Module 4: Quality of Earnings (QoE) Self-Defense: Defending and Haircutting Add-Backs

Learning objectives and prerequisites

By the end of this lesson, you should be able to challenge adjustment evidence and translate findings into cash risk. Prerequisite knowledge: The normalized EBITDA bridge. Work from assumptions to equations, explain the acquisition consequence, and record the evidence needed to rely on the answer.

What quality of earnings actually asks

Quality of earnings analysis examines whether reported performance is supported, correctly measured, recurring, and useful for forecasting a transferable business. It is broader than checking arithmetic and narrower than declaring that every business risk has been resolved. A QoE review does not automatically constitute an audit, a tax opinion, or a guarantee of future results; its reliability depends on the scope, evidence, and procedures actually performed.

A buyer approaching Vanguard Industrial Machining already has a $616,000 adjusted EBITDA bridge. The next question is whether that bridge survives challenge. The company's customers may be concentrated, the owner may work more hours than disclosed, and repair spending may have been postponed. A mathematically correct normalization schedule can still underwrite the wrong operation.

This chapter uses actual US dollars and the fictional recurring assumptions from Module 3. It develops a defensible process for accepting, reducing, rejecting, or conditioning claims. The objective is not to win every disagreement with the seller. It is to price and finance a business using earnings that withstand independent scrutiny.

Why both seller and buyer need disciplined skepticism

Owners accustomed to minimizing taxable profit may describe many expenses as discretionary when marketing the company. During a sale, higher adjusted earnings can increase value by several times the expense being discussed. The incentives are clear even when the owner sincerely believes their explanation.

Buyers have incentives too. An acquirer who has invested months in a transaction can become emotionally committed to closing and accept weak add-backs to make debt capacity work. Alternatively, a buyer may reject every adjustment as a bargaining tactic even when evidence is strong. Neither behavior produces a reliable operating forecast.

The appropriate response is symmetrical evidence. A verified personal expense that disappears can be accepted. A market-rent increase that reduces profit should also be accepted. Proposed revenue growth remains a forecast assumption unless already supported as a completed run-rate change. An accounting error should be corrected regardless of whose preferred price it supports.

Document the reasoning before negotiating the final number. A buyer can offer less than indicated value because of alternatives or risk, but should not relabel a valid expense to disguise that commercial decision.

The four questions behind every adjustment

First, determine whether the item reduced the starting earnings measure in the relevant period. A capitalized truck purchase does not reduce EBITDA in the same way as fuel expense. A shareholder distribution does not reduce operating profit at all. This test prevents duplicate benefits before the analyst debates recurrence.

Second, establish the amount from source records. An owner's estimate of “about $25,000” is not equivalent to invoices totaling $25,000. Trace material claims to ledger postings and confirm that credits, reimbursements, and reversals have been considered. Gross invoices can exceed the net expense that actually remained in profit.

Third, determine whether the cost disappears under the contemplated operating arrangement. A personal vacation may disappear; a customer visit may continue. Owner wages disappear only together with an appropriate replacement staffing plan. A cheap related-party lease may become more expensive rather than vanish.

Fourth, identify any substitute cash cost, investment, liability, or operating risk. A repair expense removed from EBITDA can reappear as capital replacement. A departing owner can require recruitment and transition payments. The bridge should show those consequences rather than treating every accepted add-back as unrestricted distributable cash.

An evidence hierarchy that remains practical

Evidence quality varies. A detailed ledger trace establishes where an expense was recorded. An invoice establishes what was purchased. Payment records support settlement. Contracts, payroll records, operating interviews, and external replacement quotes help establish recurrence and future economics. No single document answers all four adjustment questions.

For Vanguard's $300,000 officer compensation, payroll detail supports the historical amount, while the $180,000 replacement assumption needs a role description and credible hiring evidence. For personal travel, receipts alone may establish cost but not business purpose. For market rent, a broker's generic rate can be less useful than the actual lease available for the required facility.

Absence of documentation is not proof that an item is false, but it constrains reliance. A buyer can hold it as conditional, exclude it from the underwriting base, or require resolution before final pricing. The treatment should follow materiality and decision impact rather than a mechanical rule that all missing invoices receive the same percentage discount.

Claim component	Evidence supporting amount	Evidence supporting treatment
Owner salary	Payroll and ledger tie-out	Role mapping and replacement compensation
Personal truck expense	Receipts, ledger, payment detail	Usage evidence and replacement need
Market rent	Existing lease and ledger	Executable replacement terms and occupancy comparison
Exceptional repairs	Work order, invoice, expense posting	Multi-year maintenance history and condition review
Inventory correction	Subledger and count reconciliation	Ownership, condition, recoverability, and period of origin

Haircuts: reduce a claim for a reason

A haircut is a reduction in an asserted adjustment or earnings amount. It should describe an evidence or economic conclusion, not conceal uncertainty behind a round percentage. If only $18,000 of a $24,000 vehicle claim is supported and $6,000 of replacement transport is required, the accepted net adjustment is $12,000. Calling this a 50% haircut is mathematically true but analytically incomplete.

The remaining $12,000 is not all the same kind of rejection. Six thousand lacks support as removable expense, while another six thousand is replaced by an identified recurring cost. Those categories imply different follow-up work. Additional receipts might resolve the first; they cannot eliminate the second if the business truly needs transportation.

Similarly, a repair claim can be reduced because the amount was misreported, because the event is recurring, or because capital investment replaces the expense. Different reasons produce different model effects. Preserve the original claim, accepted amount, rejected amount, conditional amount, and explanation.

A conditional claim should not quietly flow into lender coverage as if verified. Show a conservative case excluding it and a documented upside case including it. Scenario treatment makes the dependence transparent without pretending that a negotiated discount creates accounting certainty.

Reunderwriting the Vanguard bridge

Assume the initial Module 3 case supports $616,000 adjusted EBITDA. In a separate diligence sensitivity, new evidence changes four assumptions. The replacement manager costs $200,000 rather than $180,000. Only $18,000 of truck expense is supported, and required replacement transport costs $6,000. Only $8,000 of the $12,000 travel claim is clearly personal. Comparable lease terms require $165,000 occupancy cost rather than $150,000.

The exceptional repair adjustment remains $20,000, and the $40,000 inventory correction remains valid. These assumptions produce the following revised bridge. They are a challenge scenario, not changes to the base packet that Modules 5 and 6 will use unless explicitly requested.

Item	Original accepted effect	Revised accepted effect	Change
Correct inventory	−$40,000	−$40,000	$0
Remove owner compensation	+$300,000	+$300,000	$0
Replacement management	−$180,000	−$200,000	−$20,000
Net truck adjustment	+$24,000	+$12,000	−$12,000
Travel adjustment	+$12,000	+$8,000	−$4,000
Rent normalization	−$60,000	−$75,000	−$15,000
Excess repair adjustment	+$20,000	+$20,000	$0
Total net adjustment to $540,000	+$76,000	+$25,000	−$51,000
Adjusted EBITDA	$616,000	$565,000	−$51,000

At an illustrative 4.0x multiple, the $51,000 earnings reduction changes EV by $204,000. That amplification explains why the buyer should investigate replacement costs carefully. It also shows why negotiations based solely on a multiple miss important changes in the denominator.

Revenue quality beyond the add-back schedule

Even a perfect expense bridge cannot rescue unsupported sales. Review revenue by month, customer, product family, and job. Compare sales growth with receivables, credit notes, returns, delivery evidence, and gross profit. A final-quarter revenue surge paired with slow collections can reflect timing errors, aggressive invoicing, or a change in customer terms rather than genuine commercial momentum.

Suppose Vanguard records $200,000 of December sales to its largest customer but collects only $20,000 by March. Determine whether the parts were delivered and accepted, whether payment terms changed, and whether a dispute exists. Do not automatically remove the entire $180,000 unpaid amount from EBITDA: a collectible timing delay and an unearned sale require different treatment.

Customer concentration should be measured in revenue and contribution profit. A customer representing 20% of sales but 35% of contribution profit creates a larger earnings exposure than revenue share suggests. A low-margin customer might absorb working capital without supporting much overhead. Those relationships belong in the operating forecast and downside case, not only in a generic discount-rate premium.

Expense quality and the missing-cost problem

An owner preparing to sell can preserve current earnings by leaving positions vacant, delaying preventive maintenance, reducing insurance, or postponing software replacement. Those actions may be legal and visible in the accounts, yet still make current earnings unsustainable. QoE asks which operating capacity the buyer receives for the reported expense base.

At Vanguard, an unfilled quality-control role might save $70,000 this year while increasing rework and customer risk. If the replacement-manager budget does not include that function, normalized earnings need a negative adjustment. Verify that the cost is not already included in direct labor or another department before deducting it.

Inspect vendor bills and post-period payments for unrecorded liabilities. A missing $25,000 expense can overstate profit and understate payables together. Correct both sides; do not reduce EBITDA while leaving an artificially low liability balance in the opening model. If the liability is also classified as debt-like for pricing, coordinate with the working-capital definition to avoid overlapping adjustments.

Seller Tricks & Red Flags: temporary austerity presented as permanent efficiency.

A short period of low repairs, overtime, or headcount may represent deferred work rather than a structural improvement. Ask how the company will deliver the forecast sales with the proposed people, equipment, and spending.

Historical correction versus forecast improvement

A signed price increase effective before closing can support a run-rate analysis, but annualizing it requires retention and volume assumptions. A proposed future price increase is an initiative. A buyer's ability to combine purchasing with another portfolio company is a synergy. Neither should be inserted into corrected historical EBITDA without clear labeling.

Show separate columns for reported results, accounting corrections, recurring normalization, implemented run-rate changes, and buyer initiatives. This format does not prohibit underwriting improvements. It forces the buyer to identify which portion of the investment depends on execution rather than acquired earnings.

Seasonality also affects annualization. Three profitable summer months do not necessarily represent a full year. Compare corresponding periods and use monthly customer and cost data. A growth forecast that annualizes peak sales while retaining average staffing expense creates a mismatch even if every individual input appears defensible.

Earnings quality must reconcile to cash quality

Suppose revised EBITDA is $565,000, cash taxes are $90,000, maintenance CapEx is $140,000, and NWC investment is $85,000. Cash available before financing under this simplified buyer bridge is $250,000. Scheduled cash debt service of $260,000 would create a $10,000 deficit before distributions, growth investment, or additional reserves.

That deficit cannot be solved by asserting that EBITDA is strong. The business may need lower debt, more equity, a lower price, better collection terms, or a revised operating plan. An annual model should also be paired with short-term liquidity because a business can meet annual service while running out of cash during a seasonal inventory build.

Inventory corrections are particularly instructive. Recognizing a noncash $40,000 loss reduces earnings, but the cash loss may have occurred when materials were purchased. Use a consistent gross-versus-net working-capital bridge so that the same write-down does not create another $40,000 cash outflow in the forecast.

Translate findings into decisions rather than accusations

A strong findings memo states the issue, supported amount, affected period, evidence, model treatment, and unresolved consequence. “Seller inflated EBITDA” is less useful than “the proposed $24,000 truck adjustment includes $6,000 without support and requires $6,000 replacement transport; accepted net adjustment is $12,000.” The second statement can be reviewed and acted upon.

Some findings require corrections before price is finalized. Others require contract protection, a working-capital adjustment, transition obligations, or an explicit buyer risk decision. An escrow can secure a particular exposure, but it cannot transform recurring low margins into higher sustainable earnings. A seller note changes financing and risk allocation; it does not prove the income statement.

QoE should end with a range of reliance, not just a single impressive number. The buyer should know verified recurring earnings, conditional claims, downside cash generation, and the remaining evidence necessary for approval. The purchase decision then becomes a controlled judgment rather than a bet on the most optimistic spreadsheet.

Self-Study Problem Set

Problem 1: construct a reasoned haircut

A seller requests a $50,000 consulting add-back. Invoices establish $42,000 actually reduced EBITDA; the rest was capitalized. Of the expense, $12,000 recurs annually, and eliminating the consultant requires $5,000 of internal software spending. Assume that spending is recurring operating expense. Find the accepted net adjustment.