Quality of earnings: the report that moves the price

I price private companies for a living. More than a thousand of them, on deals up to roughly $26 billion, and for years I did it from inside a Fortune 500 corporate development team, which means I have spent a lot of time on the side of the table that commissions the quality of earnings report rather than the side that receives it. The single most common mistake I see, on both sides, is treating the QoE as an accounting exercise that either passes or fails. It is not. It is a pricing document with an accountant's handwriting, and its entire purpose is to produce one number that gets multiplied.

The evidence that this is where deals die is not anecdotal. Axial's Dead Deal Report, published in January 2026, analyzed 75 lower middle market transactions that fell apart in 2025 across eight firm types and eight industries. Non-QoE diligence findings were the single most common cause at 25.3 percent of broken letters of intent. Quality of earnings EBITDA discrepancies came second at 21.3 percent. Renegotiation failures accounted for a further 14.7 percent, seller decisions 13.3 percent, financing constraints 10.7 percent, and business underperformance 8.0 percent. Roughly one in five signed deals died specifically because the earnings number changed, and close to half died somewhere inside diligence. That is the context for everything below.

A QoE is not an audit, and that is the point

Start with what the document is, because the name misleads people. A quality of earnings study is a financial due diligence engagement, normally performed by a transaction advisory team, normally led by a CPA, and it produces no audit opinion and no assurance of any kind. It is not an attestation that the statements comply with GAAP. Nobody signs off on anything. There is no license legally required to do the work, which is why the market runs from Big Four transaction services groups down to two-person boutiques that do nothing else.

What it does instead is answer a narrower and far more commercial question: how much of the reported profit is repeatable, and how much of it transfers to a new owner? Those are two different tests and a dollar of earnings has to survive both. Revenue that arrived once and will not arrive again fails the first. Profit that exists only because the founder answers the phone at nine at night, or because a related party charges below-market rent, fails the second. An audit asks whether the numbers are stated correctly. A QoE asks whether the numbers are worth anything to you.

On SBA deals, since October 2026, it is not optional

Everything above describes a document buyers chose to commission. On one significant slice of the market that choice is gone. SBA's SOP 50 10 8.1 took effect on 1 October 2026, and on a 7(a) loan financing a business expansion or an initial acquisition priced at $3 million or more, the lender must obtain an independent quality of earnings report in addition to the business valuation it already required.

The threshold is measured on the purchase price "before the application of buyer equity, seller debt, or other financing sources", so structuring the loan down below $3 million does not avoid the requirement. Two transaction types are exempt: owner buyouts and ESOP or cooperative deals, on the reasoning that the existing owners already carry the operational knowledge. There is an anti-avoidance test attached. On an owner buyout, anyone not employed by the business for at least 24 months may acquire less than 50 percent and may not become the largest direct or indirect shareholder, and a transaction that fails that test is processed under the initial acquisition credit standards, quality of earnings included. Styling an outside purchase as a buyout does not avoid the report.

The scope is specified rather than left to the provider. The report has to reconcile four separate records, the accountant-prepared financial statements, the tax return, the internal financial statements and IRS transcript data, into one normalized adjusted earnings figure reflecting recurring, arm's-length operations. It must include a cash proof, which reconstructs cash receipts and disbursements against bank statement data, performed on both a trailing twelve-month basis and the last two fiscal years. If the business has been trading for less than two years, that scope can be reduced.

Two provisions change who can do the work. The report may not be prepared by or for the seller. A seller-commissioned quality of earnings study, which was a reasonable piece of sell-side preparation a year ago, cannot be the document the lender relies on. And where a buyer commissioned their own report before approaching a lender, the lender may not rely on it without a review by one of its own approved vendors, with that review placed in the loan file alongside the original.

Then the number does something. Where a report is required, the lender must calculate debt service coverage from the report's findings, and if that coverage does not support the valuation and the proposed debt structure, the loan amount must be reduced. This is the part worth sitting with. The adjusted earnings figure in the report is no longer advisory. It sets how much money the buyer can borrow.

Below $3 million a business valuation is still required, from a named qualified source: ASA, CBA, ABV, CVA or BCA. Only where the purchase price is $350,000 or less may the lender perform its own valuation instead, and not even then if there is a close relationship between buyer and seller.

It is worth stating what the rule does not do, because a good deal of the early commentary overstated it and some of it is still circulating. It does not require an independent valuation on every change-of-ownership loan at every size, and it does not remove 7(a) Small or SBA Express from acquisition lending, though those loans must still meet the full debt service coverage standard. What it does tighten is where the coverage number comes from: on a change of ownership the ratio "must be satisfied using either the last fiscal year-end or an average of the last two fiscal year-end statements", 1.25 to 1 for an initial acquisition and 1.15 to 1 for a business expansion, so projections do not satisfy it except for an owner-occupied special purpose property whose price is fully supported by its appraisal. The full breakdown of the rule change is here, and here is what SBA altered in the text six days before it took effect.

The one number the whole report exists to produce

Everything in the binder builds toward a single line: adjusted EBITDA in the middle market, or adjusted seller's discretionary earnings at Main Street scale. The report gets there through a bridge, and the bridge is the part worth reading. It starts at reported net income, adds back interest, taxes, depreciation and amortization to reach reported EBITDA, and then applies the adjustments that make up the actual argument: owner compensation normalized to what a market replacement would cost, genuine one-time items removed, timing and cut-off corrected, and costs the business has not been paying but the buyer will added back in.

Here is the thing sellers underestimate and buyers should say out loud. Add-backs are judgment calls, not arithmetic. There is no authority that rules on whether the owner's car is an add-back. Both sides can hold defensible positions on the same line, which is why the schedule of adjustments, not the headline number, is what actually gets negotiated. A seller who hands over a list of add-backs with no supporting documents has not made an argument. They have made a request.

Why a small change to earnings is a large change to price

This is the mechanic that makes the QoE the most expensive document in the process, and it is pure arithmetic. Small companies are priced as a multiple of earnings, so any adjustment to earnings is multiplied before it reaches the price. Across the 7,601 small-business transactions in my Felpel Private Company Index, the median company sold at 2.67 times SDE, with the middle half of deals between 2.0 and 3.4 times. At a three-times multiple, every dollar removed from earnings removes three dollars of enterprise value.

Work an illustrative case. A seller presents $1,000,000 of adjusted SDE and the letter of intent is signed at three times, so $3,000,000. Diligence disallows $150,000 of add-backs that turn out to recur every year, and reclassifies $50,000 of one-time project revenue out of the run rate. Adjusted SDE is now $800,000. At the same multiple the price is $2,400,000. A $200,000 disagreement about earnings became a $600,000 disagreement about price, and $600,000 is enough to break a deal that both sides wanted. That is why 21.3 percent of Axial's broken LOIs died here rather than somewhere cheaper.

The four findings that actually move the number

In my experience the material findings cluster into four groups, and they recur with enough regularity that a seller can go looking for them before a buyer does. Revenue that is not repeatable is the first: one-time projects booked inside the run rate, a large customer already in notice, revenue concentrated in a handful of accounts, or contracts that are terminable at will and were presented as recurring. Axial's report named customer concentration and contract issues explicitly among the non-QoE diligence findings that killed a quarter of its sample.

Add-backs that do not survive are the second, and the argument is nearly always about the word non-recurring. An expense that appears in three consecutive years is a cost of doing business wearing a costume. So is an owner's compensation add-back that assumes the buyer will run a $6 million business without paying anyone to do the owner's job. Timing and accounting policy is the third: cash-basis books restated onto an accrual basis, revenue cut-off at period ends, deferred revenue, and costs capitalized that should have been expensed. This is where cash-basis Main Street companies lose the most, because the restatement itself often moves the earnings number before anyone has argued about a single add-back.

The fourth group is the one buyers care about most and sellers think about least: costs the business has not been paying but the new owner will. Deferred maintenance. A role that has been vacant for a year. Below-market rent from a building the seller happens to own. An owner working sixty hours a week whose replacement is two people. None of these are errors in the financials. They are real, and they come straight out of the earnings a buyer can actually keep.

Working capital: the second price, and nobody negotiated it

A QoE earns its fee a second time in a place most first-time buyers never look. Alongside the earnings bridge it produces a net working capital analysis, and that analysis sets the peg: the level of working capital the seller is required to leave in the business at closing, with a dollar-for-dollar adjustment to the price for any shortfall or surplus. In SRS Acquiom's 2026 M&A Working Capital Purchase Price Adjustment Study, which covers more than 1,500 private-target acquisitions worth more than $385 billion with finalized adjustments, working capital purchase price adjustments now appear in more than 90 percent of private-target transactions, up from about 50 percent a decade ago.

The peg is set from a historical average of net working capital over an agreed look-back period, and the choice of period is not a technicality in a seasonal business. A twelve-month average and a three-month average produce different pegs, and the difference is paid in cash at closing by whoever loses the argument. Sellers who negotiate the multiple for six weeks and accept the peg methodology in an afternoon have optimized the wrong number. Agree the working capital definition and the look-back window in the letter of intent, not in the purchase agreement, because by the purchase agreement you have already spent the leverage.

What it costs and how long it takes

The honest answer at the small end of the market is that public fee data is thin, so I will give you what exists and label it properly. DueDilio, a marketplace that collects competing bids for diligence work, published a median quoted fee of $12,000 with the middle 50 percent of quotes at roughly $9,100 to $15,000, across 17 multi-bid small business and lower middle market projects, and it states plainly that these are proposal prices rather than confirmed engagement fees. Seventeen projects is a small sample and it is a bid sample, not an outcome sample. Treat it as an anchor, not an industry average, and do not let anyone quote it to you as though it were one.

What drives the number up is work, not prestige: multiple legal entities, several accounting or point-of-sale systems, cash-basis books that must be rebuilt before analysis can start, inventory, international activity, and a compressed deadline. On timing, plan on a few weeks from the moment usable data actually exists, which is the condition buyers consistently underestimate. Compare that to the calendar the deal is running on. In the IBBA and M&A Source Market Pulse survey for the second quarter of 2026, which collected responses from 255 business brokers and M&A advisors covering 181 completed transactions, Main Street deals took six to ten months from engagement to close and lower middle market deals eleven to twelve, with the $2 million to $5 million segment stretching to 11.5 months from nine. The QoE is a few weeks of work sitting inside a process that runs for the better part of a year, and it is almost always started too late.

Set the fee against what it is testing. The Stanford Graduate School of Business 2026 Search Fund Study, which tracks more than 850 search funds in the United States and Canada, reports a median purchase price of $16 million for firms acquired in 2024 and 2025, around twenty months to acquire, and an acquisition rate of about half for funds launched between 2021 and 2024 against a 58 percent aggregate rate since 1996. A five-figure fee against a price like that is not the expensive part of the transaction. The expensive part is the deal you close on numbers nobody tested.

Where the QoE meets the price you already agreed

There is one more reason the earnings number tends to fall rather than rise, and it has nothing to do with accounting. The letter of intent is usually priced off the story told at the top of the process, and that story starts from an asking price. In my index, verified-sold deals cleared at about 2.58 times SDE while active listings were asking about 3.26 times, a gap of roughly 26 percent between what sellers ask and what buyers pay across 2,820 verified-sold transactions. The QoE is frequently the first moment in the process where those two numbers are made to meet in a room.

Read that way, a price cut at the end of diligence is not always a buyer being opportunistic. Sometimes it is the market arriving late. The seller who wants to avoid it has one real defense, and it is not arguing harder in week ten.

How to run your side of it

If you are buying, four things. Scope the QoE and issue the data request before you sign the letter of intent, so the work starts on day one of exclusivity instead of week three. Put the delivery date inside the exclusivity period with room to renegotiate afterwards. Ask for the add-back schedule with supporting documents as a condition of the LOI, not as a diligence item, because a seller who cannot produce it in two weeks has told you something. And agree the working capital definition and look-back period in the LOI, where you still have leverage.

If you are selling, the asymmetry is brutal and it is entirely in your favor if you move first. A finding you disclose and support costs you nothing. The identical finding, discovered by the buyer's accountant in week nine of exclusivity, costs you price, then time, then sometimes the deal. Build the add-back schedule yourself, three years, month by month, with a document behind every line. Restate cash-basis books onto an accrual basis before anyone asks. Know your own concentration numbers. Sell-side QoE exists for exactly this reason, and its real product is not the report, it is the fact that nothing in it is a surprise.

The takeaway from the side that commissions them

A quality of earnings report looks like accounting and behaves like pricing. It takes a reported number, subjects it to two tests that have nothing to do with whether the books are correct, and hands back a smaller number that gets multiplied into a price. That is the whole mechanism, and once you see it, the strategy on both sides becomes obvious. Buyers should start it earlier than feels necessary. Sellers should run it on themselves before anyone runs it on them. The deals that die in diligence are almost never killed by a problem. They are killed by a problem that showed up in week nine, when there was no time left to price it and no goodwill left to absorb it.

Common questions

What is a quality of earnings report?

A quality of earnings report, or QoE, is a financial due diligence study that tests how much of a company's reported profit is repeatable and transferable to a new owner. It is not an audit. It carries no opinion and no assurance, and it is not an attestation that the financial statements comply with GAAP. It answers one commercial question instead: what is the sustainable earnings number a buyer should apply a multiple to. The output is an adjusted EBITDA or adjusted SDE figure, a bridge showing every adjustment from reported profit to that figure, a view on revenue quality and customer concentration, and a proposed net working capital target. Because the earnings number is multiplied to reach a price, a QoE is best understood as a price-adjustment document rather than an accounting exercise.

How much does a quality of earnings report cost?

At the small end of the market the honest public benchmark is thin. DueDilio, a marketplace that collects competing bids for diligence work, published a median quoted fee of $12,000 with the middle 50 percent of quotes at roughly $9,100 to $15,000, drawn from 17 multi-bid small business and lower middle market projects, and it labels those as proposal prices rather than confirmed engagement fees. Treat that as a starting anchor and not an industry average, because the sample is small. Fees rise with the things that create work: multiple legal entities, several accounting or point-of-sale systems, cash-basis books that have to be rebuilt on an accrual basis before any analysis can start, inventory, international activity, and a compressed deadline. Set against the Stanford Graduate School of Business 2026 Search Fund Study median purchase price of $16 million for firms acquired in 2024 and 2025, a five-figure QoE fee is small relative to the price it is testing.

How long does a quality of earnings report take?

Plan on a few weeks from the point where usable financial data actually exists, which is the condition most buyers underestimate. DueDilio makes the same point: small business engagements are typically planned around a few weeks once usable data is available, and the timing then depends on scope, data quality, management responsiveness and deal complexity. The clock that matters is not the analyst's, it is the seller's. If the books are on a cash basis, if there is no monthly trial balance, or if the owner answers questions once a week, the calendar stretches and the exclusivity period runs down with it. The practical fix is to scope the QoE and request the data list before signing the letter of intent, so that the work starts on day one of exclusivity rather than week three.

Who prepares a quality of earnings report, and who pays for it?

Transaction advisory or financial due diligence teams prepare them, usually led by a CPA, but the report is not an audit product and the firm issues no audit opinion on it. There is no license or designation that is legally required to perform the work, which is why the market ranges from Big Four transaction services groups to independent boutiques that do nothing else. Historically the buyer commissioned and paid for it, and on a buy-side engagement the report belongs to the buyer and its lender. Sell-side QoE has become common in the other direction: the seller commissions the study before going to market so that the add-back schedule is built and supported on its own timetable rather than under a diligence deadline. Both versions cover the same ground. They differ in who chose the scope and who sees the findings first.

What are quality of earnings adjustments?

They are the additions and subtractions that bridge reported profit to sustainable earnings, and they fall into four groups. Owner and related-party normalization: replacing an owner's actual compensation, personal expenses and below-market related-party rent with what the market would charge a new owner. Non-recurring items: legal settlements, one-off projects, and costs or revenues that genuinely will not repeat, which is the group buyers challenge hardest because an item that appears in three consecutive years is not non-recurring. Timing and accounting policy: cash to accrual restatement, revenue cut-off, deferred revenue, and costs capitalized that should have been expensed. Run-rate and cost-to-replace items: expenses the business has not been paying but the buyer will, such as deferred maintenance, an underpaid or unfilled role, or the second hire needed to replace an owner working sixty hours a week. Add-backs are judgment calls rather than arithmetic, so every one of them should arrive with support attached.

Related reading

• Due diligence: what actually gets checked, from the side that does the checking.
• How I price a company in 48 hours, including the normalized earnings bridge.
• What acquirers really pay for when they buy your company.
• Private markets information asymmetry: where the edge and the risk both live.

Go deeper

If you are under a letter of intent and want an independent read on the earnings number before the QoE lands, or you are preparing to sell and want the add-back schedule built properly first, let's work the numbers together. The underlying transaction data sits in the Felpel Private Company Index.

Tomasz Felpel is an investor, founder, and advisor in private markets and healthcare, based in New York. He is a three-time founder of Value Alpha, an AI-powered private-markets valuation platform, Sonnerie VC, an early-stage healthcare venture firm, and Pond. Previously he led corporate development and M&A at Fortune 500 scale, pricing more than 1,000 private companies. Columbia Business School EMBA. Read the full story. This essay is educational and is not accounting, legal, or investment advice; engage qualified professionals for any transaction.