Open in interactive viewer → charts, metric popovers & call review

Astrana Health: Record Numbers, a 22% Drawdown, and a Cyber Incident Nobody Will Size

The delegated-risk operator beat, held guidance, and re-cut its own revenue math — but Medicaid acuity, an unquantified breach, and an accounting transition are what the tape is pricing.
ASTH · Earnings Call · 2026-11-05

The beat the market declined to buy

Astrana Health's third quarter was, on the numbers, its best ever: revenue of $1.02 billion, up 7%; adjusted EBITDA of $77.2 million, up 13% and near the top of a $72.5–77.5 million guide; adjusted diluted EPS of $0.71. Nine-month free cash flow of roughly $105 million funded $116 million of debt paydown, pulling net leverage to about 2.2x against the 2.5x pro forma when Prospect closed. The Care Partners business is doing what management said it would — Texas cohorts maturing, medical cost trend "slightly better" than the ~5.2% embedded in guidance, G&A down to 6.2% of revenue year-to-date. And yet the stock enters this report down 21.7% over 90 days, in a single unbroken decline from a July 1 peak of $49.81. At 0.67x revenue, it is a business priced as though its revenue is low-quality and its balance sheet is fragile — a reading reinforced by $541 million of net debt and 2.1x interest coverage. That gap between operating reality and valuation is the whole dossier. The question is whether Q3 closed it or widened it.

The new number that answers an awkward question

The single most company-unique thing on this call was an accounting invention: Premiums at Risk, or Gross Equivalent Consolidated Premiums. The reason is structural. ACO REACH, which sunsets at end-2026, lets Astrana book the full Medicare benchmark as revenue with medical costs in expenses; MSSP books only its share of savings, net. As the ACO program mix rotates toward net-basis arrangements and the new ACO LEAD model, reported revenue can shrink without a dollar of economics changing.

We're introducing a supplemental measure: Gross Equivalent Consolidated Premiums, or Premiums at Risk... On this basis, we expect approximately $4.45 billion in premiums at risk for 2026, compared to our reported revenue guidance of $3.9 billion to $4 billion.

Chan Basho, Chief Financial Officer and Chief Operating Officer · 2026-11-05
Read that as an inoculation. Management is telling investors in advance not to confuse a GAAP revenue headwind with a business headwind — and CFSO Chan Basho's admission that "changes in our reported GAAP revenue won't reflect necessarily changes in the total Medicare spend we're managing" is the tell. This is a genuinely new disclosure for the company and, per the analyst on the line, the first time he had seen a peer do it. It also explains the curious keyword clustering this quarter around net basis and "apples to apples" — language of a company defending a denominator.

Medicaid stops being a footnote

The keyword ranking inside Astrana's own transcript moved decisively. membership attrition is now the top theme, replacing the full-risk conversion language that dominated the prior quarter. The specific change is acuity, and management said so plainly in a way it had not three months earlier: “Earlier this year, we said that we weren't seeing as much adverse selection as we expected. We're starting to see signs of that ticking up a bit, modest changes in the risk profile of the members that are remaining as the year progresses.” — Brandon Sim, President and Chief Executive Officer · 2026-11-05 Contrast that with August, when the same CEO described Medicaid as fully contained: “Medicaid is in line with that trend number, and that was inclusive of potential adverse selection effects in our guidance.” — Brandon Sim, CEO or Senior Executive · 2026-08-06 And with May, when he said trend had come in "better in Medicaid as well relative to our expectations" with "lower adverse selection." Three calls, three different reads on the same risk pool. That is the shift the tape has been pricing. The forward number is now explicit: “We've been incorporating these developments into our planning for some time and currently anticipate a 10% to 15% decline in Medicaid membership in 2027.” — Brandon Sim, President and Chief Executive Officer · 2026-11-05 With 10–20% of those members sitting in UIS status moving to fee-for-service, and work requirements plus redeterminations ahead, management guides a mid-single-digit net EBITDA headwind. This is not idiosyncratic — Medicaid work requirements sits on the market's 90-day advancers list, carried by MOH, UHS and MMS, which is a reminder that the policy variable is industry-wide even if the acuity outcome is Astrana's alone.

What management won't quantify

Two items deserve the skeptical eyebrow. First, the September cybersecurity incident, which appeared in the Q&A and nowhere in the prepared remarks:

We experienced an incident and obviously involves criminal activity, and there's an ongoing investigation, including with law enforcement... I'm not sure it's extremely productive to speculate, you know, and get ahead of that process.

Brandon Sim, President and Chief Executive Officer · 2026-11-05
No size, no timeline, no remediation date. For a company whose entire equity story rests on being a proprietary data and AI powered care enablement platform, an unpriced criminal breach of that platform is the risk the model cannot see. Second, the fourth-quarter range. Guidance of $48.5–61.5 million of adjusted EBITDA against $77.2 million in Q3 implies a wide band, and the answer was a rerun: “the wider range is really just, frankly, an artifact.” — Brandon Sim, President and Chief Executive Officer · 2026-11-05 In August the identical question got the identical word — “if you're focused on the width of the range, I think that's primarily an artifact.” — Brandon Sim, CEO or Senior Executive · 2026-08-06 Two consecutive quarters of "artifact" is a pattern, not a one-off. And the 2027 anchor has quietly migrated. On the May call the framing was “we continue to believe that the 2027 guide is within the range of outcomes” — Brandon Sim, President and Chief Executive Officer · 2026-05-08 — the $350 million EBITDA framework. In August it was reaffirmed as “mid to high teens EBITDA growth, not just for 2027, but into the medium term years as well.” — Brandon Sim, CEO or Senior Executive · 2026-08-06 Now it is "at least mid-teens" medium-term growth, with a verbal bridge built from new membership, rate-versus-trend and expansion markets, net of Medicaid. The target has become a rate, not a level.

The offsets are real, and the market is ignoring them

The bull case did not disappear. Texas is the proof point: ~7,000 MA members moved from gain share to global risk plus 3,000 professional-risk lives, and “our Texas business is performing ahead of plan, with medical cost trends running favorably to our underwriting assumptions.” — Brandon Sim, President and Chief Executive Officer · 2026-11-05 Star ratings are inflecting — “around 30% of our members today are in 4-star plans, and we expect that to be... above 50% of our membership in 4-star plans going forward into '27” — Brandon Sim, President and Chief Executive Officer · 2026-11-05 — and the 2027 rate notice, at roughly 5% effective growth, is expected to match trend and hold MA margins stable. The under-appreciated lever is health system distribution. Care enablement is already “over $300 million of segment revenue, you know, operating at very healthy gross and operating margins” — Brandon Sim, President and Chief Executive Officer · 2026-11-05 — and it just signed a Connecticut multi-specialty group and a California hospital, with no dedicated business-development headcount. Layer on the Atlas provider platform launch, annual wellness visit penetration at 65% of eligible members, and 4-star plans momentum, and the platform's non-premium-risk revenue stream is quietly becoming a second engine.

The verdict

Astrana is a company whose operations are accelerating while its multiple compresses: 3.5% operating margin, record EBITDA, 2.2x net leverage, and a stock 67% below its 2021 peak. The Q3 call's real content was defensive plumbing: a new Premiums at Risk metric to neutralize an accounting transition, an admission on Medicaid acuity that three months ago did not exist, and an unquantified breach of the very platform the story is built on. The company is telling you the economics are intact while the tape tells you the confidence is not. Re-rating requires two proofs, not one: a Medicaid number that stops drifting, and a cyber number management is finally willing to say out loud.