Story

The Full Story

For five years management told a single story: a 12–15% adjusted EPS compounding machine, a "balanced and resilient" portfolio, and a "flywheel for growth" anchored by Carelon. That story held through 2023 — and then collapsed. Adjusted EPS fell from $33.14 in 2023 to a guided at least $25.50 in 2026, the long-term enterprise margin target was quietly cut from 6.5–7.0% to 5.0–6.0%, and a $935M risk-adjustment accrual to CMS arrived in Q1 2026. Credibility eroded in two distinct waves — a Medicaid acuity miss in late 2024 and an ACA morbidity miss in mid-2025 — and the rebuild rests on a 2027 promise to "return to at least 12% growth" off a baseline that has been reset twice.

1. The Narrative Arc

Annotated timeline

The arc has three legs: a build phase (2019–2022) where vertical integration and rebrand were the story; a peak credibility year (2023) where the "flywheel" framing was uncontested and CAGR sat at 16%; and a two-stage reset (2024 Medicaid → 2025 ACA → 2026 CMS) that turned a coverage growth story into an execution and rate-recovery story.

2. What Management Emphasized — and Then Stopped Emphasizing

Topic frequency across the nine quarterly transcripts shows the narrative pivots clearly. The "flywheel" disappeared. "AI" exploded. "Redetermination" — once weekly — went silent. "ACA" became dominant in 2025.

What the heatmap reveals:

• "Flywheel for growth" — peaked at 6 mentions in Q4 2023, the single most-used framing of the strategic narrative. Dropped to zero by Q4 2025 and Q1 2026. The metaphor disappeared without explanation as the engine seized.
• "AI" went from 0–2 mentions through 2024 to 13 mentions in Q1 2026 alone. Boudreaux now describes "more than $1 billion in digital and AI-enabled capabilities." It became the new unifying story.
• Redetermination dominated 2023–24 calls, then nearly disappeared as the cycle ended — but resurfaced in Q2 2025 (11 mentions) once it became clear the cost impact had outlasted the membership impact by 18+ months.
• ACA exploded from 5–6 mentions per quarter to 34 in Q2 2025 when the morbidity miss landed, and remained elevated. What had been a high-conviction growth lever in 2023 ("up 30%+") became the largest single contributor to the 2025 guidance cut.
• CMS / risk adjustment went from zero historical mentions to 9 in Q1 2026, when the $935M accrual was disclosed.

Quietly dropped initiatives

3. Risk Evolution

The risk factors section of the 10-K is conservative by nature, but its emphasis migrates over time. Comparing 2020 to 2025 shows how the threat surface changed.

What's newly visible. The 2025 10-K is the first to name Medicaid community engagement requirements, the expiration of enhanced PTCs at the end of 2025, and the V28 HCC model as discrete risks — language that was either absent or generic in 2020. Goodwill at $39.5B is now 32.5% of consolidated assets, a meaningful impairment exposure as Carelon-related deals (Paragon, Kroger Specialty, BioPlus, CareBridge) season.

What became less important. COVID-19 dominated the 2020 risk factors and was effectively gone by 2023. The "general economic downturn" framing softened. State-specific concentration risk language is largely unchanged.

What was always there. The core risk — "slight differences between predicted and actual medical costs… can result in significant changes in our results of operations" — is identical in 2020 and 2025. The 2024–2025 EPS misses were not from a novel risk; they were the headline risk of the entire managed care category, exactly as described.

4. How They Handled Bad News

Two episodes test management credibility.

Episode 1 — Medicaid acuity, Q3 2024

In Q4 2023, the message was: "premium rate adjustments in recognition of medical cost trend… continue into 2025." In Q2 2024, the message tightened: "acuity has increased due to attrition of healthier members" but "we expect to achieve our full-year adjusted diluted earnings per share guidance of at least $37.20." In Q3 2024, that guide was cut to ~$33 with explicit language:

"While the rate increases we've received are the highest in the past decade, they are still inadequate to cover 2024 cost trend that we now expect to be three to five times historical averages."

Two things stand out about how this was handled. First, the framing — "the issues impacting Medicaid are time bound" — was repeated three times in the Q3 2024 call and twice in Q4 2024. It turned out not to be true: Medicaid margin guidance for 2026 is still negative 1.75%, more than two years after the framing first appeared. Second, the disclosure was clean and the financial reset was taken in one quarter — not dripped out across three.

Episode 2 — ACA morbidity, Q2 2025

By Q4 2024 the talking point was "another year of strong growth" in individual exchange. By Q1 2025: "effectuation rates… tracking a little bit lighter." By Q2 2025 the EPS guide was cut from ~$34.50 to ~$30 with this line:

"We recognize that revising guidance for the second consecutive year is disappointing… we are choosing to act now, not later… It is not based on assumptions of a near-term recovery."

The Q2 2025 call is the most candid in the set. Boudreaux opens with: "This adjustment is disappointing, and we're taking concrete actions to address it." No framing of "time bound." No reference to a pending recovery. The ACA risk pool was acknowledged to be structurally worse — "morbidity has stabilized" but at a higher level — and 2026 ACA pricing was explicitly built around the assumption that enhanced subsidies expire.

Episode 3 — CMS RADV notice, Feb 2026

The third episode is more contained. In February 2026 CMS issued a notice tied to historical risk adjustment data. By Q1 2026 the company had recorded a $935M accrual and disclosed an extended compliance window through July 31, 2026. Boudreaux's language was unusually defensive: "It's not about how we operate the business today… we stand firmly behind the integrity of our risk adjustment program." The accrual was excluded from adjusted EPS, but the underlying issue — historical Medicare Advantage coding — sits in the most-scrutinized area of MA economics and is a quiet flag for class-action investigation announcements that have already been published by multiple firms.

5. Guidance Track Record

The first column of this table is what management committed to at the start of each year. The second is where they revised it mid-year (when applicable). The third is what was actually delivered.

What this shows. From 2018 through 2023, initial guidance was a floor — the company beat by 1–4% every year. In 2024 and 2025 the relationship inverted: initial guidance was an aspiration, missed by 11% and 12% respectively. The 2026 guide of "at least $25.50" is set 25% below 2024's initial guide of $37.20 — without an underlying revenue collapse — implying a structural margin reset, not a one-year disturbance.

Why 5/10. Pre-2024, this management team would have rated 8–9. Two years of large mid-year cuts on the same basic issue (medical cost trend) drop that materially. The mitigating factors: the cuts themselves were communicated cleanly, not dripped; the 2026 guide is consistent with peer behavior (UnitedHealth and Centene reset similarly); the Q1 2026 raise was real, even after stripping the nonrecurring $1. The aggravating factors: the prior-quarter reassurances before each cut, the long-term margin target reduction, the CMS RADV accrual landing on top, and the fact that Carelon — the engine of the 2027 recovery story — just lost its founding leader (Pete Haytaian) with the CFO assuming oversight.

6. What the Story Is Now

The story has changed shape three times in 30 months. It is now narrower, more conditional, and more dependent on things outside management's control.

What has been de-risked

• Medicare Advantage 2026. Deliberate exits + portfolio repositioning produced a high-teens membership decline — but the math suggests a real path to ~2% MA operating margin in 2026 vs. negative recently. Star Ratings for PY2027 also recovered (59% of MA members in 4-star plans vs. 40% the prior year).
• ACA repricing. The 2026 book is priced for the elevated morbidity and the expiration of enhanced subsidies. Q1 2026 results showed the bronze-mix shift broadly tracking expectations — the early read is the worst pricing assumption did not break the book.
• Carelon Services external mix. Less exposed to the affiliated health-plan headwinds; CareBridge and risk-based oncology are real, scaled offerings, not slideware.

What still looks stretched

• Medicaid margin recovery to historical norms. Guidance is still negative 1.75% for 2026. Rates lag trend by 12–24 months in normal times; "work requirements" under the One Big Beautiful Bill Act add a fresh attrition vector that, by management's own admission in Q2 2025, hasn't been operationalized at scale outside Indiana and Georgia.
• CarelonRx margins. Long-term target was just cut to "mid-single-digit." The specialty pharmacy build-out (BioPlus + Kroger + Paragon) was sold as accretive; the segment guide for 2026 is "moderated by lower health plan membership."
• The 2027 12% promise. Off a $25.75 base, that's ~$28.85 — still 13% below 2023's $33.14, four years later. Hitting it requires rate catch-up and MA margin expansion and AI-driven OpEx leverage all working.
• CMS RADV resolution. The $935M accrual is "current best estimate"; the final number, and whether sanctions trigger after July 31, 2026, are open.

What the reader should believe vs. discount

• Believe: the 2026 guide is conservative by recent management standards; Boudreaux/Kaye learned from two consecutive cuts and built more cushion in. The Q1 2026 raise gives early evidence.
• Believe: the long-term margin target cut to 5–6% is honest. Better to have a target the business can hit.
• Discount: the "return to 12% growth" framing. It's directionally correct off the new base, but the 2018–2023 track record shouldn't anchor expectations going forward. The compounding regime ended in 2023.
• Discount: "AI" as a P&L catalyst at the magnitude implied. The narrative shift from flywheel to AI is real, but the dollars-of-savings claims are aspirational and the operating expense ratio has only moved 50–70 bps over two years despite the rhetoric.

The cleanest read: this is a managed-care category bottom story disguised as a single-name reset. Whether 2026 is the trough depends less on management execution than on Medicaid rate cycles, the post-subsidy ACA risk pool, and the CMS resolution — three things they don't control.