A quality of earnings report is a financial due diligence study of a private company, commissioned for a transaction and normally prepared by a CPA firm’s transaction advisory group. It answers the one question an audit never asks: how much of last year’s profit will still be there next year, under a different owner. The deliverable is an adjusted EBITDA figure, the schedule showing how that figure was built out of the general ledger, and an honest list of everything the preparer could not verify.
It is not an audit. Nobody signs it under attestation standards, and it carries no opinion. It is nevertheless the number the deal closes on - the figure the buyer multiplies, the figure the bank sizes its loan against, and the figure your working capital peg is calculated from. Sellers who have never seen one tend to treat it as a formality. It is the single most consequential document in the process after the purchase agreement.
What is actually in the report
Scope varies by firm, but the spine is consistent, and it is built around the trailing twelve months rather than the fiscal year. Weaver’s transaction advisory group makes the contrast explicit: an audit focuses on annual results, while a QoE emphasises monthly performance, because buyers and lenders are “particularly interested in the trailing twelve-month results”. Expect five things:
- The EBITDA bridge. Reported net income walked up to reported EBITDA, then to the seller’s adjusted EBITDA, then to the preparer’s. Every line is an adjustment somebody has to defend. Which ones survive is its own subject - the rules are in what counts as an adjusted EBITDA add-back.
- A proof of cash. Bank statements reconciled to recorded revenue, month by month. This is where revenue recognition problems surface, and it is the part owners most often have never had done.
- Net working capital. A monthly trend and a normalised level. The peg in your purchase agreement comes from this chart, and a peg set 10% high is a dollar-for-dollar cut to your proceeds at the post-closing true-up.
- Net debt and debt-like items. Accrued PTO, deferred revenue, customer deposits, unfunded bonuses, capital-lease obligations. Each one typically comes off the price as if it were borrowed money.
- Customer and margin concentration. Revenue and gross margin by customer and by line, usually three years deep, which is where a buyer decides whether your earnings are a business or a relationship.
Why a lender will take a QoE instead of audited financials
This is the part that surprises owners. An audit is the more expensive, more regulated document, and lenders routinely accept the QoE in its place.
You can read the substitution in the loan documents themselves. In the credit agreement Limbach Facility Services signed with PNC Bank on September 9, 2026, guaranteed by its parent Limbach Holdings - a $200 million revolver plus $100 million of term facilities - the conditions on any permitted acquisition require that, at least fifteen days before closing, the lenders receive, “if the annual revenue of the Target exceeds $15,000,000, either the audited financial statements of the Target for its two most recent fiscal years and for any fiscal quarters ended within the fiscal year to date or a quality of earnings report.”
Either. A bank syndicate lending against a target does not insist on an audit; it insists on knowing what the earnings are. Weaver’s practitioners put the reason plainly - “an audit is not designed to assess the sustainability of earnings or the reliability of earnings on a go-forward basis” - and sustainability is the only thing a lender is underwriting. A clean audit confirms that last year’s statements comply with GAAP. It says nothing about whether the owner’s brother-in-law is on payroll, whether the largest customer renewed, or whether the margin came from one unrepeatable job.
The clause that shows what the report is really for
The same credit agreement contains the more revealing provision. When it defines which add-backs may be used to compute Consolidated EBITDA - the number every covenant in the document is measured against - it draws a hard line between adjustments a QoE supports and adjustments the borrower merely asserts.
Restructuring and integration costs, board and officer expenses, and earn-out and retention charges each get a seat at the table, but they share a single ceiling: the aggregate of those three categories “shall not exceed 15% of Consolidated EBITDA for such period.” Anything outside the enumerated list needs “such other add-backs approved by the Required Lenders in their sole discretion” - which is lawyer for no.
Then comes the exception. The agreement separately permits “adjustments in the amounts and for such periods specifically reflected in” two named quality of earnings reports delivered to the administrative agent for two acquisitions, plus “adjustments reasonably acceptable to the Administrative Agent in any other quality of earnings report delivered with respect to Permitted Acquisition.” That category carries no percentage cap at all.
Read the two side by side and the economics of the report fall out of the page. An adjustment you assert is rationed to a shared 15% bucket or left to lender discretion. The identical adjustment, evidenced in a QoE, is uncapped. The report is not paperwork that describes the number; it is the mechanism that makes the number spendable.
What the gap costs: a bridge on an invented company
Take a made-up manufacturer with $34 million of revenue. Its books, and then its diligence, look like this.
| Line | Amount |
|---|---|
| Net income (trailing twelve months) | $2,150,000 |
| + Interest | $420,000 |
| + Taxes | $180,000 |
| + Depreciation and amortisation | $650,000 |
| = Reported EBITDA | $3,400,000 |
| + Owner compensation above market | $450,000 |
| + Litigation settled and closed | $300,000 |
| + Non-operating property carried on the company | $120,000 |
| + Losses on a discontinued product line | $230,000 |
| = Seller’s adjusted EBITDA | $4,500,000 |
| − Accrued vacation never booked | ($85,000) |
| − Discontinued line struck (its revenue left too) | ($230,000) |
| − Repairs capitalised, reclassified to expense | ($140,000) |
| = Diligence-adjusted EBITDA | $4,045,000 |
The gap is $455,000, and none of the three strikes is unreasonable. The vacation accrual is a real liability. The discontinued line cannot be added back as a cost when its revenue is also gone from the base. Capitalised repairs are an accounting choice a buyer is entitled to undo.
Now price it. GF Data’s contributed set put average valuations in the $25M-$50M enterprise value band at 6.8x trailing-twelve-month adjusted EBITDA through the first nine months of 2025. At 6.8x, $455,000 of EBITDA is about $3.1 million of enterprise value. Check the range for your own sector against EBITDA multiples by industry and run your own figure through the business valuation calculator before you accept anyone else’s.
The second effect is larger than it looks and almost never modelled. Senior debt in that same band averaged 3.4x adjusted EBITDA, so the $455,000 also moves borrowing capacity by roughly $1.5 million - and borrowing capacity is what funds the cash portion of the price. In any structure financed against the company’s own cash flow, the diligence number is not an input to the price. It is the cash at close.
It is not one bank’s drafting habit either. Oil-Dri Corporation of America’s revolving credit agreement with BMO Bank N.A., as amended by the Ninth Amendment of October 7, 2026, writes the document into the definition of Consolidated EBITDA outright. Clause (s) of the add-back list reads: “adjustments reflected in any quality of earnings report prepared by a nationally or regionally recognized accounting firm, in connection with any acquisition or other investment permitted under the terms of this Agreement”. No ceiling, no lender discretion, no further test - the report is the qualification. Two unrelated lenders, two drafting styles, the same conclusion about which document they are willing to size debt against.
The same report sets the working capital peg, and that is a second negotiation
The report that produces the EBITDA bridge also produces the net working capital target - the peg. The mechanism is plain enough: the parties agree a number, usually an average of the trailing twelve months, and the balance actually delivered at closing is measured against it. Deliver less and the price falls dollar for dollar.
Here is the trap, and it is the reason the bridge and the peg have to be read together rather than in sequence. Take the $85,000 vacation accrual that diligence struck out of the bridge above. If that accrual is booked into every historical month used to build the peg and into the closing balance sheet, the peg drops by $85,000, the delivered balance drops by $85,000, and the true-up is neutral - the accrual has cost you once, through the multiple, which is what the bridge already showed. If it is booked only into the closing balance sheet, you pay for the same $85,000 twice: once in a price struck on the lower EBITDA, and again in the closing adjustment. At 6.8x that first charge is about $578,000, and the second is the $85,000 in cash.
Nothing in the report flags this, because it is not an accounting question. It is a drafting point in the purchase agreement - whether the peg and the closing statement are built on the same accounting - and the moment to raise it is before signing.
The other half of that argument is which months the average is drawn from. A trailing-twelve-month average and an average of the months actually surrounding the closing date are different numbers in any business with a season, and the gap between them is paid in cash at closing rather than multiplied. The working capital schedule in the databook is what supports arguing for one convention over the other, which is a second reason to have the report before the purchase agreement is drafted rather than after.
It is worth knowing which way this one usually goes. SRS Acquiom’s 2026 M&A Working Capital Purchase Price Adjustment Study analyses “more than 1,570 private-target acquisitions ($385 billion) that closed from 2020 through Q3 of 2025 with a finalized PPA”. Across that set, 91% of private-target deals carried a purchase price adjustment at all, and of the adjustments that were finalised, 51% landed in the buyer’s favour against 38% in the seller’s, with 11% delivering no adjustment either way. The average buyer-favourable adjustment ran to 0.76% of transaction value, the median to 0.27% - and the average one consumed 74% of the escrow set aside for it.
The asymmetry in how those claims get contested is the part to plan around. Of the 57% of deals where the buyer opened with a claim in its own favour, 22% were disputed; of the 34% where the buyer proposed a surplus in the seller’s favour, 6% were. SRS counts a dispute broadly - its own footnote includes the cases where its shareholder-representative team “successfully implements a mathematical or accounting correction (e.g., not a substantive dispute)” - so read 22% as how often the number gets challenged, not how often it turns into a fight. Either way, the seller is the party being asked to disagree after closing, with the money already in escrow.
What the post-closing mechanism will not fix
Owners often read the true-up as a general-purpose correction: if something in the diligence file turns out to be wrong, the independent accountant named in the agreement will sort it out afterwards. Delaware has said otherwise, in a dispute large enough that the reasoning has been borrowed ever since.
In Chicago Bridge & Iron Company N.V. v. Westinghouse Electric Company LLC (Del. No. 573, 2016, decided June 27, 2017, revised June 28, 2017), the buyer took the position that the seller’s historical accounting practices had not complied with GAAP, and the opinion records that it was prepared to ask the independent auditor to make the seller “pay over $2 billion” on that basis. The seller’s own characterisation, quoted in the opinion, was that around $1.93 billion of those claims were really allegations that the represented financial statements were wrong - which is a breach-of-representation claim, not a purchase-price adjustment. The Court of Chancery had read the true-up clause as giving the buyer “a wide-ranging, uncabined right to challenge any accounting principle used by” the seller. The Supreme Court reversed, holding that “the Independent Auditor did not have a wide-ranging brief to adjudicate all disputes between Chicago Bridge and Westinghouse under the Purchase Agreement, but rather one confined to a discrete set of narrow disputes.”
The reading that matters at $3M of EBITDA has nothing to do with the billions. It is that the accounting arguments you intend to win have to be won in diligence, in writing, before the agreement is signed. The narrow mechanism that follows closing is there to measure an agreed definition, not to reopen what the definition should have been. Anything still unsettled at signing gets handled the other way - by an earnout, an escrow or a holdback, every one of which moves the money to the far side of the closing.
Sell-side, buy-side, and who ends up paying
A buy-side QoE is commissioned by the buyer after the letter of intent, scoped to find problems, and shared with the seller only in summary. A sell-side QoE is commissioned by the owner before going to market, scoped to find the same problems while there is still time to fix or explain them.
The cost is real but small against the arithmetic above. Carter Morse & Goodrich, writing in 2020, put the engagement at “between $50,000 to $150,000”, and notes that a QoE typically analyses the last three full fiscal years where an audit “typically cover[s] only the latest fiscal year end.” On the invented company above, the low end of that fee is about 1.6% of the value at stake in a single $455,000 disagreement.
The argument for paying it yourself is sequencing, not optics. Every item in that bridge is fixable in the eighteen months before a process and unfixable afterwards. A vacation accrual can be booked properly for four quarters so it is simply part of the baseline. A discontinued line can be closed cleanly a year early so the comparative periods are honest. Capitalisation policy can be corrected and consistently applied. None of that is available once a buyer’s accountant has already written the finding into a report, and a buyer who discovers three unexpected adjustments rarely stops at re-pricing - they re-open the earnout, the escrow, and the indemnity caps at the same time.
When to commission one
Eighteen to twenty-four months before you intend to run a process, not three. The report is only worth its fee if there is time to act on what it finds, and the adjustments that matter most - compensation normalisation, related-party rent, accrual discipline - need full clean periods in the trailing twelve months to be credible. Commissioning a sell-side QoE six weeks before the data room opens buys you an expensive preview of bad news.
It is also the right moment to establish what the number means to you rather than to the buyer. The same diligence-adjusted EBITDA sets the price in a sale to a private equity fund, the debt capacity in a management buyout, and the valuation your own rollover is struck at. One number, three very different consequences - which is the real reason to know it before anyone else does. How private equity values a company walks through what a fund does with the figure once it has it.
Frequently asked questions
Is a quality of earnings report the same as an audit?
No. An audit tests whether historical statements comply with GAAP and is issued with an opinion under attestation standards. A QoE is a consulting engagement with no opinion, built around trailing-twelve-month and monthly data, that asks whether the earnings are sustainable for a new owner. A company can have a clean audit and a badly damaged QoE.
Do I need an audit if I have a QoE?
Often not, at lower-middle-market size. Credit agreements routinely accept “either the audited financial statements of the Target ... or a quality of earnings report” for acquisitions above a revenue threshold - the Limbach-PNC facility sets that threshold at $15,000,000 of target revenue.
What does a quality of earnings report cost?
Published middle-market ranges run from roughly $50,000 to $150,000, scaling with revenue, entity count and how clean the accounting records are. Smaller lower-middle-market engagements sit at the bottom of that band.
How long does it take?
Plan on several weeks of elapsed time, most of which is you. The preparer’s work is fast once the data is complete; the delay is almost always the general ledger detail, bank statements and contract files the company has to produce.
Will a sell-side report stop the buyer running their own?
No, and it is not meant to. It changes what theirs finds. A buyer’s accountant who confirms a number you have already published is a very different negotiation from one who discovers it.
Does a QoE help if I am not selling the whole company?
It helps more. Any partial transaction - a recapitalisation, a rollover, an internal buyout - prices the equity you keep off the same adjusted EBITDA as the equity you sell. Getting the number wrong costs you twice.
Sources
- Credit Agreement among Limbach Facility Services LLC, Limbach Holdings, Inc. and PNC Bank, National Association, dated September 9, 2026 (SEC EDGAR, Exhibit 10.1)
- SEC EDGAR full-text search, “quality of earnings report” - 389 filings in the twelve months to September 30, 2026
- Weaver, All About Quality of Earnings Reports and How to Read Them
- Carter Morse & Goodrich, Why Quality of Earnings Reports Are Becoming ‘Standard Operating Procedure’
- Withum, Why a Sell-Side Quality of Earnings Report Is Essential in M&A Transactions
- Chicago Bridge & Iron Company N.V. v. Westinghouse Electric Company LLC, No. 573, 2016 (Del. June 27, 2017, revised June 28, 2017) - the opinion itself, read for the quoted passages
- Oil-Dri Corporation of America, Ninth Amendment to Credit Agreement with BMO Bank N.A., dated October 7, 2026 (Exhibit 10.1 to Form 8-K), definition of Consolidated EBITDA, clause (s)
- SRS Acquiom, 2026 M&A Working Capital Purchase Price Adjustment Study - deals closed 2020 through 2025 Q3
- Capital Pad, Lower Middle Market EBITDA Multiples (GF Data valuation and leverage tables)
The bottom line
A quality of earnings report is not an accounting formality and not a smaller audit. It is the document that decides which version of your earnings is real for the purposes of a price, a loan and a covenant - and lenders have written that preference into their contracts, accepting it in place of audited statements and exempting the adjustments it supports from caps that apply to everything else.
The owner’s decision is narrower than it looks: whether the first complete version of your number is produced by someone you hired, with time left to act on it, or by someone the buyer hired, with a signed letter of intent already in the room.
That number matters in every direction, not just towards a sale. An Independent Buyout is financed against the same diligence-adjusted EBITDA a private equity fund would use, so the work you do to defend it is work you keep regardless of which path you take. If your company is doing $3M+ in EBITDA and you want to think through what your number supports, talk to IBO Advisors.
Curious whether an Independent Buyout fits your business?
Learn More