The pro-forma EBITDA framework — the negotiated earnings number.

Every agency deal turns on a single number — the earnings the multiple gets applied to — and that number is not an objective fact. Pro-forma EBITDA is built, line by line, from a base of reported earnings plus a stack of negotiated adjustments. Treat it as a given and you're accepting whatever the seller's accountant produced; build it deliberately and you're pricing the actual earning power of the book. The framework is how you do the second thing.

§ 01 · A negotiated numberWhy recasting matters.

Pro-forma EBITDA is an adjusted-earnings number used for valuation and earnout calculation — and it is explicitly not a GAAP definition. It's a negotiated term, which is exactly why its architecture decides deals. Recasting — restating reported earnings to reflect true owner economics — lifts the number meaningfully: 20–40% on sub-$3M agencies, and 60–80% on $3M+ agencies where an "enterprise illusion" (an owner paying themselves $1M+) masks the real earning power. The lift isn't manufacturing earnings; it's revealing the earnings a new owner would actually keep. And it feeds straight into the multiple through margin bands: an EBITDA margin above 25% is excellent and earns a premium 8–10× multiple; 17–25% is at or above norm and earns a standard multiple; 10–17% is below norm and takes a −0.25× to −0.5× discount; below 10% is significantly below norm, where a revenue multiple may be more appropriate. The recast that turns reported into normalized is the same discipline detailed in normalized EBITDA.

§ 02 · The margin bandsWhere the multiple is set.

The margin bands are the bridge from the earnings number to the price, and they're why the recast matters so much: lifting EBITDA can move a deal from the 10–17% discount band into the 17–25% standard band, changing both the number and the multiple applied to it. The largest add-back driving that lift is almost always owner compensation, normalized through a six-step procedure: document current total owner comp (salary, bonuses, benefits, perks); determine the replacement cost for the role at the agency's size; calculate the excess as current minus replacement; validate the replacement cost against market data; document personal expenses run through the business; and sum the add-back as excess plus perks. The benchmark scales with size — an owner under $500K of revenue typically is the business (no add-back), while an owner above $3M who pays themselves $1M+ supports a roughly $795K add-back. As a quick screen, owner comp above 30% of revenue signals a strong add-back, 20–30% a partial one, 15–20% no adjustment, and below 15% may even need a replacement-cost add-down.

§ 03 · The methodology hierarchyWho holds home-field advantage.

Because pro-forma EBITDA is negotiated, the methodology for calculating it has to be pinned down in the purchase agreement — especially when an earnout will be measured against it later. The right hierarchy is explicit: the historical accounting practices the seller used pre-close come first, GAAP with negotiated deviations second, and generic GAAP last, as the least-favorable fallback. The ordering matters because the party that controls the methodology controls the number, and absent a written hierarchy the buyer holds that advantage by default. Two more adjustments commonly enter the build. Rent-to-market: agency rent benchmarks around 4.9% of revenue, scaling from 5.5% at the smallest agencies to 4.0% at the largest, and a below-market or above-market lease produces an add-back (or add-down) worth $72K–$360K of valuation impact at a 6× multiple. And concentration risk enters through a carrier matrix — acceptable below 25%, elevated at 25–40%, high above 40% — handled with an EBITDA haircut, a −0.25× to −0.5× multiple reduction, or a structure shift toward an earnout tied to carrier retention.

§ 04 · Concentration and loss ratioThe risk overlays.

Two risk overlays finish the number. Carrier concentration runs a three-tier matrix: below 25% is acceptable with no adjustment, 25–40% is elevated (a soft multiple haircut or an earnout gated on carrier retention), and above 40% is high transfer risk — material multiple reduction, a structure shift, carrier-specific reps and indemnification, or a walk. Loss ratio runs its own four tiers: below 40% is excellent (max contingency), 40–50% is good (standard qualification), 50–60% is marginal (reduced or no contingency), and above 60% is poor (carrier scrutiny, contingency at risk). Three levers translate concentration risk into the deal: an EBITDA haircut that subtracts the attributable contingency income, a −0.25× to −0.5× multiple reduction, or a structure shift from cash-at-close toward an earnout tied to carrier retention. Build the pro-forma deliberately — recast owner comp, fix the methodology hierarchy, apply the risk overlays — and you've priced the real earning power of the book, which is the only foundation a sound structure can sit on. The market backdrop that frames how aggressively to structure around these is in the 2026 market context.

◆