Issue 1 · 2026-07-04
RuleCascade — Issue 1 · published 2026-07-04
On July 2, CMS proposed paying hospitals average sales price minus 33.4% for drugs bought through the 340B program, a $4.55 billion first-year cut that statute forces CMS to recycle into higher payment for every other outpatient service. Only public and nonprofit hospitals can hold 340B status. HCA runs roughly 190 for-profit hospitals, none eligible, so it takes none of the cut and keeps the offset. CMS's own impact analysis puts for-profit hospitals up 7.4% under the rule. That redistribution is the top of this issue's stack: long HCA into the November 2026 final rule, with five more calls ranked behind it that I'm watching or flagging rather than grading.
| # | Call | Status | Mechanism |
|---|---|---|---|
| 1 | HCA long | Graded call | 340B cut recycles $4.55B into higher non-drug OPPS rates; for-profits take none of the cut and keep the offset |
| 2 | RDNT long | Watch | Site-neutral imaging drops the hospital rate 60% at excepted off-campus sites; the price umbrella over freestanding centers comes down |
| 3 | DGX / LH short (genetic testing) | Watch | Two payers tightened genetic-testing and lab-panel rules in one month; denial risk lands on lab volume |
| 4 | INSP short | Flag | Aetna added the hypoglossal-nerve-stimulation codes to precert; one payer, and a company with one product |
| 5 | AMGN short | Flag | Aetna site-of-care review steers Xgeva to biosimilars outside the hospital |
| 6 | WAY / RCM long | Flag | Fourteen payer notices adding prior-auth codes in June; each added code is workflow-vendor demand |
Sections follow in rank order, and depth follows rank. Only #1 is a graded call this issue; watches and flags, if later upgraded, are graded from the document that upgrades them.
The CY2027 OPPS/ASC proposed rule (CMS-1850-P) landed July 2 with a CMS fact sheet and rule text on the Federal Register. The operative language:
"we are proposing for CY 2027 to pay for 340B acquired drugs at the drug's Average Sales Price (ASP) minus 33.4%."
Current law pays ASP plus 6%, so the proposal swings 340B drug reimbursement by about 39 points. The next sentence is the one that turns a sector policy into a single-name trade:
"This proposal is estimated to reduce Original Medicare drug payment by $4.55 billion dollars and beneficiary drug payments by $1.15 billion in the first year. Statute requires that this policy be implemented in a budget neutral manner, so this proposal would increase OPPS payments for non-drug services by an equivalent amount."
The AHA, which opposes the rule, works that budget-neutral arithmetic out to an 8.14% conversion-factor increase on non-drug services. Treat 8.14% as the opponent's estimate, not settled math: the AHA has every incentive to headline the largest offset it can, and I have not cross-checked it against the rule's conversion-factor tables. It reconciles to a non-drug OPPS base of roughly $56 billion ($4.55B ÷ 0.0814), the right order of magnitude, so I use it as a directional input. The headline payment update sitting alongside all this is ordinary: 2.4%, a 3.2% market basket less 0.8 points of productivity.
Three other provisions in the same document matter for the call. The remedy recoupment left over from the last 340B fight accelerates:
"CMS is proposing to revise the annual offset percentage for non-drug items and services from 0.5% to 3% effective CY 2027, excluding hospitals that enrolled in Medicare after January 1, 2018. This 3% reduction would remain in effect until the estimated payment reduction reaches $7.8 billion, which CMS estimates will occur in CY 2029."
Site-neutral policy widens by one category. Imaging without contrast at excepted off-campus hospital departments would be paid at the Physician Fee Schedule rate. HFMA breaks the effect into a $190 million Medicare payment reduction in the first year, with additional beneficiary premium and cost-sharing effects on top; the $190 million is the number that hits provider revenue. That provision is #2 in the stack, below. And the inpatient-only list loses another 638 procedures in year two of a three-year phase-out.
Comments run through August 31, 2026. The final rule typically lands in early November — November 2026 for this cycle — effective January 1, 2027.
July 2 coverage led with the cut. Healthcare Dive went with "slashing," the AHA statement reads as a sector-wide grievance, and HFMA's whole-rule math nets the sector to about +1.9% after offsets, which makes it sound like not much happened. The aggregate hides the event: the money moves from one class of hospital to another, and the class that loses writes the press releases.
Follow the $4.55 billion channel by channel:
The 2.5-point recoupment drag is the one give-back big enough to argue with. It runs against the gross for-profit gain STAT reports at 7.4%, in the same 2027-2029 window the trade holds, and takes back a real share of the first-year offset. That is why the conviction is moderate. The net still points the same way: the recoupment is a known finite $7.8 billion recovery every OPPS hospital carries, while the 340B redistribution is the new money and it flows toward the for-profit class. The instrument question is simpler than usual, since the rule is federal and applies to every OPPS hospital — no state-footprint mapping risk — and HCA is the largest listed pure-play operator.
The legal design matters as much as the arithmetic. The 2018 ASP-22.5% cut died in AHA v. Becerra because CMS skipped the required acquisition-cost survey. This time CMS ran the survey, January through April 2026, about 1,300 hospitals, 41.4% response rate per HFMA. TD Cowen (via STAT) still expects a court challenge but reads the survey basis as the cure for the prior defect. My working odds: call it 60% the rate finalizes near proposed. CMS softening after comments is the realistic alternative, maybe one chance in four, and there is direct precedent covered below. The tail is a court enjoining it before January 1, 2027 — one chance in eight or so. A lawsuit is near-certain regardless.
Sell-side had the direction inside a day. TD Cowen, quoted in Healthcare Dive, called the non-340B for-profit chains "major beneficiaries" on July 2, and the for-profit basket traded up on it. Direction is consensus. What the basket framing skips is where the 8.14% can legally show up. The budget-neutrality statute operates inside the hospital outpatient system: the offset raises the OPPS conversion factor and nothing else. Ambulatory surgery centers are paid off a separate conversion factor with its own budget-neutrality arithmetic, so the ASC schedule gets the ordinary 2.4% update and none of the 340B recycle. Capture is proportional to hospital-billed (HOPD) Medicare outpatient revenue, and the names in the basket differ on that dimension by a lot.
| Ticker | Market cap | 8.14% offset capture | Mix |
|---|---|---|---|
| HCA | ~$91B | Full | 190 hospitals; outpatient is 38.4% of FY25 patient revenue, the bulk billed as HOPD under OPPS |
| THC | ~$17.5B | Roughly half | 50 hospitals capture it; USPI's 533 ASCs and 26 surgical hospitals, about half of EBITDA, bill the ASC schedule |
| UHS | ~$9.6B | Under half | About half of EBITDA is behavioral, paid under the inpatient psych system with no OPPS exposure; the acute half is inpatient-weighted |
| CYH | ~$0.55B | Full, levered | 60 hospitals after the June divestitures; 6.5x net debt/EBITDA turns each point of uplift into equity torque |
| SGRY | ~$1.8B | None | 300+ facilities on the ASC schedule; commercial is ~51% of revenue; its lever in this rule is the inpatient-only-list migration, not the recycle |
Tenet is the interesting second. Hospital and ambulatory now split its EBITDA close to evenly — Q4'25 ran $603 million hospital against $580 million ambulatory, and Q1'26 ambulatory was $484 million at roughly 40% margins — so only half the company touches the offset. The USPI half rides a different provision, the 638 procedures leaving the inpatient-only list, which pushes surgical volume toward the settings USPI operates. That makes THC a blended play on two provisions, with lower per-share sensitivity to the 340B recycle than the basket framing implies, against a $4.485-4.785 billion FY26 EBITDA guide.
UHS is the wrong instrument for this mechanism. Behavioral hospitals bill under the inpatient psychiatric facility system, so about half of its $2.59 billion in FY25 adjusted EBITDA has no OPPS exposure at all (Q3'25 same-store ran $428.3 million acute against $404.5 million behavioral), and the acute half skews inpatient. Cheap after a weak 2026, but cheap on other grounds.
Community Health is the levered expression: a $550 million cap under 6.5 turns of net debt, so each point of Medicare outpatient uplift moves the equity more than anywhere else on the list. A proposed rule that can soften in November is a bad thing to be levered to, and the divestiture program is still reshaping the base. An option, not a call, and I pass. Surgery Partners does not belong in the thesis at all: it captures none of the offset, its book is half commercial, and the inpatient-only-list migration story its management leans on is a real but separate trade.
So the call goes to HCA — the largest pool of OPPS-billed outpatient revenue, the cleanest capture of the 7.4% net figure, and the deepest liquidity on the list. One more thing about that 7.4%, because the day-one notes flattened it: CMS's for-profit figure is net of the give-backs inside the same rule, the recoupment acceleration and the imaging repricing that for-profits also eat, which is why it sits below the 8.14% headline offset. Whatever edge remains after the July 2-3 repricing lives in that netting and in the capture split above, and the November 2026 final rule grades both.
The cited numbers: $4.55 billion of drug payment moves in year one, the offset is AHA's 8.14% on non-drug OPPS payments, and CMS says the for-profit class gains 7.4% on its Medicare outpatient payments.
Turning 7.4% into HCA EPS requires HCA's Medicare OPPS revenue, and the company doesn't publish that split, so sizing stops there. As a guardrail only: against HCA's roughly $75 billion revenue base, Medicare outpatient is a minority slice of a minority payer bucket, and 7-8% on that slice is plausibly a few hundred million dollars a year of pre-tax swing, which is real money and short of a re-rating on its own. The call is about whether that number survives to January 1, 2027, and management's own sizing of it is the first item on the calendar below.
Day-one coverage was everywhere: STAT, Healthcare Dive, Becker's, HFMA, 340B Report, plus same-day sell-side commentary. Secondary reporting has HCA up around 5% in the sessions after July 2. I could not verify that move against a quote feed before publication and I don't have an XLV-relative figure for the window, so treat the 5% as reported and let the Performance page carry the graded numbers.
I treat the announcement as priced and the November path as open: whether the rate survives the final rule intact, and the 2027-2029 recoupment arithmetic that most of the day-one coverage skipped.
Direction long, held at that moderate conviction, because this is a proposed rule with a litigation history, and part of the announcement move has already happened.
The call is graded the way every call on the record is graded: entered at the close of the first NYSE session after the evidence was public. The rule text dated July 2, 2026, and NYSE was dark July 3, so the entry is the July 6, 2026 close, measured against XLV from there. That entry sits after the reported 5% pop and I take it anyway, because the thesis is a four-month regulatory hold, not a day-trade on the announcement. Invalidation is the final rule: if CMS drops or softens ASP-33.4% in November 2026, or a court enjoins the 340B cut before January 1, 2027, the call closes and grades where it stands.
| Date | Event | What I'm watching |
|---|---|---|
| Jul 21-28, 2026 | HCA Q2 print (typical window) | Whether management sizes the rule at all. |
| Aug 31, 2026 | Comment deadline | Volume and legal posture of nonprofit and AHA comments; litigation signaling. |
| ~Nov 2026 | CY2027 OPPS final rule | The decision point. ASP-33.4% and the offset survive near proposed form, or they don't. |
| Jan 1, 2027 | Rates effective | Cash flows begin. |
Confirms: the final rule retains the ASP-33.4% rate and the budget-neutral offset near proposed form; HCA management quantifies a 2027 benefit on the Q2 or Q3 call.
Breaks if CMS softens in the final rule, and there is precedent. In the CY2026 cycle CMS proposed accelerating the same recoupment offset to 2% and dropped it after comments. Breaks harder if a court enjoins the 340B cut, because budget neutrality takes the 8.14% offset down with it. The branch I'd hate is the recoupment acceleration surviving while the cut is enjoined, which leaves HCA the drag with no offset; low probability, and I close the call on it immediately rather than argue with it.
The same rule extends CMS's section 1833(t)(2)(F) volume-control authority, the tool it used on clinic visits in 2019 and drug administration in 2026, to the four imaging-without-contrast APC levels when furnished in excepted off-campus provider-based departments. Those APCs cover radiography, ultrasound, mammography and DXA, and CT and MR performed without contrast. The proposed rate is the PFS-equivalent 40% of the OPPS rate, a 60% cut on the affected lines, effective January 1, 2027 with no phase-in proposed; the clinic-visit policy phased over two years, this one doesn't. Rural sole community hospitals are exempt, as they were in the prior two rounds. The boundaries matter as much as the coverage: contrast-enhanced CT and MR sit in separate APCs and keep full OPPS rates, as do PET and nuclear medicine, on-campus departments are untouched, and non-excepted off-campus sites have billed at PFS-equivalent rates since the 2015 site-neutral statute.
The dollars are small next to the 340B move — $190 million of first-year provider revenue by HFMA's figure above, in a rule that shifts $4.55 billion — and the 2026 drug-administration expansion was sized about the same. What earns the provision its own section is the sequence: clinic visits, then drug administration, then non-contrast imaging, each finalized essentially as proposed, and no obvious reason contrast studies stay off the list in a future cycle. The legal question is also closed in a way the 340B piece is not. The AHA challenged the clinic-visit policy, the D.C. Circuit upheld CMS's authority in AHA v. Azar in 2020, and the Supreme Court denied cert in 2021.
The hospital lobby's objection is on file — the AHA says HOPDs "disproportionately care for patients with more complex health conditions" and serve "rural or underserved communities," and the AAMC filed the academic-medicine version. The ACR has not taken a position; its July 2 note says it is reviewing the rule and will publish a detailed summary. The hedge is structural: hospital-based radiology practices lose technical-fee economics under this provision while office-based and freestanding groups gain relative position, and the College represents both.
The trading readthrough is a price wedge, not a payment. A non-contrast MR at an excepted off-campus hospital department bills Medicare at the full OPPS rate under current policy; from January 2027 it would bill at 40% of that, roughly what a freestanding center collects. RadNet is the only public pure-play on the freestanding side: ~440 centers, $2.04 billion of FY25 revenue, $300.2 million of adjusted EBITDA, $575.6 million of Q1'26 revenue, payer mix 57.1% commercial and 26.6% Medicare. The rule sends RDNT no new dollars, since it already bills freestanding rates. The mechanism is relative: the hospital price umbrella at excepted off-campus sites comes down, payer steerage toward freestanding sites gets easier to justify, and hospitals rethinking off-campus imaging feed RadNet's joint-venture and acquisition pipeline. Akumin, the next-largest fixed-site operator at ~145 centers, is private under Stonepeak, so there is no second instrument.
The 2019 precedent argues for patience. When the clinic-visit rate went site-neutral, volumes did not visibly move; hospitals absorbed the Medicare cut, commercial rates were untouched, and referral patterns stayed put. No rigorous utilization study exists either way, so read that as absence of evidence of migration rather than proof of absorption. If the same holds for imaging, the wedge shows up over years, in hospital exits and joint-venture formations, not in 2027 volumes — and $190 million of repricing spread across every excepted off-campus department in the country rounds the direct 2027 effect on a $2 billion revenue base to zero.
So RDNT holds the second slot in the stack, a watch rather than a trade, and the calendar is the reason as much as the size. The CY2027 Physician Fee Schedule proposed rule is due around mid-July 2026 (last cycle: July 14), and it sets the rates freestanding centers are actually paid; the last cycle's efficiency adjustment cut a long list of imaging codes. Going long the freestanding winner right before its own rate sheet prints is bad sequencing. Two triggers would upgrade this to a thesis: a July 2026 MPFS proposed rule that leaves imaging codes roughly whole, and a November 2026 final rule that keeps the imaging provision.
The one substantive item in Cigna's otherwise-housekeeping June update was a set of tightened lab guidelines effective July 1: a new hereditary-cancer genetic testing guideline plus substantive changes to 21 genetic and molecular testing guidelines, co-branded with eviCore. A second document from the same window works the adjacent seam: Molina's Washington plan posted a notice on lab panel billing and unbundling, tightening how multi-test panels get coded and paid. Different payers, different mechanisms, same destination: lab reimbursement mechanics got harder in June.
The readthrough runs to the testing companies, not CI, and the direction is short the testing volume. Quest and Labcorp hold the broad exposure, genetic testing as one line among hundreds. The concentrated version sits with the pure-play genetic names, where hereditary-cancer panels are the core book rather than a product line; Myriad and Natera bill the codes these guidelines govern. Two documents is a pattern worth a slot, not a thesis. What upgrades it: a third payer publishing comparable genetic-testing rules, or a lab quantifying denial-rate impact on a call. What complicates it: eviCore itself sits under strategic review, which clouds how hard Cigna's guidelines get enforced through a sale, and guideline changes take a quarter or two to surface as denials, so the first legible print is Q3 2026. Nothing is priced. Neither DGX nor LH moved on a coverage-policy PDF, and nothing here reaches an estimate until denials do.
Aetna's June OfficeLink update adds codes C8007-C8013, hypoglossal nerve stimulation, to its precertification list. Those codes describe one procedure and, in practice, one company's device: Inspire Medical, which has no second product. A single-product company picking up a precert requirement at a major national payer is a utilization headwind in the only revenue line it has. Precertification is friction rather than denial, and Aetna is one payer, so this stays a flag. Direction short. The upgrade is a second national payer adding a comparable requirement, or Inspire calling out prior-auth drag on a print; the kill is Aetna approvals proving routine while implant volume holds. Nothing is priced — the change ran in a provider newsletter, and INSP does not trade on precert additions until they show up in procedure counts.
The same Aetna document adds a site-of-care review that steers Xgeva (denosumab) toward biosimilar substitution when infused outside the hospital. Short AMGN is the direction, and scale is what caps it at a flag: Xgeva is one franchise inside a revenue base north of $30 billion, and biosimilar denosumab has been eroding it since the 2025 launches, so a steering policy accelerates a decline every model already carries. The upgrade is a second payer publishing similar steering, or Xgeva volume data running ahead of the modeled erosion curve; the kill is quarterly Xgeva sales landing on the curve. The erosion is priced. The acceleration is not, and it is small.
June's provider bulletins ran heavy with new prior-authorization requirements. CareSource added PA codes on its marketplace plans and published new-code lists in Georgia and Arkansas. Molina updated its Iowa PA code list and put varicose-vein treatment behind PA in Michigan. AmeriHealth Caritas posted a Pennsylvania notice, Aetna's June OfficeLink carried its own additions, and Colorado and Maine ran parallel changes on the fee-for-service side. Fourteen documents in the window with the same shape, none big enough to price alone. The read that survives is long the vendors that process the paperwork: every added code is another authorization to submit, track, appeal, and bill for, and that workflow is Waystar's and R1's product. The mechanism is diffuse and slow, which is why it sits last. The upgrade is a vendor attributing volume or bookings growth to utilization-management expansion on a call, or the document count holding through August 2026; if the count mean-reverts, June was a busy month and not a theme. Nothing is priced, and no analyst models payer-bulletin volume.
One call in this issue is graded: long HCA, entered at the close of the first NYSE session after the July 2 rule text — July 6, 2026 — and measured against XLV in the called direction from that level. The dated entry lives on the Performance page. The grading rule is short: entry follows the evidence date, and nothing is backdated.
The corrections: two calls stamped July 3, 2026 — an ADUS short and an MOH long — went up on a closed session, NYSE being dark for the Independence Day observance, and were withdrawn the same day. Both theses had also failed the document read: Addus sold its entire New York operation in 2024, so a New York records-retention reminder touches no Addus revenue, and the Molina item is a recurring SNF claims-hygiene reminder with no rate or coverage content. Both were withdrawn.
The Medicare GLP-1 Bridge went live July 1. Eligible beneficiaries pay $50 a month, manufacturers net $245 per 30-day supply, and KFF counts about 3.8M eligible in a population that previously had no covered channel for weight-loss drugs. Lilly holds two of the three eligible products, including the only pill, and the coverage is saturated: LLY printed a record high into launch week. No trade yet; the economics were public since the May 6 pre-announcement. The open question is uptake, which spans an 8x range in KFF's cost scenarios, and the first read comes with LLY's Q3 2026 print.
The payer-side version of the utilization theme dissolved on the document read. UHC is publicly cutting prior authorization by 30%. Aetna's headline radiology cut re-implements a CMS policy in force since 2016 while keeping its broadest code-edit list behind a provider login. There is no trade in the payers themselves until one prints a notice with a published code list and a dollar figure attached; the vendor side of the same bulletins sits at #6 above.
RuleCascade tracks 50+ primary regulatory and payer sources across the sector; every claim in an issue links to its primary document, and undated documents are disclosed as undated, never backdated. Calls are entered at the first trading session after their evidence was public and graded against XLV in the called direction; nothing is backfilled. Corrections are published in the issue and stay up. I publish 10 to 12 times a year, on no set schedule.