Bid Cap
Company library Healthcare

Company report

UnitedHealth Group Incorporated UNH

Three-pass checked

The bet you're really making is that UnitedHealth stops paying out more in medical bills than it planned, so its profit climbs back toward where it stood two years ago. You're betting the wave of extra surgeries and doctor visits that older Americans started using in 2024 has crested, and that the premiums governments and employers pay are catching up to what care actually costs. Right now it is turning: the first half of 2026 was the biggest ever, profit from operations up 19%, and for every premium dollar collected the company spent about 85 cents on care, better than the 88 it promised for the year. The thing to watch: sales barely grew, and the Justice Department is looking at how it bills Medicare. You pay 26 times trailing earnings, near the top of its range since 2014, but only 15 times what it is expected to earn by 2028.

Key data

Price$397.14
52-week range$255.97 – $461.62
P/E (trailing / FY28)26.6x / 15.1x
EV/EBITDA15.7x

UNH · price with moving averages

Daily · 6MWeekly · 3Y
$208$317$427$536$646 Oct '23May '24Dec '24Jul '25Feb '26Oct '26 Bid Cap
EMAs82140

Source: market data.

The business

UnitedHealth is two businesses under one roof. UnitedHealthcare collects premiums from employers and from the government for Medicare and Medicaid members, then pays their medical bills. Optum is the services arm: it fills prescriptions (OptumRx), employs and contracts doctors and clinics (OptumHealth), and sells data, billing software and consulting to the rest of the health system (OptumInsight, sitting on a $10.7B backlog of signed work). Premiums are about 78 cents of every revenue dollar. The two halves feed each other, the insurer routing members to Optum's pharmacies and physicians while Optum's tools help the insurer price risk. That loop, at this scale, is the moat. It is also why one bad stretch on medical costs, 2024 into 2025, dragged the whole machine down at once.

The numbers

Revenue has compounded about 12% a year since 2021, from $288B to $448B. Earnings went the other way.

QuarterRevenueNet incomeDiluted EPS
Q2 2025$111.6B$3.4B$3.74
Q3 2025$113.2B$2.3B$2.59
Q4 2025$113.2B$0.1B$0.05
Q1 2026$111.7B$6.3B$6.90
Q2 2026$112.0B$5.5B$6.04

Q3 2025 was the floor, $2.3B of net income, gutted by medical costs running ahead of premiums (Q4 2025 detail is not in this pull; adjusted EPS that quarter was $2.11, the actual trough). Then the snap back: Q1 and Q2 2026 earned $6.3B and $5.5B. On the adjusted basis the Street tracks, Q2 came in at $6.38 against a $4.94 estimate, a 29% beat, after a 12% beat in Q1. The recovery is arriving faster than the desk penciled in.

YearRevenueNet incomeDiluted EPS
2021$287.6B$17.3B$18.08
2022$324.2B$20.1B$21.18
2023$371.6B$22.4B$23.86
2024$400.3B$14.4B$15.51
2025$447.6B$12.1B$13.23
2026, 1H to Jun$223.8B$11.8B$12.94

Operating income peaked at $32.4B in 2023 and fell 41% to $19.0B in 2025; net income nearly halved. First-half 2026 operating earnings of $17.0B are already up 19% year over year, and the $12.94 of half-year EPS is running past all of 2025.

PeriodPremiums, $BMedical costs, $BCare ratio
Q2 202587.978.689.4%
Q2 202687.075.486.7%
1H 2025174.4152.087.1%
1H 2026174.5148.885.3%

That ratio is the whole story: medical costs fell 4% year over year while premiums held, and the ratio held through Q2 at 86.7%, comfortably inside the 88.1% full-year guide. At $397 you pay 26 times trailing earnings but 15 times what it is expected to earn by 2028, about 10% above the 2023 peak. The first-half ratio says the margin recovery is real and early, yet the stock still carries a top-of-range trailing multiple because two things cloud the forward path, a top line that grew just 1.2% and a Justice Department inquiry into Medicare billing. The print that settles it is the Q3 2026 care ratio against that 88.1% guide, since the ratio seasonally climbs through the year.

Management

No insider bought a share in the last twelve months; the only activity was three small sales by Optum's Patrick Conway totaling about $945K, plan status not disclosed. Capital return tells more: the company paid $2.1B in dividends at $2.32 a quarter, a 2.3% yield, and resumed buybacks in Q2, $3.2B for the half after spending nothing in Q1. That pause-then-restart pattern fits a management that throttled repurchases while earnings were falling and turned them back on once the care ratio broke its way, which is the disciplined version of the story rather than the reckless one.

How it fails or surprises you

The care ratio reaccelerates. The ratio rises seasonally into the back half, and the full-year guide of 88.1% implies H2 lands near 90% against H1's 85.3%. If Q3 prints above 89%, the first-half recovery was seasonal timing, not structural repair, and the forward EPS path collapses with it.

The DOJ escalates. The Medicare Advantage billing inquiry moving from question to civil action or settlement resets sentiment on its own timetable, independent of earnings. A material number, or a coding-practice change that shrinks risk-adjusted revenue, hits the profit engine directly and dates cannot be predicted.

The recovery overshoots (right tail). If the care ratio holds in the mid-80s while Optum keeps growing, EPS clears the $26 FY2028 path early, and a 15x forward multiple re-rates toward the 20x the business earned in calmer years. On restored peak earnings that is a materially higher stock, and today's price pays for none of it.

Closing thoughts

The Q3 care ratio will tell you whether the first-half recovery is structural or seasonal. If it holds near the guide at 88%, the earnings engine is repaired and you are buying it at 15 times forward; if it spikes above 89%, the first half was timing and the forward path collapses. The evidence in hand, medical costs falling while premiums hold and a care ratio already inside the guide, says the operating recovery is underway; the market's top-of-range trailing multiple and depressed forward multiple say it half-believes it and is discounting the legal overhang. The fatter tail is up, because you are buying a repaired earnings engine at 15 times forward while sentiment still prices the trough, but the left tail is real and binary: a DOJ action or a Q3 ratio blowout would prove the first half was rented, not owned.

The bet is still that UnitedHealth stops paying out more in medical bills than it planned, so its profit climbs back toward where it stood two years ago. What breaks it is the care ratio giving way or the DOJ inquiry turning into an action. The one pair of numbers that tells you first: the Q3 2026 care ratio against the 88.1% guide, read next to the legal-proceedings footnote. Hold inside the guide with the footnote quiet and the discount closes; either one gives and you learn the recovery was borrowed.

Methodology

Sector frame: managed care. Anchored to the Form 10-Q for the quarter ended June 30, 2026, filed August 10, 2026, with income statement, revenue-mix and share figures taken as filed from the company document and SEC XBRL company facts. Q4 2025 quarterly figures derived as FY2025 less first nine months. The care ratio is computed as medical costs divided by premium revenue (total revenue less products, services, and investment income) from the filing; the full-year 88.1% guide is management's stated outlook. Adjusted EPS, beats and consensus are vendor-sourced. Price, 52-week range and multiples are vendor market data as of September 6, 2026. Documentation prepared with AI assistance. Not investment advice.

Fact check: 1 hallucination corrected (2028 consensus EPS is 10% above 2023 peak, not a restoration), 1 approximation corrected (200-day MA percentage). Bundle financials reconciled to filed 10-Q. Management's FY26 care ratio guide (88.1%) cited but not independently verified in this run (would require earnings call transcript). Final analysis verified as of Sep 6, 2026.

Bid Cap

Daily ideas, a 390-name database, and a model long/short book from an investor who mostly covers financials. $70 a month or $700 a year.

Subscribe on Substack