How EBITDA multiples step up with scale, why the gap exists, and how full sales, equity rollovers, and joint ventures let founders capture more of it.
Practice owners often hear two valuations that seem impossible to reconcile. Their own practice, they are told, is worth a modest multiple of EBITDA. Yet the private equity platforms that acquire practices like theirs are bought and sold at far higher multiples — and large national operators higher still.1,2 Both are real. The gap between them is the single most important economic fact in physical therapy M&A.
This paper explains why the gap exists, how buyers profit from it, and the three ways a seller can participate in the upper rungs of the ladder — along with the risks that accompany each.
| Rung | Typical Profile | Relative Valuation | Basis |
|---|---|---|---|
| 1. Owner-dependent practice | One to a few clinics; owner is the top treating clinician and holds key referral relationships | Lowest rung | Mihama transaction experience; heavy key-person and concentration discounts |
| 2. Professionally managed independent | Multiple clinics, clinic directors, diversified referrals, clean financials | Top of the independent range | Mihama transaction experience; position driven by size, KPIs, and structure |
| 3. Add-on to a platform | Practice acquired by an existing PE-backed or strategic platform | A step above independent practices | Livingstone add-on dataset; published figures vary by dataset and vintage1 |
| 4. Platform, under $200M EV | Regional consolidator with corporate infrastructure and an M&A engine | Materially higher; stepped up since 2020 | Livingstone platform transaction data1 |
| 5. Platform, over $400M EV | Multi-state operator, institutional management, diversified payers | Highest; national operators at the very top | Livingstone platform data and commentary1,2 |
A single clinic can lose a key therapist, a top referring surgeon, or a payer contract and see EBITDA fall 30%. A 100-clinic platform absorbs the same event as noise. Lower volatility means a lower discount rate — and a higher multiple.
Corporate overhead falls as a share of revenue as a platform grows, and the leanest overhead levels are generally reached only at substantial scale.1 Every dollar of overhead a platform avoids on an acquired practice is new EBITDA.
Platforms with balanced growth — a mix of de novo openings and acquisitions — and sustained same-clinic visit growth earn premiums; heavily de novo or heavily acquisition-driven models draw more diligence.1
Larger platforms can be sold to larger sponsors, strategics, or public markets, and they support more acquisition debt. The breadth of the eventual exit market is itself priced into today's multiple.
When a platform valued at a high multiple acquires a practice at a lower one, the acquired EBITDA is re-rated the moment it joins the larger enterprise. That re-rating is the buyer's profit engine — and the reason platforms compete for well-run practices.
Illustrative only. Real arbitrage is reduced by integration costs, financing costs, retention payments, and the risk that acquired EBITDA declines after closing — but the structure explains why a competitive process among platforms can move price within the add-on range.
The arbitrage spread is the buyer's room to pay more. A single bidder in a one-off negotiation keeps the spread for itself. Multiple bidders competing for the same practice transfer part of that spread to the seller — through a higher multiple, better terms, or both. That is the economic case for a run process rather than accepting the first unsolicited offer.
Maximum certainty and liquidity at closing; no exposure to a platform's leverage or future exit. Best suited to owners who plan to step back from ownership and want a defined outcome.
Forgoes any participation in the platform's re-rating. Purchase price may still include earnouts or holdbacks tied to post-closing performance, which carry their own risk.
The seller takes most of the price in cash and reinvests ("rolls") the remainder into equity of the acquiring platform. Example: a seller takes 70% of the purchase price in cash and rolls the remaining 30% into platform equity. If the platform's equity value doubles by its eventual sale, the rolled stake doubles with it.
With proper structuring, rollover gain can often be tax-deferred until the platform exit (see Mihama's tax planning materials).
Rollover equity is a minority, illiquid position in a leveraged company. It may sit behind preferred equity and debt, be diluted by future raises, and carry limited governance rights. It can also lose value if the platform underperforms, overpays for acquisitions, or carries too much debt. Diligence the sponsor, the capital structure, and the operating agreement as carefully as the buyer diligences you.
In a partnership model, the buyer acquires a majority of the practice and the founder retains a minority interest — commonly 20–50% — with a back-end purchase at the original multiple.1 The multiple is fixed at entry in the operating agreement; it does not expand at the second sale. Wealth is created by growing your clinic's EBITDA.
A founder retains 30% with a buyout at the entry multiple. If clinic EBITDA grows from $1.0M to $1.4M, the retained stake is worth 40% more at buyout than at entry, because the same multiple now applies to a larger number. If EBITDA declines, so does the buyout. Unlike platform rollover, the upside is tied to the performance you control — and capped by the locked multiple.
The Key Distinction: Platform rollover participates in the platform's multiple and leverage; a clinic-level JV participates only in your clinic's EBITDA at a predetermined multiple. Neither is inherently better — they suit different owners, ages, and appetites for risk. Mihama models both side by side before a seller chooses.
| Lever | Why Buyers Pay for It | Evidence They Will Ask For |
|---|---|---|
| Management depth | Removes key-person risk; the business runs without the owner on the schedule | Clinic director roles, owner's share of visits, retention agreements |
| Payer contracts | Higher net rate per visit is near-100% margin | Executed contracts, rate tables, renegotiation history |
| Referral diversification | Pipeline survives the owner's departure | 24-month referral report showing no small group of sources dominates |
| Clean QoE and compliance | Defensible EBITDA with no successor-liability surprises | Accrual-basis financials, documentation audits, billing reconciliations |
| Growth engine | Demonstrates repeatable expansion, not one-time luck | Same-clinic visit growth, de novo ramp curves, pipeline of sites |
| Operating KPIs | Shows the platform what it can scale | Twelve-KPI dashboard (see Mihama's Buyer's Scorecard) |
| Outcomes and value-based readiness | Aligns with where Medicare and commercial payers are moving | PROM collection rates, MVP scores, payer or hospital partnerships |
| Revenue diversification | Reduces dependence on fee-for-service reimbursement | Employer contracts, specialty service lines, cash programs |
A consortium of Select Medical's co-founder Robert Ortenzio, executive Martin Jackson, and Welsh, Carson, Anderson & Stowe acquired the company at $16.50 per share, valuing it at approximately $3.9 billion; the deal closed June 30, 2026.3,4 Select operated critical illness recovery and rehabilitation hospitals alongside 1,917 outpatient rehabilitation clinics at year-end 2025.5 Livingstone places Select at a meaningful discount to scaled rehabilitation platforms and attributes much of the gap to its diversified, hospital-heavy mix, including its long-term acute care hospital division.1,2
On August 1, 2025, a consortium of existing stockholders holding over 90% of voting shares, led by Knighthead Capital Management and Marathon Asset Management, completed a merger taking ATI private; remaining shares were acquired for $2.85 in cash.6 A reminder that scale alone does not guarantee equity value — leverage and execution matter.
With Select and ATI private, USPH is the remaining publicly traded pure-play outpatient PT operator.1 It reported 2025 net revenue of $781.0 million and adjusted EBITDA of $95.0 million, and continued to acquire majority stakes in practices while founders retained minority interests.7,8 In 2026 it bought 67% of a 12-clinic practice for $16.4 million and formed two health-system alliances, including one with NYU Langone, together covering 72 clinics; it guides 2026 adjusted EBITDA to $102–$106 million.9
Livingstone expects 2026 to bring platform-level M&A as pandemic-era sponsors seek exits, after several sale processes stalled in diligence in 2024–2025.1 Platforms preparing for sale want clean, accretive add-ons — but the 2027 Medicare cut will tighten underwriting on Medicare-heavy practices.
Reduce the owner's share of visits over 24 months and transfer referral relationships to clinic directors.
Institutional buyers value accrual EBITDA; cash-basis books understate a growing practice.
Rate increases in force before the process are captured as pro forma EBITDA — and multiplied.
Buyers price clinics individually. Show which sites carry the margin and which are ramping.
Find and fix the issues before the buyer's accountants do — and before they become price reductions.
Platforms preparing for exit, hospital systems, and strategics value the same practice differently.
Compare after-tax, risk-adjusted outcomes for each structure before you sign a letter of intent.
Competition is how a seller captures part of the buyer's multiple arbitrage.
Mihama compares full-sale, rollover, and joint-venture offers side by side — modeling cash at close, rollover risk, and long-term value — so you choose the structure, not just the headline multiple.
347-878-2941
Confidential consultation, no retainer required
info@mihamainc.com
We respond within one business day
www.mihamainc.com
Resources, white papers & transaction experience
Valuation tiers reflect published Livingstone Partners datasets, which vary by vintage and methodology, and Mihama's transaction experience for independent practices. Rollover and joint-venture examples are illustrative and assume constant multiples unless stated.