Transaction advisory
Normalized EBITDA starts at the transaction level
Add-backs are only as strong as the data behind them. The ones that survive are the ones you can point to, line by line.
Jaylon Smith · July 28, 2026 · 4 min read
Every seller walks into a process with a list of adjustments: owner compensation above market, the one-time legal settlement, the lease on the building the business does not really need. Every buyer walks in expecting to disallow half of them. The difference between the add-backs that hold and the ones that get negotiated away usually comes down to one question: can you show me?
An adjustment supported by a sentence in a memo is an opinion. An adjustment supported by the underlying transactions, tagged, totaled, and reconciled to the general ledger, is a fact. Buyers' diligence teams know the difference, and they price it.
Building adjustments from the ground up
The method is straightforward and tedious, which is why it is rarely done well. Start with the transaction-level detail for the lookback period. Tag each record that supports a proposed adjustment. Total the tags by period and by category. Reconcile the totals to the ledger accounts they came from. Then write the one-paragraph rationale for each category, with the total and the record count beside it.
Done this way, the adjustment schedule becomes something a buyer can audit in an afternoon rather than argue about for a month. It also tends to surface adjustments the seller had not thought to claim, and occasionally ones they should drop before someone else finds them.
Where the data engineering comes in
Tagging a few thousand transactions by hand is possible. Tagging them consistently across three years and four systems, and keeping the tags current as the close process continues, is not. That is a data problem, and it is solved with the same tools used to build any other reliable dataset: deterministic matching rules, documented exceptions, and a reconciliation that runs every time the data changes.
The result is a normalized earnings figure that moves with the business, holds up under scrutiny, and gives the seller something more valuable than a higher number: a number nobody can take apart.