Market-Selection Intelligence

Thousands of public federal records, resolved into a market-by-market read of where to grow, who to acquire, and who already organizes risk — generated from the answer tree, every figure reproducible against the lake. 100% public · zero PHI.

Conventions · reconciled to the 6/6 brief

Two denominators are named so nothing reads as a contradiction: MA % is reported on CMS’s published ma_prtcptn_rate and the computed benes_ma ÷ benes_total (same lives, different base); group universe is the PCP-led count (primary specialty = primary care), distinct from the broader “≥1 PCP” count.

The market-selection layer · outside-in

Where to grow, scored on public data

Every U.S. county (≥40k Medicare benes) scored on one transparent composite — PCP supply + low VBC penetration + low REACH competition + RAF headroom, equal-weight percentiles, no hidden weights — across 368 sizable markets. Cut to your core states (NY/PA/OH/NC):

State	Market	Medicare benes	PCPs	VBC density	REACH ACOs	RAF headroom	Target score	Natl rank	Top competitor present
NY	NY-New York	305,603	8,193	0.78	34	+0.028	3.016	27	CareMount Value Partners IPA
NC	NC-Durham	56,158	928	0.81	9	+0.011	2.834	45	Vytalize Health REACH, LLC
NY	NY-Monroe	177,779	2,686	0.86	1	+0.037	2.801	50	None
NY	NY-Broome	47,833	547	0.84	3	+0.052	2.717	63	agilon health Northeastern ACO, Inc. (DBA: Senior Health Connect ACO 6)
PA	PA-Dauphin	64,363	961	0.82	3	+0.003	2.619	77	D0115
PA	PA-Lackawanna	56,912	545	0.83	2	+0.023	2.613	78	Pearl Primary Care Network, LLC
NC	NC-Buncombe	68,879	1,032	0.83	12	+0.041	2.608	79	D0245
PA	PA-Chester	113,969	1,146	0.81	11	-0.004	2.474	99	Pearl Primary Care Network, LLC
PA	PA-Delaware	124,259	1,209	0.78	14	-0.022	2.471	100	American Choice Healthcare, LLC
NC	NC-Cumberland	62,127	821	0.78	12	-0.047	2.447	103	D0245
PA	PA-Lehigh	82,973	1,732	0.83	6	-0.056	2.357	116	D0045
PA	PA-Philadelphia	274,655	4,954	0.79	18	-0.049	2.322	122	Temple Care, Inc

Live national percentiles over 368 counties; CMS Geographic Variation PY2024, MSSP County RAF PY2024, ACO REACH PY2023, NPPES. Reproduce any row on demand.

The competitive lens

Who organizes risk — including you

The same federal footprint names who else contracts risk in each market and audits your own positions in the same pass. In NY-Broome the dominant REACH entity is your own agilon health Northeastern ACO, Inc. (DBA: Senior Health Connect ACO 6) — surfaced from competitors’ CMS filings, not your claims. That is the market map an internal RAF pipeline cannot produce. Live

The acquisition layer · independence-screened

PCP-led targets, ownership- and integrity-screened

171

NC independent
· 33 captured

487

NY independent
· 30 captured

144

OH independent
· 27 captured

208

PA independent
· 24 captured

1,010

Independent candidates
across core states

PCP-led groups, ownership-screened (M9 named-owner overlay) and integrity-screened (OIG LEIE). Live

State	Group	PCPs	LEIE flags
OH	Central Ohio Primary Care Physicians, Inc.	491	0
NY	Westchester Medical Group, P.C	409	2
NY	Interfaith Professional Physician Services PC	388	1
OH	MVHE Inc	386	0
OH	Alliance Physicians Inc	315	0
NY	Lawrence Medical Associates PC	231	1
NY	New YORK Medical Physician Associates PC	213	0
NY	TJH Medical Services PC	212	0

Boundary (a selling point): “independent candidate” = no ownership tie on the federal record — not confirmed independent. The federal ownership graph is facility-scoped (only 7 of 1,598 core-state PCP-led groups appear in it), so confirmed independence is where your records turn our read into proprietary truth (§7).

Risk-documentation headroom

Directional — validation only

Per-county documented vs demographic-expected risk (MSSP RAF, person-year weighted) yields a coding-headroom signal Directional — a proxy, never expressed in dollars. Surfaced only to validate a market that already ranks outside-in; it never leads.

What the surface misses · the investigation, with receipts

Where the public read goes deeper

“Independent” is a trap word. A name screen passes payer-/system-/academic-owned groups (Optum, Emblem, UPMC, Mount Sinai). The ownership overlay removes them, each tagged with its named parent — true targeting needs the ownership graph, the next-build moat. Live
The lens audits you, not just the market — your own ACO surfaces as a competitor from CMS filings alone, independent confirmation of your footprint. Live
Federal naming has gaps — resolved deterministically, never guessed: some REACH competitors render as raw entity IDs where the name table doesn’t join; treated as unresolved, not approximated. Live

Evidence & confidence

Every figure carries its source

Market metrics: High (fresh, in-band, active sources, PY2024/PY2023). RAF headroom: Directional. Acquisition independence: Directional (candidate-not-confirmed). Every figure carries its source + as_of and reproduces against the lake.

Sources: CMS Geographic Variation PUF · CMS MSSP County RAF · CMS ACO REACH aligned-bene PUF + registry · NPPES · PECOS clinic groups + OIG LEIE · CMS ZIP↔county. Verified-metric SQL version-controlled (reference/agilon/metrics/).

The exchange · Track B

Public predicts — your data calibrates

This pack predicts from public data. The two proxy points — true independence (§3) and documented-risk headroom (§4) — are exactly where your data turns prediction into proprietary calibration: the groups you’ve screened, the ownership you’ve confirmed, the RAF you actually book. Public graph predicts → you validate → the validation becomes an edge no competitor can buy.

Boundaries: relationship is Phil’s; pricing is a later, separate conversation. Generated artifact — every Live figure reproduces against the lake on request.

100% U.S. public federal records · Zero PHI · Every Live figure reproducible against the lake on request · powered by Talon AI LLC