DX / Differentials / BEHAVIORAL HEALTH

D-05 — PREPARING FOR SALE · BEHAVIORAL HEALTH

Growth has to read as a system — behavioral health

Growth has to read as a system, not a good year

DIFFERENTIAL — LIKELY CAUSES, MOST LIKELY FIRST

1. 01

   Census growth is not attributed to anything a buyer can diligence

   Growth with no documented demand origin reads as luck or as one relationship. Diligence tests durability, and durability means a named, transferable source for the admissions you booked.

   Self or individual referral accounts for 49.7% of admissions nationally and is the largest single source — larger than every professional source combined. But of 1,625,833 TEDS records, 267,269 had an unknown referral source and 66,540 did not collect it at all. Even the federal dataset loses a substantial share of attribution; a private operator with no CRM discipline loses more.

   SAMHSA TEDS 2023, Table D-1

   NOT THIS IF — Every admission for 24 months carries a named referral source in the system of record and the source-level trend is visible.

2. 02

   You cannot produce the operating table the comparables publish

   Buyers diligence against public comparables because those are the only disclosures that exist. If the two public pure-plays publish a specific same-facility table and you cannot, the gap itself is the finding — regardless of how good the underlying business is.

   UHS publishes, for behavioral health: average licensed beds 24,087, average available beds 23,987, patient days 6,415,058, average daily census 17,575.5, occupancy 73.0% on licensed and 73.3% on available beds, admissions 469,571 and length of stay 13.7 — same-facility and all-facility, with prior-year comparatives. Acadia publishes same-facility revenue +4.9%, patient days +2.1%, admissions +2.3%, ALOS −0.2% and revenue per patient day +2.8% for FY2025.

   Universal Health Services FY2025 Form 10-K; Acadia Healthcare FY2025 Form 10-K

   NOT THIS IF — You already close eight quarters of admissions, patient days, ADC, ALOS, occupancy and revenue per patient day per facility.

3. 03

   Scale, not performance, is the constraint on who will look

   Much of the M&A data operators read excludes deals below a revenue floor. Below it, you are not absent from the market — you are absent from the dataset the market reads, and from the buyer list that reads it.

   Provident Healthcare Partners, which states its methodology as "excludes subscale transactions with less than $5M of revenue," recorded 4 platform, 13 add-on and 5 secondary transactions with 15 unique buyers in Q1 2026 — and zero substance use disorder transactions at that threshold. Mertz Taggart, which applies no revenue floor, counted 5 addiction treatment deals the same quarter, down from 7 in Q4 2025 and 8 in Q1 2025. The Braff Group notes that mental health is dominated by providers with one — maybe two — offices and that the number of multi-office providers with $10–15M or more in revenue is extremely limited; Census County Business Patterns 2023 corroborates it, with 36,309 of 46,513 offices of mental health practitioners under five employees.

   Provident Healthcare Partners, Behavioral Health M&A Market Update Q1 2026; Mertz Taggart, Q1 2026 Behavioral Health M&A Report; The Braff Group, Behavioral Health M&A Year in Review 2023; US Census Bureau, County Business Patterns 2023

   NOT THIS IF — You are comfortably above the $5M revenue threshold with multiple sites — then the question is quality of earnings, not visibility.

4. 04

   Concentration in payer, referral and labour is doing the pricing

   A buyer models the labour line and the payer mix before it models growth. Both are structural in behavioral health, and both are where a strong-looking business gets repriced.

   UHS reported salaries, wages and benefits at 54.0% of behavioral health net revenues in 2025 (53.6% in 2024). Acadia's payer mix is Medicaid 57.7%, commercial 24.6%, Medicare 14.3%, other 3.4%. NABH's tabulation of the 2016 CMS Hospital Cost Report found more than half of the nation's 1,738 inpatient psychiatric facilities had negative net operating margins, averaging −5% (2016 vintage — do not read as current). NABH's 2019 survey of 62 facilities put regulatory compliance cost at just under $1 million per facility, over $18,000 per licensed bed, and $6,747 per 100 days of inpatient psychiatric care.

   Universal Health Services FY2025 Form 10-K; Acadia Healthcare FY2025 Form 10-K; NABH, The High Cost of Compliance (2019)

   NOT THIS IF — Payer mix is diversified, labour is at or below the public comparable's share of net revenue, and no single referral source carries more than a modest share of admissions.

5. 05

   The multiple you are anchored on is probably not real

   Anchoring on a fabricated valuation range sets the wrong preparation priorities and the wrong timeline. In this vertical the most SEO-visible multiple ladders are broker pricing narratives, and two of them carry attributions to firms whose actual documents contain no multiples at all.

   No credible source publishes an EBITDA multiple for residential substance use treatment as a distinct segment. The Braff Group has the deepest sub-segmentation available anywhere — SUD split into residential high-end, residential mid-range-to-value, outpatient PHP/IOP/counseling, MAT, sober living and specialty — and publishes counts only, never multiples. Capstone Partners' Behavioral Healthcare Services Market Update contains no valuation multiples of any kind. Provident's Q2, Q3 and Q4 2025 and Q1 2026 behavioral health newsletters contain none. The detailed ladders in circulation attributing ranges to Capstone and Provident are not in those firms' documents.

   The Braff Group, Behavioral Health M&A Year in Review 2023; Capstone Partners, Behavioral Healthcare Services Market Update, 22 October 2025; Provident Healthcare Partners quarterly behavioral health newsletters, Q2 2025 – Q1 2026 — all read directly

   NOT THIS IF — Your range came from a live process with named comparable transactions and a stated basis. Then it is a real range for your business, which is different from an industry figure.

HOW TO TELL THEM APART

How to tell these apart in your own numbers

Each of these is a measurement you can run yourself, without us.

WHAT RESOLVES EACH

What resolves this, and how you will know it resolved

Census growth is not attributed to anything a buyer can diligence	Rx 05 · preparing for sale →	Attribution is the first pre-exit fix because it takes the longest to accumulate. Two years of named-source admissions data cannot be assembled retroactively in a diligence window.
You cannot produce the operating table the comparables publish	Rx 04 · marketing attribution →	Close the same-facility table monthly and reconcile it to the income statement. The format is already public — the two behavioral health comparables print it in their 10-Ks and it costs nothing to adopt.
Scale, not performance, is the constraint on who will look	Rx 05 · preparing for sale →	If you are sub-scale, the honest answer may be that the right move is another 18 months of same-facility growth or one tuck-in, not a process now. We would say so rather than dress the business for a buyer set that is not looking at it.
Concentration in payer, referral and labour is doing the pricing	Rx 05 · preparing for sale →	Surface concentration in the materials with the mitigation attached. Labour and payer mix are structural in this vertical and a buyer knows it; what damages price is discovering them late.
The multiple you are anchored on is probably not real	Rx 05 · preparing for sale →	There is no honest published residential SUD multiple, so we will not quote one and we would treat anyone who does as a signal about the adviser. What we can do is prepare the disclosure set the real buyers ask for, which is what moves a range in a live process.

HOW THIS DIFFERS BY SCALE

How this differs by scale

Single site	Single site: the buyer is often a local operator or an individual, and the diligence is the founder. The work is transferring the referral relationships onto named staff and out of one person's phone. Note the market floor is real — Praesum was acquired out of a Chapter 11 auction for $18.5 million in the same period that venture money was going elsewhere in behavioral health.
Group	Small group: the same-facility table per site is the deliverable, and site-level variance will be priced — which makes census and length-of-stay reporting a pre-exit exhibit rather than an operating nicety. Segment structure matters — Braff's counts show high-end residential ran 13–18 deals a year in 2014–2016 and 3 in 2023, while outpatient PHP/IOP/counseling hit a record 17 in 2023. Which sub-segment you are in changes who is buying.
Platform	Platform: private-equity-backed platforms have accounted for more than 60% of all behavioral health deal flow since 2018, and 2025 closed with 180 transactions on Mertz Taggart's count against 176 in 2024 — mental health 111, autism/I-DD 36, addiction treatment 33. Note the counters disagree by method (LevinPro counted 104 publicly announced transactions in 2025); pick one series and stay in it rather than blending them in a board deck.

OTHER PRESENTATIONS — BEHAVIORAL HEALTH

• Census is flat while admissions hold
• Cost per admission is climbing
• Inquiries arrive; admissions do not
• Same brand, three different businesses