An add-back that gets rejected does not cost you the add-back. It costs you the multiple on top of it.
Disallow $100,000 of claimed adjustments on a business trading at five times EBITDA and $500,000 of enterprise value evaporates. Disallow $300,000 and you have lost more than most owners will spend on professional fees across the entire transaction. This arithmetic is why the adjustment schedule, which looks like a bookkeeping exercise, is one of the highest-stakes documents in a deal.
The thing to understand up front is that adjusted EBITDA is not a number you calculate. It is a number you negotiate. Two competent analysts working from identical books will arrive at different figures, because most of the interesting adjustments involve judgment about what is recurring, what is discretionary, and what a new owner would actually spend. Your job is not to compute the right answer. It is to build a position that survives contact with someone paid to dismantle it.
Adjusted EBITDA is not a GAAP measure and has no standard definition. The SEC requires public companies to reconcile it to net income specifically, not operating income. Private companies face no such rule, which is why the buyer’s analyst ends up doing the regulating.
Source:
SEC Division of Corporation Finance, Non-GAAP Financial Measures Compliance and Disclosure Interpretations
Where the Bridge Starts
Before any adjustment gets debated, there is a question about the starting point that owners frequently get wrong.
Your business probably has three versions of net income. The management P&L that gets reviewed monthly. The tax return, which reflects elections and timing choices made to minimize liability. And the year-end financials the CPA prepared, which may differ from both. These are not the same number, and a bridge built from the wrong one starts with a credibility problem before a single add-back appears.
Buyers generally want the bridge built from book financials, with a separate reconciliation to the tax returns. If your books do not tie to your returns, that reconciliation becomes the first argument of the diligence process rather than a footnote.
The period question, and why monthly data is not optional
The other structural decision is which twelve months get measured.
Fiscal year figures are clean but stale. A buyer signing in September does not want to underwrite a period that ended nine months ago, so they will ask for the last twelve months through the most recent closed month. That is an LTM figure, and building it requires monthly financial statements prepared consistently across at least three years.
If your books close annually, you cannot produce an LTM bridge without reconstructing the monthly detail. That reconstruction is doable and it is also exactly the kind of prepared-under-pressure work that gets discounted relative to contemporaneous records.
There is a second wrinkle in seasonal businesses. LTM captures a full seasonal cycle, which is fair, but if your strongest quarter sits at the front of the LTM window and your weakest at the back, the trend line reads as decline even when nothing has deteriorated. Anticipating that reading and preparing the seasonality analysis to explain it is a small piece of work that prevents a bad conversation.
The Three Tests Every Adjustment Has to Pass
Analysts do not evaluate add-backs by intuition. Each one gets run through three questions, and knowing them tells you exactly what your documentation has to accomplish.
The frequency test asks whether the expense actually recurs. An item claimed as non-recurring that appears in multiple years is not non-recurring, whatever it is called on the schedule.
SEC rules bar public companies from labeling a charge non-recurring if a similar charge occurred in the prior two years, or if it is reasonably likely to recur within two years. Private buyers apply the same logic informally, which is why analysts pull three years of adjustments.
Source:
SEC, Conditions for Use of Non-GAAP Financial Measures
The nature test asks whether the expense relates to operating the business. Owner perks, personal costs, and expenses tied to assets the buyer is not acquiring fail this test in the seller’s favor. Costs a new owner would still incur do not.
The evidence test asks whether the claim can be traced to something outside your own assertion. An invoice, a contract, a payroll register, a settlement agreement, a salary benchmark.
An adjustment that clears all three is usually accepted without much conversation. Clear two and you are in a negotiation. Clear one and it is gone. The practical consequence is that documentation is not the last step in preparing a schedule, it is the thing that determines whether an item belongs on the schedule at all.
Which Add-Backs Actually Survive
Adjustments sort into fairly predictable groups once you have seen enough of them.
The ones that generally clear
Owner compensation above market rate, provided the market rate is benchmarked rather than asserted. More on the methodology below, because this one is more involved than it looks.
Transaction and one-time professional fees, supported by engagement letters that establish the scope and the non-recurring nature of the work. Investment banking fees, deal legal, a one-time systems implementation.
Legal settlements with the settlement agreement attached, where the matter is resolved and not part of a pattern.
Discontinued operations and closed locations, when the cut is clean and you can isolate both the revenue and the full cost structure that departed with them.
Costs associated with assets the buyer is not acquiring. The aircraft, the boat, the lake property, the vehicles that are going with you. These clear easily because they self-evidently do not transfer.
Related-party rent above market, when supported by a broker opinion or comparable lease data rather than your estimate of what the space is worth.
The ones that get haircut
Personal expenses running through the business almost always survive partially and rarely survive in full. You claim $80,000 of owner discretionary spend, the analyst can substantiate $52,000 from receipts and statements, and the difference disappears. The gap is not a dispute about principle. It is a dispute about evidence, and it resolves against whoever failed to produce it.
Family members on payroll get scrutinized in proportion to how much work they actually do. A spouse who genuinely does not work in the business is a clean add-back with the payroll register and an org chart. A relative who does twenty hours a week of real work gets normalized to market rate for those hours, not added back entirely.
Discretionary bonuses invite a straightforward question: if the business paid them in each of the last three years, on what basis are they discretionary. Sometimes there is a good answer. It needs to be prepared.
Travel and entertainment sits permanently in the contested zone, because some of it genuinely is business development that a new owner would continue and some of it is not, and the receipts rarely make the distinction obvious.
The ones that almost never survive
Pro forma synergies. Cost savings the buyer will achieve are the buyer’s value to capture, not yours to sell, and no analyst will let you charge for them.
Run-rate adjustments for contracts signed but not yet performed. This is a forecast wearing an adjustment’s clothing, and forecasts do not get multiplied.
Lost revenue from a bad year, a departed customer, or an external event. The instinct is understandable and the answer is consistent: the revenue did not happen, and buyers underwrite what happened.
Owner opportunity cost, meaning the value of work you performed without compensating yourself, beyond the market-rate normalization already discussed. This gets claimed more often than you would expect and it fails every time.
Future cost reductions you have identified but not executed. If you know a vendor renegotiation will save $60,000, execute it, run it for two quarters, and let it show up in the actual numbers. Claiming it as an adjustment converts a real improvement into a rejected line item.
The One-Time Expense That Happens Every Year
Here is a specific pattern that damages sellers more than any single rejected item, and it is entirely avoidable.
Analysts do not evaluate the adjustment schedule for one period. They lay adjustments out across three years and look at the categories. When “one-time legal fees” appears in 2023, 2024, and 2025, the conclusion is not that you had three unlucky years. The conclusion is that legal expense is part of your run rate and someone has been carving it out of earnings annually.
The damage extends past that one line. An analyst who catches this expands the sample. Items that would have passed on a spot check now get tested individually, and the credibility of the entire schedule is in question because the person who assembled it has demonstrated a pattern.
The defense is uncomfortable and effective: run the three-year view yourself before anyone else does. Where a category appears in multiple years, either drop it from the schedule or address it directly in the supporting memo. If you had genuinely unrelated legal matters in consecutive years, say so and identify each one, since three distinct matters with three sets of documentation is a very different story from a recurring cost being relabeled. Calling out your own pattern costs you the add-back. Having it discovered costs you the benefit of the doubt on everything else.
Owner Compensation Is More Complicated Than Adding It Back
This is the largest adjustment in most owner-operated businesses and the one most often constructed incorrectly.
The wrong version adds back the owner’s entire compensation. That is the SDE calculation, discussed below, and it produces a number a private equity buyer will not use. The correct version for adjusted EBITDA normalizes compensation to what the business would pay someone at market rate to do the job.
Which raises the question the methodology actually turns on: what job. Owners of businesses in this size range typically perform several. They are the chief executive, and also the head of sales, and often the person approving every disbursement. A buyer will not accept a single general manager salary as the replacement cost for three functions, and you should not propose one, because proposing an obviously inadequate replacement cost invites the analyst to build their own estimate, and theirs will not be generous.
Build it honestly. List the functions you perform and the approximate time in each. Price each function at market and cite the source you used.
The Bureau of Labor Statistics publishes Occupational Employment and Wage Statistics with wage percentiles by occupation and metropolitan area, updated annually and free to access. A public federal dataset is a considerably more defensible compensation benchmark than an estimate.
Source:
U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics
Published salary guides from staffing firms provide a useful second reference point, and industry association compensation surveys are stronger still where they exist for your sector.
The ones that almost never survive
If the owner has been underpaying themselves, and many do for tax reasons, normalization runs the other direction. Market rate exceeds actual compensation, the difference becomes an additional expense, and adjusted EBITDA goes down.
Sellers almost never volunteer this. Buyers find it every time, because it is one of the first things a competent analyst checks. Discovering it yourself and presenting it in your own schedule is worth more than the dollars it costs, because a schedule containing an adjustment against your own interest reads as an honest document, and honest documents get less scrutiny elsewhere.
The Adjustments Buyers Bring That You Did Not
Which points at the structural asymmetry in how most sellers prepare.
You arrive with a list of items that increase EBITDA. The buyer’s team arrives with a list that decreases it. Only one party built both lists, and it is not usually the seller.
Deferred maintenance and underinvestment come first. If capital expenditure has run below depreciation for several years, the analyst will normalize toward a sustainable level, because the business has been earning partly by not spending what it needed to spend. Equipment-heavy businesses are especially exposed here.
Below-market related-party rent works exactly like the above-market version, in reverse. If you own the building and charge the business a favorable rate, the buyer normalizes rent up to market and EBITDA down accordingly. Sellers who add back above-market rent while staying quiet about below-market rent are making an argument that only points one direction, and analysts notice.
Unrecorded liabilities matter here too. Accrued paid time off that was never booked, warranty obligations, deferred compensation. These affect both the earnings picture and the closing balance sheet.
Understaffing is the subtlest one. If the business runs without a controller because the owner does that work, or without a safety manager because nobody got around to it, the buyer prices in what the position costs. The same logic covers insurance coverage that is thinner than a professional buyer would carry and compliance work that has been deferred.
None of this is a reason to despair. It is a reason to build the negative side of the bridge yourself, because a seller who presents both directions is negotiating from a defensible position, and a seller who presents only the favorable direction is waiting to be corrected.
SDE and Adjusted EBITDA Are Not the Same Number
A definitional confusion that costs owners real money, particularly in the middle of the size range where the metric itself is in play.
Seller’s discretionary earnings adds back one working owner’s full compensation and benefits, because the buyer is an individual who will take that job and that pay. Adjusted EBITDA normalizes owner compensation to market rate, because the buyer is an institution that will hire someone to do it.
The same business produces two different numbers, and the gap is roughly the market-rate salary. It also carries two different multiple ranges, with SDE multiples running lower precisely because the earnings figure is larger.
The transition between them sits somewhere between $1M and $2M of earnings, and sources genuinely disagree about where. Below roughly $1M, SDE dominates and the buyer is typically an individual, often with SBA financing. Above roughly $2M, adjusted EBITDA takes over as private equity and strategic buyers enter. Inside that band, expect buyers to run both calculations and frame the conversation around whichever suits them.
The practical error to avoid is comparing an SDE-based expectation against an EBITDA-based offer, or the reverse. An owner who has been told their business is worth three and a half times SDE and then receives an offer at five times adjusted EBITDA may be looking at a better deal or a worse one, and the multiple alone will not tell them which. Run both, know both, and be clear which one any given number refers to.
Building the Schedule
The format matters more than it seems, because a well-built schedule answers questions before they get asked and a poorly built one generates them.
Structure it with a row per adjustment and columns for the description, the category, the amount in each of the last three fiscal years plus the LTM period, the general ledger account it flows through, the supporting document reference, a written rationale, and a flag for whether the item appears in more than one year.
The three-year layout is the important part, and it is the opposite of what instinct suggests. Instinct says to show only the current period so that recurring items are less visible. The analyst will build the three-year view anyway. Showing it yourself means you frame the pattern rather than defend it.
Attach the support rather than referencing it. A schedule where every line links to the underlying document is a different artifact from a list of numbers with a note saying support is available on request. The first one gets reviewed. The second one gets tested.
On volume, fewer and stronger beats more and weaker. Six well-documented adjustments totaling $400,000 will fare better than thirty adjustments totaling $450,000 where the long tail is unsupported. Analysts sample, and a sample that turns up two failures in the small items produces a much more thorough review of the large ones. Every marginal add-back you claim without support puts the well-supported ones at risk.
The Bottom Line
Adjusted EBITDA is a negotiated figure, and the negotiation is settled by evidence rather than argument.
Every adjustment you claim is a proposition someone will test against frequency, nature, and documentation. The ones that clear all three become part of the valuation. The ones that fail cost you the item and the multiple on it, and enough failures cost you credibility on the items that should have passed.
The preparation is unglamorous and it is not complicated. Build the bridge from books that tie to your returns. Produce monthly data so an LTM figure is available on request. Benchmark owner compensation properly against the functions actually performed. Run the three-year view on your own adjustments and address the patterns before someone else finds them. Build the buyer’s side of the bridge alongside your own.
Do that and the adjustment schedule stops being the document you defend and becomes the one that establishes you are worth believing.
Where to Go From Here
The judgment calls are what make this difficult. Which adjustments are worth claiming, what documentation will actually satisfy an analyst, how to benchmark compensation defensibly, and which unfavorable adjustments to surface yourself rather than wait for.
At Frak Finance, building defensible adjustment schedules and normalized earnings bridges is core transaction work, whether you are two years from a sale or already under a letter of intent.
Schedule a free consultation and we’ll look at what your adjusted EBITDA can actually support.ng.
