a middle-market leveraged credit:

Documentation Requirements

• ☐ Three full fiscal years of audited or reviewed financial statements
• ☐ TTM financials with management-prepared statements and general ledger tie-out
• ☐ Trial balance for the most recent completed fiscal year
• ☐ Detailed general ledger for all flagged addback line items
• ☐ Third-party Quality of Earnings (QoE) report from an independent accounting firm
• ☐ Executive compensation benchmarking analysis (independent source)
• ☐ Supporting invoices, contracts, or settlement agreements for each claimed non-recurring item
• ☐ Related-party transaction schedule with market comparables for all below-market arrangements

Questions to Ask Management and the Sponsor

• ☐ Has this "one-time" cost appeared in any prior fiscal year? What is the litigation or operational history?
• ☐ What is the market cost to replace the owner's operational functions? Has this been benchmarked?
• ☐ Are the claimed run-rate savings already reflected in the TTM P&L? If not, what is the contracted implementation timeline?
• ☐ Are all related-party transactions priced at arm's length? What market comparables were used?
• ☐ Does the revenue recognition methodology conform to GAAP? Are there any side arrangements or deferred revenue items?
• ☐ What is the maintenance capex requirement, and has it been consistently invested? Is the reported D&A a true proxy for economic depreciation?

Analytical Steps

• ☐ Independently reconstruct EBITDA from the trial balance, do not accept the management bridge as a starting point
• ☐ Apply both addbacks and adddowns: test for below-market rent, underpaid owner, and deferred capex
• ☐ Run two stress scenarios: flat EBITDA and 15–20% haircut; test DSCR in each
• ☐ Calculate leverage ratio on lender-adjusted EBITDA, not sponsor-adjusted EBITDA
• ☐ Reconcile EBITDA-to-cash flow: assess working capital requirements, capex intensity, and debt service obligations before declaring a coverage ratio

Key Takeaways

• EBITDA is a starting point, not a conclusion. Institutional lenders independently reconstruct normalized EBITDA from primary financial records. The seller's schedule is a document to be tested, not accepted.
• Every dollar of inflated EBITDA compounds through the entire credit structure. At a 7.0x multiple, a $1M EBITDA overstatement inflates enterprise value by $7M, distorts leverage ratios, and creates covenant headroom that doesn't exist in reality.
• Adddowns are as important as addbacks. A credible normalization analysis adjusts in both directions. Missing adddowns is one of the most common and most consequential errors in junior underwriter analysis.
• Documentation is the underwriting standard. An adjustment without documentation is an assumption. The discipline is simple: if you cannot trace an adjustment to an invoice, contract, legal resolution, or independently benchmarked rate, it does not belong in the credit EBITDA.
• Current credit data demands greater conservatism. With 22% of middle-market borrowers reporting EBITDA declines, 25% operating below 1.0x interest coverage, and private credit default rates rising through 2025, the margin for EBITDA error has narrowed materially. Conservative normalization and rigorous stress testing are not optional; they are the minimum standard of professional credit practice.