S02.8 · Healthcare & Life Sciences
A fragmented, labor-intensive US care segment where Medicaid rate exposure drives returns and AI touches administration, not delivery.
Behavioral health, post-acute and specialty care spans two categories whose combined size has no single defensible figure given differing scope definitions: US behavioral health services at $150-190B, and US post-acute care — home health, skilled nursing and hospice combined — at roughly $250-280B (2025E, composite of Precedence Research and CMS NHE post-acute categories). Forward growth of 6-9% rests on demographic aging and a payer-side push toward lower-cost care settings. But the variable that actually decides returns is narrower than any of that: state-level Medicaid reimbursement rates. The segment remains overwhelmingly a US M&A market for the same reason.
The historical 6-8% growth rate is projected to run 6-9% forward. State-level licensure and Medicaid reimbursement rates produce material regional differences in economics — the reason this is essentially a US phenomenon rather than a global one. Concentration is low. PE-backed roll-ups in behavioral health, home health and hospice each hold single-digit share of their respective categories, and thousands of independent operators remain in the market.
Staffing, real estate and facilities sit upstream; Medicare, Medicaid and commercial payers, plus referring health systems, downstream. Labor dominates the model: clinical staffing runs 60-70% of cost, and scaled operators run 10-20% EBITDA. Medicaid and Medicare rates vary by state and change independent of anything an operator does, which makes rate exposure the primary factor driving returns across the segment — operating performance is necessary but not sufficient.
Agentic AI is compressing the administrative cost line — clinical documentation, staffing and scheduling optimization — in a business where labor is otherwise the dominant cost. The more telling fact is what AI does not touch. Direct-care delivery — a home-health visit, an inpatient behavioral health stay — requires an in-person clinician, and software does not substitute for it the way it substitutes for documentation or scheduling. The core labor-delivery moat holds. The horizon splits accordingly: 2-5 years for administrative automation, with direct-care delivery itself remaining largely AI-resistant near term.
Telehealth and virtual behavioral care bolt-ons add revenue — access extends, and the lower-cost delivery channel carries margin for conditions that do not require in-person care. Multi-state licensure consolidation adds nothing new; it is regulatory scale, letting an operator navigate the patchwork of state Medicaid rules more efficiently rather than a bet on new volume. Home-based primary care is the further adjacency, connecting the segment back to traditional provider economics.