All Three Operating Segments Grew and the Stock Still Fell 5%: 2027 Is the Real Disclosure

Results vs. Consensus

CVS Health reported before the open on August 5, 2026 and held its call at 8:00 a.m. ET. The print was the strongest of the recovery on every headline measure, and the shares still finished the session down 5.1%. Reconciling those two facts is the work of this note, and the answer is not in the quarter. It is in the stretch of prepared remarks the CFO spent on a year he was not obliged to discuss.

Q2 2026 Scorecard

Two providers put adjusted EPS at $1.83 and $1.87; we use $1.85. Revenue estimates ranged from $98.3B to $100.0B, with the higher figure carried by two of the three sources we checked.

The size of the earnings beat deserves less weight than it first appears, because the consensus was not an independent forecast. The Q1 call guided the full year to $7.30 to $7.50 and described a "roughly 60-40 split of earnings between the first half and second half." Sixty percent of the $7.40 midpoint is $4.44; subtract the $2.57 already reported in Q1 and you get $1.87, which is the consensus to the cent. The Street was solving management's own phasing arithmetic rather than modelling the quarter. A 40% beat against that is a statement about how conservative the guide was, not about how far the business ran ahead of informed expectations.

Year-over-Year Comparison

The GAAP line is close to uninterpretable year over year. The prior-year quarter carried $833M of legacy litigation charges (a $542M False Claims Act penalty in retail and a $291M charge on pharmacy benefit manager remuneration reporting) plus a $471M premium deficiency reserve, and it was taxed at 38.5% against 24.7% this year. Adjusted EPS growth of 42.5% is the honest headline, and even that overstates the underlying move for the reason set out two sections below.

Sequential Comparison

The sequential table is the one that reframes the quarter. Enterprise adjusted operating income was flat, up $7M on $5.15B, because the composition rotated completely: Health Care Benefits gave back $615M while the two pharmacy segments added $522M between them and Corporate/Other improved $100M. The two quarters produced almost identical earnings from almost opposite sources. That is a healthier enterprise than Q1's one-segment result, and it is also the reason the second-half guide looks the way it does, because the segment that carried Q1 is the one whose seasonality now turns hard against it.

Revenue

Growth of 7.3% accelerated from 6.2% in Q1 and was, for the first time in this recovery, broad across all three operating segments on both the top and bottom lines. Health Services contributed $5,342M of the $7,181M consolidated increase, Health Care Benefits $1,280M and Pharmacy & Consumer Wellness $235M. Inside Health Care Benefits the mix remains the story: Government premiums grew 9.9% to $28,494M while Commercial premiums fell 19.7% to $6,625M as the individual exchange exit ran through a second quarter. CVS is still growing revenue by taking on more government medical risk and less commercial insured risk, and by passing through more drug cost.

Margins

The consolidated improvement of 101 basis points is more evenly distributed than Q1's was. Health Care Benefits went from 3.6% to 6.5% of segment revenue, Pharmacy & Consumer Wellness from 4.0% to 4.4%, and Health Services slipped from 3.4% to 3.3%. That last figure is worth pausing on. Health Services grew gross profit 11.7% and cut operating expenses 20.2%, and its adjusted operating margin still slipped, simply because revenue grew 11.5% against adjusted operating income growth of 10.0%. Read the GAAP line instead and you get the opposite distortion: segment operating income rose 45.5%, but only because the prior-year quarter carried $332M of charges that the adjusted line already excludes, being the $291M pharmacy benefit manager remuneration charge and a $41M loss on Accountable Care assets. The segment is doing better than its adjusted margin suggests and considerably worse than its GAAP swing suggests.

EPS

Adjusted EPS of $2.58 grew 42.5% on adjusted operating income growth of 35.4%. The gap is below the line, and it is the same two components as last quarter. The effective tax rate of 24.7% is running roughly 1,000 basis points below the first half of 2025 and has now gone two consecutive quarters without explanation in either the release or the call. Diluted shares of 1,287M are up 1.3%, and the full-year guidance assumes 1,289M, so the share count is a headwind through 2026 by design.

The Core Margin Bridge

This section exists because the single most decision-relevant number in the quarter is not in the release. Management disclosed the input; the arithmetic below is ours.

Health Care Benefits reported a medical benefit ratio of 87.4% against 89.9%, an improvement of 250 basis points, and segment adjusted operating income of $2,426M against $1,308M, an increase of 85.5%. Both figures contain items that relate to plan years already closed. The CFO sized this year's:

"These results include the impact of changes in our individual exchange risk adjustment position associated with the 2025 plan year as well as the impact of favorable prior year development. Together, these items contributed approximately $500 million or 140 basis points to our MBR in the quarter." — Brian Newman, EVP and CFO

The prior-year quarter had its own distortions, all of them disclosed in that quarter's filing: a $471M premium deficiency reserve in the Group Medicare Advantage product line, which the company itself sized at 140 basis points of medical benefit ratio, and roughly $300M of favorable prior-year development, being the $1.9B disclosed for the first half of 2025 less the $1.6B disclosed for the first quarter. Setting both quarters on the same basis:

The same exercise on the segment's earnings line: $2,426M less the $500M of prior-year items is $1,926M of core adjusted operating income, against $1,308M plus the $471M reserve, or $1,779M, a year ago. That is growth of about 8%, not 85%.

Assessment: the Aetna repair is real and is now two years in, but the reported margin trajectory continues to be amplified by legacy-year true-ups in a way the headline does not disclose. This matters more this quarter than last, because the largest single component, the individual-exchange risk adjustment for the 2025 plan year, relates to a business CVS has exited. There is no 2026 plan year to true up. Whatever that item contributed cannot repeat in 2027, which is a substantial part of why management's own 2027 baseline of $7.46 sits $0.54 below its 2026 guidance midpoint.

Segment Performance

Health Care Benefits

Revenue of $37,538M grew 3.5% and adjusted operating income of $2,426M grew 85.5%, against a prior-year quarter that carried the $471M premium deficiency reserve. Health care costs of $30,692M actually fell 0.2% in absolute dollars while premium revenue grew 2.7%, which is the mechanical source of the margin gain. Segment operating expenses grew 3.1%, marginally faster than revenue, taking the expense ratio from 12.5% to 12.4%. The composition of the improvement is set out in the bridge above; management's own framing of what sits underneath it was narrower than the headline invited.

"Excluding these items, our core performance in the quarter still exceeded our expectations. This core outperformance was largely concentrated in our Medicare business, where we continue to experience pockets of favorability as a result of strong medical cost management and disciplined pricing." — Brian Newman, EVP and CFO

Two disclosures moved in the right direction and one moved the wrong way. Total medical membership of 26,023 thousand rose 18 thousand sequentially, the first sequential increase since the individual exchange exit began, with Medicare Advantage up 27 thousand, Medicaid up 21 thousand and Commercial insured up 25 thousand. Days claims payable of 41.7 sits 0.8 days above the 40.9 recorded a year ago, having been 0.3 days below the comparable figure at Q1. Against that, days claims payable fell 1.2 days sequentially from 42.9, which is the normal direction after a first quarter that recognised $1.1B of development but is worth watching in a book where reserve adequacy is the whole debate.

Assessment: this is the best-evidenced quarter of the Aetna recovery on the metrics that are hardest to manufacture, namely membership stabilisation and a year-over-year build in days claims payable. It is also the second consecutive quarter in which the reported margin gain was materially larger than the underlying one, and management has now guided the segment to lose money in the second half. Both things are true, and the market spent the session working out which one to weight.

Health Services

Revenue of $51,795M grew 11.5% and, for the first time in this cycle, gross profit kept pace: $2,537M against $2,272M, up 11.7%, with gross margin flat at 4.9%. That is a genuine reversal of the Q1 pattern, where 11.0% revenue growth produced a 3.7% gross profit decline and a 70 basis point margin contraction. Pharmacy claims processed grew 0.9% to 473.0 million, the first volume growth of the recovery, though the release characterises volume as having "remained consistent on a 30-day equivalent basis." Adjusted operating income of $1,733M grew 10.0%. Health care delivery, principally Oak Street Health, grew revenue "nearly 23%" per the call and performed in line with plan.

For the half, however, gross margin is still down: 4.8% against 5.1%. One quarter of stability does not retire the erosion argument, and the quarter contained the same disclosure that appeared in Q1, in the same unsized form.

"We also experienced a pull forward of value previously expected to occur in the second half of the year. After adjusting for the pull forward, our underlying results were in line with our expectations as outperformance across the broader Caremark business, including higher specialty generic penetration rates, offset the pressure in 340B." — Brian Newman, EVP and CFO

Note the construction: adjusted for a pull-forward of unstated size, the segment was "in line," and the outperformance elsewhere was fully consumed offsetting 340B. The full-year guide of at least $7.25B was reiterated rather than raised, which leaves at least $4,028M for the second half against $3,973M in the comparable period of 2025, or 1.4% growth. That is the same arithmetic we set out after Q1, now with one quarter of it delivered.

Assessment: the near-term data improved and the forward outlook deteriorated in the same forty minutes. Gross margin stopped falling, volume grew, and management then disclosed that Caremark will lose members in 2027 and that 340B is a 2027 headwind it will not size. For a segment whose bear case is structural erosion, a quarter that stabilises the present while pre-announcing pressure in the next year is not a rebuttal.

Pharmacy & Consumer Wellness

This was the quarter's cleanest result and the least discussed. Revenue of $33,816M grew only 0.7%, but gross profit of $6,534M grew 8.4% and gross margin expanded 140 basis points to 19.3%. Operating expenses fell 3.2%. Adjusted operating income of $1,475M grew 10.2%, and the segment's adjusted operating margin reached 4.4% against 4.0%. Prescriptions filled grew 4.3% to 457.0 million on a 30-day equivalent basis; same-store prescription volume grew 7.0%; same-store front store sales grew 1.0%. The full-year guide moved to at least $6.4B from at least $6.18B, an increase of $220M, which is more than double the $90M raise the segment received after Q1.

"You can already see this in the second consecutive year of mid-single-digit growth reflected in our updated guidance." — Brian Newman, EVP and CFO

The gap between 0.7% revenue growth and 4.3% script growth is the same regulatory price reduction, generic introduction and reimbursement pressure that has defined this segment for three years. What changed is that the gross margin line finally moved, and the segment head attributed it to the pricing model rather than to mix.

"This is also helping us get to a more consistent margin profile in this business." — Prem Shah, CVS Health

Assessment: after Q1 we wrote that the volume signal was good but that nothing yet showed volume converting to profit while reimbursement compressed. This quarter shows it: 140 basis points of gross margin expansion, a 10.2% earnings gain, and a guidance raise more than twice the size of the last one. Retail is now the most clearly evidenced pillar of the whole thesis, and it is the one the market spends the least time on.

Corporate/Other

The adjusted operating loss of $477M widened $64M year over year but narrowed $100M sequentially from Q1's $577M. For the half the loss is $1,054M against $743M, a deterioration of $311M, and the residual implied by the full-year segment guides puts the second half at roughly $1,046M more, or about $2.1B for the year against $1,687M in 2025. The CFO acknowledged the line as the destination for investment spend.

"We had some investments in capabilities and technology, and that continues. We have it across our business segments. We've been investing in colleagues and capabilities at PCW for some time, and we think we're seeing the fruits of that labor. We continue to do it. You saw some of it pop up in the corporate segment this past quarter." — Brian Newman, EVP and CFO

Assessment: a roughly $400M year-over-year deterioration in the corporate line is now a standing offset to the segment recovery, and it is the only place in the P&L where the unquantified technology programme is visible. It absorbed a small fraction of this quarter's segment-level improvement and drew a single question.

Key KPIs

Medical membership (000s)

After Q1 we flagged the membership table as the quietest important disclosure in the release, because CVS was shrinking the risk pool while improving the margin on what remained, which works until it has to grow again. This quarter is the first sign it can. Total medical membership rose 18 thousand sequentially, and it did so in the three lines that carry medical risk: Medicare Advantage up 27 thousand, Medicaid up 21 thousand, Commercial insured up 25 thousand. The fee-based ASC book, which carries no medical risk, fell 39 thousand. Year over year the picture is still a 698 thousand decline dominated by the 1,121 thousand loss of Commercial insured lives from the individual exchange exit, but the sequential direction inverted, and Aetna's leadership described the underlying dynamic without prompting.

"We've seen less contraction in our membership than we expected. favorable member mix. We have leading star scores." — Steven Nelson, Aetna

Operating KPIs

Key Topics & Management Commentary

Overall Management Tone: confident on the quarter and deliberately, unusually forthcoming about the next year. Management chose to bring forward guidance commentary it was not required to give and used it to introduce two headwinds and a number equal to consensus, which is the behaviour of a team managing expectations down rather than defending them. Where the posture tightened was on sizing: three separate items material to 2027, the 340B headwind, the Caremark membership decline and the second-half pull-forward, were named and then explicitly left unquantified. Tone on the printed quarter was the most assured of the past five, and tone on the forward year was the most guarded.

1. Management pre-announced 2027 a quarter early, and that is why the stock fell

CVS normally discusses next-year headwinds and tailwinds on its third-quarter call and gives detailed guidance on the fourth. This year the CFO moved that conversation forward by a quarter, unprompted, and framed the decision as a transparency commitment.

"This year, in the spirit of this commitment, we are pulling forward our commentary on 2027 by a quarter as we wanted to discuss some dynamics in our businesses that we are aware of today." — Brian Newman, EVP and CFO

The substance of the pull-forward was a number and two headwinds. The number:

"While we would not normally comment on 2027 consensus this early in the year, an outlook of at least $8.44, consistent with current consensus appears reasonable at this juncture. This represents EPS growth of about 13% off an adjusted baseline of $7.46." — Brian Newman, EVP and CFO

Read that construction carefully, because two things are happening at once. The $8.44 is equal to where the Street already sat, so the best case management would volunteer for 2027 is no better than what was already in models. And the 13% growth rate is measured off $7.46, not off the $8.00 midpoint of the guidance the company had raised minutes earlier. The $0.54 difference is prior-year development and prior-year individual-exchange items, which management strips out by convention. Against the reported 2026 midpoint, $8.44 is growth of 5.5%.

Assessment: this is a company telling the market, voluntarily and three months early, that the earnings it just raised are worth $0.54 less than they look and that next year grows single digits on the reported base. The transparency is genuinely creditable and the content is genuinely negative, and a market that had run the stock up 64.6% in twelve months resolved the two in the obvious direction.

2. The composition of the Aetna margin gain, and what cannot repeat

The bridge is set out in full above; the point that belongs here is what the largest component was. The individual-exchange risk adjustment true-up relates to the 2025 plan year of a business CVS exited at the start of 2026. It is not a recurring seasonal benefit; it is the final settlement of a book that no longer exists. The prior-year development component was roughly $100M, being the $1.2B disclosed for the first half less the $1.1B disclosed for the first quarter, which is a sharp step down from the roughly $300M in the comparable quarter of 2025 and the $1.6B in the first quarter of that year.

Management's confidence statement on reserves was the same single sentence it has used for two quarters.

"We remain confident in the adequacy of our reserves." — Brian Newman, EVP and CFO

Assessment: the declining scale of prior-year development is, on balance, good news for reserve quality, and days claims payable moving above the prior-year level supports that. What it also means is that the earnings tailwind from development is fading fast, and the one-off that replaced it this quarter is by construction terminal. Both facts point to the same place: 2027 has to be carried by trend, which is exactly what management is now saying.

3. The full-year guide embeds a second-half Health Care Benefits loss

The Health Care Benefits full-year range moved to $5.03B to $5.37B from $4.00B to $4.34B, an increase of $1.03B that the CFO attributed to two things.

"This increase reflects a portion of our strong underlying core performance in the first half of the year as well as the approximately $500 million cumulative impact of our net risk adjustment update and the prior year development experienced in the second quarter." — Brian Newman, EVP and CFO

The segment has already produced $5,467M of adjusted operating income in the first half. The entire guided range sits below that figure. Subtracting gives an implied second-half range of negative $437M to negative $97M, against an actual second-half loss of $362M in 2025. The medical benefit ratio path management described makes the same point from the other side.

"We now expect the increase between our first quarter and fourth quarter MBR in our Health Care Benefits business to be slightly higher than 950 basis points after adjusting for the impact of prior year development in the first quarter." — Brian Newman, EVP and CFO

Adjusting Q1's 84.6% for its $1.1B of development on $33,792M of premiums gives roughly 87.8%, so the guide implies a fourth-quarter medical benefit ratio of about 97.3% or higher. Solving the full-year 89.75% guide against the 86.0% delivered in the first half puts the second half at roughly 93.4% across any plausible premium assumption.

Assessment: this is Part D seasonality under the Inflation Reduction Act working exactly as it did in 2025, when the segment lost $676M in the fourth quarter alone. It is not new and it is not a warning. What is notable is that the full-year improvement of roughly $2.26B at the midpoint against 2025 is entirely a first-half phenomenon: $2,166M of it was delivered in the first six months, and the guide asks for about $95M more across the remaining two quarters. An investor extrapolating the first half into a run-rate is extrapolating something management has explicitly declined to guide.

4. Caremark will lose members in 2027, and the retention rate is already below normal

The second element of the pull-forward was the more damaging of the two, because it goes to volume in the segment that carries the largest share of enterprise profit.

"In our Pharmacy Services business, we expect the previously discussed market dynamics in our 340B business to continue and result in a headwind in 2027. Additionally, we believe we will see membership declines in Caremark next year." — Brian Newman, EVP and CFO

The segment head gave the mechanism, and the contrast with the prior selling season is stark. Last year's pipeline produced over $6B of new sales, which management called meaningfully above its historical average. This year's posture is different by choice.

"As we went into 2027, we took a disciplined and prudent approach to the selling season to ensure that we had the right underwriting and contracts in place for the new business and renewals that we had. And at this point in the year, we're trending to a retention rate that's slightly lower than our historical performance, but more consistent with industry norms." — Prem Shah, CVS Health

Two distinct causes were named: CVS declining business it does not want at the price on offer, and clients leaving markets of their own accord.

"Second, product actions and market exits by some of our health plan customers will result in a membership impact." — Brian Newman, EVP and CFO

Assessment: a retention rate reverting to "industry norms" from a position above them is a euphemism for losing share, whatever the reason. The disciplined-underwriting framing is credible and may well be correct, but a business whose claims volume was flat for a year, has just returned to 0.9% growth, and is now guiding membership down in the following year has a volume problem that pricing discipline does not solve. The offset management named, specialty pharmacy and the 2027 generic wave, is real and was cited three separate times, which is itself a signal about how much weight it is being asked to carry.

5. 340B returned as a headwind and management refused to size it

340B pressure appeared in the quarter and was fully absorbed by strength elsewhere in Caremark, and then reappeared as a stated 2027 headwind. The mechanism is manufacturer behaviour rather than anything CVS controls.

"Pharma manufacturers are imposing restrictions on covered entities that led to some of the impact that we're seeing in the second quarter." — Prem Shah, CVS Health

A second driver was volunteered in the same answer: as large specialty drugs go generic, the economics of the programme compress. Asked directly for a number, the CFO declined.

"We continue to feel confident in our full year outlook for HS. However, as we discussed, we do expect 340B pressure to be a headwind next year." — Brian Newman, EVP and CFO

Assessment: the honest read is that this is a genuine external headwind of unknown magnitude in a segment already guiding membership down. The company's own framing, that 340B was "fully offset by broader Caremark performance" this year, is a statement that the offsets have already been spent. Going into 2027 the same offsets have to cover a larger headwind plus membership attrition.

6. Retail is now the enterprise's best story and nobody is telling it

Pharmacy & Consumer Wellness delivered 140 basis points of gross margin expansion, 10.2% adjusted operating income growth, 4.3% script growth, 7.0% same-store script growth and a $220M guidance raise, and it consumed a small fraction of the call. The segment head's framing tied the result to the pricing architecture rather than to one-off mix.

"we're still the best run national pharmacy in the country" — Prem Shah, CVS Health

The Rite Aid file acquisitions, which closed in the third quarter of 2025, are still a tailwind and were acknowledged as such, which means they begin to lap from here.

"Obviously, the Rite Aid was an incremental tailwind as you think about this year and the first half of this year." — Prem Shah, CVS Health

Assessment: the CostVantage transition, which we scored as complete after Q1 without evidence that it converted volume to profit, has now produced that evidence. This is the single most improved pillar of the thesis and the least reflected in the discussion. The watch item is that same-store script growth of 7.0% has to survive the Rite Aid lap, which begins in the third quarter.

7. GLP-1s: from a managed cost to a distribution business

The CEO opened the call on GLP-1s rather than on the numbers, which is a choice. The strategic content is a pivot from managing the category as a trend problem inside Caremark and Aetna to selling into it directly through retail and MinuteClinic, and the pricing was disclosed in detail: a $29 virtual weight-management visit, cut from $49 launched in the second quarter, and cash-pay therapy access "for as little as $149." A new Eli Lilly partnership joins the existing Novo relationship, and CVS Pharmacy participates in the CMS Medicare GLP-1 Bridge programme through the end of 2027 at $50 per month for eligible beneficiaries.

The segment head was unusually candid about how CVS got here.

"we may have been slightly slow to enter the cash paying market or direct-to-patient market" — Prem Shah, CVS Health

Assessment: this is the clearest example of the enterprise assembling its own assets around a single consumer need, and it converts a category that has been a pure cost headwind at Caremark into a volume opportunity at retail. It is also a low-margin, high-velocity dispensing business by construction, exactly like the rest of the pharmacy franchise, and management gave no revenue or margin framing for it. Treat it as support for the retail script trajectory rather than as a new profit pool.

8. The cash flow inflection is the quarter's most underrated fact

First-half cash flow from operations of $10,594M is up 64.2% from $6,453M, and the full-year guide moved to at least $11.5B from at least $9.5B, an increase of $2B. Working capital contributed $2,200M against $1,118M. Total debt fell to $61,410M from $64,570M at year end, $3,287M of long-term debt was repaid, and the leverage ratio came down materially.

"Our leverage ratio at the end of the quarter was approximately 3.5x, and we expect to drive further improvement as we continue to execute against our 2026 outlook." — Brian Newman, EVP and CFO

After Q1 we identified cash conversion as an active bear point: operating cash flow had fallen 6.7% year over year and the raised full-year guide of at least $9.5B still sat below the $10.6B generated in 2025 while adjusted earnings were guided roughly 10% higher. That objection is now answered. The first half alone has produced essentially the whole of 2025's full-year cash flow.

Assessment: this is the cleanest reversal of a bear point in the quarter and it removes the balance-sheet question from the thesis. It also makes the new guide conspicuously conservative: at least $11.5B for the year against $10,594M already delivered implies roughly $906M in the second half, against $4,147M in the second half of 2025. Either the second half is genuinely a working-capital give-back of that scale, or the floor is a placeholder. Neither was addressed.

9. Capital allocation: better balance sheet, same answer on the buyback

Leverage improved from 3.84x to roughly 3.5x, cash at the parent and unrestricted subsidiaries stood at approximately $2.7B, and $1,725M of dividends were paid in the half. No shares were repurchased, and the answer on when that changes was the same as last quarter, in fewer words.

"As a reminder, our current outlook does not assume any share repurchases this year. We will continue to evaluate capital deployment opportunities as our leverage position improves." — Brian Newman, EVP and CFO

The forward framing was tighter still. Asked directly whether the $8.44 figure for 2027 embeds any repurchase, the CFO confirmed it assumes only enough to offset dilution.

Assessment: the deleveraging is real and fast, and the company has now put itself within reach of a rating-consistent target. What it has not done is commit to returning anything beyond the dividend, in either 2026 or 2027, which means diluted share count remains a headwind to per-share growth for at least six more quarters. For a stock at 12.4x with a 2.7% dividend yield, the absence of a buyback is the difference between a mid-single-digit and a high-single-digit 2027 per-share outcome.

10. Technology and AI: a $20B decade commitment and a $1B savings claim, neither reconciled

The technology programme received more airtime than any topic other than 2027. The CEO restated the strategic framing and, for the first time on an earnings call this year, put a figure against it.

"we also made a commitment a year ago that we were going to invest $20-plus billion in our technology efforts over the next decade" — J. Joyner, Chairman and CEO

Health100 moved from announcement to launch, with an AI assistant attached.

"Last month, we began the targeted launch of our Health100 platform, including Haio, our new AI-powered assistant." — J. Joyner, Chairman and CEO

The CFO closed the discussion with the only benefit figure disclosed: over $1B of operating expense savings generated over the last few years through technology efficiencies and AI. Against that, first-half capital expenditure of $1,540M is up 14.1%, and the corporate segment's loss widened.

Assessment: the $20B over a decade works out to roughly $2B a year against a company that spent $1,540M on all capital expenditure in six months, so the commitment is either mostly within existing spend or represents a genuine step-up that has not been separately disclosed. Management has not said which, and the operating benefit is quoted as a cumulative historical figure rather than a forward run-rate. The strategic logic remains sound and the disclosure remains inadequate.

11. Medicare Advantage bids, Part D, and the 2027 rate environment

Aetna's leadership described the Medicare book as running ahead of plan in 2026 and took the same posture into the 2027 bids, assuming trend stays elevated rather than assuming relief.

"As I think about '27, as we contemplated the bids, we're taking that same focus to return the business to target margins and the same discipline. We assumed a continuation of the elevated trend as we put our bids in. And we think this business, combined with our leading star scores is going to be well positioned to make that continued progress towards the appropriate margins in '27, but also delivering a great experience and good health outcomes for our members." — Steven Nelson, Aetna

Group Medicare Advantage renewals are about 75% complete for 2027 pricing. Part D was described as performing in line after two years of deliberate de-risking and portfolio simplification, with the same discipline carried into the 2027 bid. The commercial book was described as strong and in line, with the No Surprises Act dispute-resolution process singled out as a cost-trend irritant that Aetna says it priced for.

Assessment: note what changed in the language. Last quarter Aetna's leadership reaffirmed "target margins in 2028" explicitly and unqualified. This quarter the destination was described by the CFO as "our pathway back to target margins over the next couple of years" and by the segment head as "continued progress towards the appropriate margins in '27." The commitment has not been withdrawn, and it has been stated less precisely. In a quarter where management pre-announced 2027 in order to be transparent, the one number it did not re-anchor is the one the whole recovery narrative rests on.

Guidance & Outlook

Every line moved the right way and two moved a long way. Lowering the full-year medical benefit ratio guide 75 basis points while halving the range around it is a stronger statement than the $0.60 earnings raise, because it is a claim about trend rather than about a period's results. The $2B cash flow increase is the largest single revision in the release.

The implied second half

Subtracting the reported first half from each full-year figure gives the shape management is actually guiding to, and it is the most informative table in this note.

First-half adjusted EPS grew 27.1% year over year. The guide implies the second half grows 5.6%. First-half revenue grew 6.7%; the revenue floor implies the second half declines 0.5%. Neither of those is a warning, because the "at least" formulations are floors and the Part D seasonality is structural and known. But the shape is unambiguous, and it is the same shape the 2027 commentary describes. Management is guiding to a business whose year-over-year improvement is concentrated in the six months already reported.

Implied quarter-over-quarter ramp: first-half adjusted operating income of $10,307M is 61.5% of the $16.75B full-year midpoint, consistent with the roughly 60-40 weighting described at Q1. Management added that within the second half, "we expect second half EPS to be more weighted to the third quarter, reflecting typical seasonality," which on 2025's split points to roughly $1.69 in the third quarter and roughly $1.15 in the fourth against the $2.84 second-half midpoint.

Street at: the pre-print consensus of $1.85 for the quarter was, on the evidence, management's own 60-40 phasing arithmetic applied to the old guide. The new $7.90 to $8.10 range sits 18.5% above 2025 adjusted EPS of $6.75. The number that now matters more is $8.44 for 2027, which management has confirmed rather than exceeded.

Guidance style: conservative, and demonstrably so this quarter. The full-year raise of $1.05B in adjusted operating income against a first half that beat by rather less, combined with a cash flow floor that implies $906M in the second half against $4,147M last year, both indicate a guide set to be beaten. The countervailing evidence is that management chose to introduce a consensus-level 2027 anchor in the same breath, which is not the behaviour of a team sandbagging into a beat.

The 2027 Framework

The single most consequential disclosure of the quarter deserves its own arithmetic. Management's framework has three components: a 2027 floor of $8.44, a 2026 baseline of $7.46, and a reaffirmed mid-teens adjusted EPS compound annual growth rate from 2025 through 2028.

Now carry it forward. The mid-teens CAGR commitment runs from the reported 2025 actual of $6.75 to 2028, and management reaffirmed it explicitly on this call. What that requires of 2028, given an $8.44 anchor in 2027:

The commitment is not broken by the $8.44 figure, but it is made steeply back-end loaded by it. A 2027 that grows 5.5% on the reported base requires a 2028 that grows roughly 18% to 22% to land in the middle of a mid-teens CAGR, in a year with no disclosed catalyst beyond Aetna reaching target margins. Management called $8.44 a floor, twice, and there is a coherent case that it will be beaten in the way the 2026 guide has been beaten twice. But the company has now told the market that the 2026 number contains $0.54 it will not carry, and it has done so before the Street had marked its own models.

Assessment: the mid-teens CAGR now depends on a 2028 doing more than double the work of 2027. That is a materially different underwriting proposition from the one available in May, and it is the correct reason for the stock to have fallen on a quarter this good.

Analyst Q&A Highlights

Seven questions were taken. The call ran barely over an hour and management ended it after the seventh.

How Health Services offsets the 340B headwind into 2027

The first question of the call went straight at the pull-forward commentary, asking what levers are within management's control and how a headwind whose policy trajectory is still unsettled gets mitigated. The answer contained the number that moved the stock, delivered as a preface rather than as the substance of the reply.

Q: "As you think about the dynamics of managing to some level of growth trajectory, how do you feel about the opportunities, whether it's conversion on formulary, whether it's specialty dynamics that are within your control? And how do you think about the pathway forward to offset and mitigate, in particular, the 340B headwinds given that it does seem to be a moving target within D.C. right now?"
— Michael Cherny, Leerink Partners

A: "I think a reasonable floor, Mike, for '27 for adjusted EPS should be -- could be $8.44, and that's obviously consistent with current consensus."
— Brian Newman, EVP and CFO

Assessment: the question asked about segment levers and the answer led with an enterprise number equal to consensus. The stock reversed from a premarket gain during this stretch of the call. Note also the hedged construction, "should be -- could be," on a figure management had just described in prepared remarks as merely appearing "reasonable at this juncture."

Medicare Advantage and Part D positioning into the 2027 bid cycle

A recurring theme of the call was whether the Aetna recovery survives contact with a second consecutive elevated-trend year. The answer was that trend relief was not assumed in the bids, which is the conservative posture and also concedes that the rate environment is not doing the work.

Q: "You mentioned that Medicare Advantage was better. But as we think about what Brian talked about for 2027, just your thoughts on the way that you bid for MA in '27, thoughts on Medicare Part D in '27 and how you see that playing out?"
— Lisa Gill, JPMorgan

A: "As I think about '27, as we contemplated the bids, we're taking that same focus to return the business to target margins and the same discipline. We assumed a continuation of the elevated trend as we put our bids in. And we think this business, combined with our leading star scores is going to be well positioned to make that continued progress towards the appropriate margins in '27, but also delivering a great experience and good health outcomes for our members."
— Steven Nelson, Aetna

Assessment: pricing to elevated trend rather than to hoped-for relief is the right discipline and is consistent with the membership stabilisation this quarter. The wording is the tell. Last quarter's answer to the same line of questioning ended with an explicit reaffirmation of target margins in 2028; this one ends at "continued progress towards the appropriate margins in '27." Same direction, less commitment.

Whether any capital return sits inside the 2027 floor

The question that most cleanly isolates the quality of the $8.44 figure was asked immediately and answered in two sentences. It matters because a number built with buybacks is a different number from one built without.

Q: "First, a quick follow-up on '27. Is there any share repo or other capital deployment assumed in that [ $8.44 ] -- or is that still outside the guide similar to the Investor Day framework?"
— Justin Lake, Wolfe Research

A: "Similar to what we talked about at Investor Day, we are only assuming offsetting dilution in '27, Justin. So nothing beyond that at this point."
— Brian Newman, EVP and CFO

Assessment: the most constructive fact in the 2027 framework. The $8.44 is a pure operating number with no financial engineering, on a balance sheet that just delevered to roughly 3.5x and generated $10.6B of operating cash flow in six months. If capital return is switched on, it is additive to the floor rather than embedded in it. That is the strongest available argument that $8.44 is genuinely a floor.

Durability of retail share gains as the Rite Aid contribution laps

The second half of the same question probed whether the retail momentum is structural or an acquisition artefact that fades, and whether cost-plus pricing stabilises gross margin per script. The answer tied the result to a three-part operating programme and, unusually for this company, to the pricing model as a margin stabiliser rather than as a margin cap.

Q: "We know you're benefiting at the moment from script share gains from Rite Aid and others that might normalize over time. So I'm curious on your thoughts of the sustainability of the strong PCW growth as share gains might normalize going forward."
— Justin Lake, Wolfe Research

A: "This is also helping us get to a more consistent margin profile in this business. And as you look at 2026, we took a prudent approach to how we thought about this, and we're excited about the encouraging progress that we're making relative to our expectations."
— Prem Shah, CVS Health

Assessment: the 140 basis points of gross margin expansion this quarter is the evidence the answer needed and it was not cited in the reply. The reasonable read is that CostVantage has moved retail from a structurally deflating gross margin to a stable one, and that the Rite Aid files are a volume overlay on top. The lap begins in the third quarter, and same-store script growth of 7.0% is the metric to watch, because it excludes acquired files by construction.

Commercial risk margins and the dispute-resolution cost overhang

One line of questioning went to the least-discussed part of Aetna, the commercial book, and to a cost driver that has been an industry irritant for two years. Management characterised the book as in line and priced for, and was notably direct about where the pressure originates.

Q: "I was hoping to get an update on the performance of the commercial risk business. Can you update us on how the margins in your guidance this year compared to what you target long term? And then additionally, how the -- no Surprises Act and independent dispute resolution process is impacting your cost trend and margins this year and maybe some context on how that compares to what you saw last year."
— Stephen Baxter, Wells Fargo

A: "With respect to IDR, look, this is one component of trend. We've seen it. We've been watching it. We've accounted for it. So it's not catching us off guard. And we've taken actions. We've gone upstream with providers trying to bring certain providers into the network at reasonable rates so we can get ahead of this dispute resolution."
— Steven Nelson, Aetna

Assessment: the question asked how commercial margins in the current guide compare to the long-term target and that half was not answered at all. The IDR half was answered fully and credibly, including the observation that the process "is clearly being abused by a small group of players." The commercial book is a $6.6B quarterly premium line shrinking 19.7% on the exchange exit while its fee-based sibling grows, and its margin position remains undisclosed.

What is driving the 340B pressure, and how large it is

A pointed follow-up noted that 340B had not been called out as a headwind since 2024 and asked what changed. The mechanism was explained clearly. The magnitude was declined, explicitly, on the record.

Q: "I wanted to follow up on the 340B headwind. I believe this is the first time you've called this out since 2024. So I'd love to better understand what's driving the incremental pressure now? Is this driven by claims penetration, manufacturer contracting or reimbursement changes?"
— Andrew Mok, Barclays

A: "not commenting on 340B sizing specifically, but we're building up the plans as we go right now"
— Brian Newman, EVP and CFO

Assessment: a headwind material enough to be pre-announced a quarter early, and to be given as one of two reasons Health Services faces pressure in 2027, but not material enough to size. The two are hard to hold together. The segment head did supply the mechanism, manufacturer restrictions on covered entities plus large specialty drugs going generic, both of which are external and neither of which CVS can manage away.

What the first-half pull-forward was, and whether it creates second-half room

The final question tried to convert the pull-forward disclosure into something usable, asking what the accelerated value actually consisted of and whether it frees capacity later in the year. The reply moved the conversation to investment spending and closed by pointing back at the guide.

Q: "You had talked a lot about the value creation activities being pulled forward in the first half of the year. I know you talked about a little bit about the technology side, but anything you can talk about in terms of like what that was operationally? And does that give you a little bit more space to do additional things that you hadn't contemplated in the back half of the year?"
— Elizabeth Anderson, Evercore ISI

A: "So you'll continue to see investments in the business, but that's captured in the full year beat and raise. And obviously, we put a floor on '27. So we have the investments currently that we've modeled in there."
— Brian Newman, EVP and CFO

Assessment: the pull-forward has now been disclosed in two consecutive quarters and sized in neither. In Q1 it was value pulled from Q2; in Q2 it was value pulled from the second half. The answer here does not describe what was accelerated, and the reply's substance is a reassurance that investment is inside the guide. This is the second-least satisfying exchange of the call after the 340B sizing refusal, and the two are related: both concern how much of the second half has already been consumed.

What They're NOT Saying

1. How the ~$500M of prior-year items splits: management gave a combined figure for the individual-exchange risk adjustment and the prior-year development but never separated them. The split matters because the risk-adjustment component relates to the 2025 plan year of an exited business and therefore has no 2027 analogue, while development at least in principle recurs. The filing's six-month development figure of $1.2B against the first quarter's $1.1B implies the development piece was roughly $100M, meaning roughly $400M is terminal. That inference is ours; the company declined to make it.
2. Whether the core medical benefit ratio improved year over year at all: the prepared remarks say core performance "still exceeded our expectations," which is a comparison to plan. No comparison to the prior year on a like-for-like basis was offered, and on our bridge the improvement is roughly 60 basis points against 250 reported.
3. The implied second-half Health Care Benefits loss: the raised full-year range of $5.03B to $5.37B sits entirely below the $5,467M already delivered. That the guide embeds a segment loss of $97M to $437M in the back half went unstated by management and unasked by seven analysts.
4. The size of the 340B headwind: refused on the record. Named as one of two reasons for 2027 pressure in the segment that generates the most adjusted operating income.
5. The magnitude of the Caremark membership decline: "membership declines" with no percentage, no lives figure, no revenue or earnings translation. The retention rate is characterised as "slightly lower than our historical performance" without either rate being given.
6. The size of the second-half pull-forward at Health Services: disclosed in prepared remarks, raised again in Q&A, unsized both times. Second consecutive quarter with an unsized pull-forward, which compounds: Q1 pulled value from Q2, and Q2 pulled value from the second half.
7. What 2028 has to be: the mid-teens CAGR from 2025 was reaffirmed and $8.44 was set for 2027, and the two together require roughly 18% to 22% growth in 2028. That arithmetic was never put on the table, and nobody asked management to reconcile the reaffirmation with the anchor.
8. Where the Medicare book's profitability now sits: the CEO volunteered last quarter that Medicare generated an adjusted operating loss in 2025. This quarter Medicare was described as "ahead of our expectations" with no absolute position given, no definition of target margins in basis points, and no statement of whether the book is now profitable.
9. Why the effective tax rate is running near 24.8%: second consecutive quarter with a rate roughly 1,000 basis points below the prior-year period, unaddressed in both the release and on both calls. On $7,915M of first-half pre-tax income the difference is material to the GAAP result.
10. Why the cash flow floor implies a second half of roughly $906M: the at-least-$11.5B guide against $10,594M already banked leaves less than $1B for six months, against $4,147M in the second half of 2025. Either that is a large working-capital reversal nobody described, or the floor is a placeholder. The $2B raise was attributed to earnings and working capital without reconciling to what the back half then requires.
11. Whether the $20B decade technology commitment is incremental: quoted for the first time on an earnings call this year, against first-half capital expenditure of $1,540M for the entire company. No split between capitalised and expensed, no annual cadence, and the only benefit figure given is a cumulative historical "over $1 billion in OpEx savings."

Market Reaction

• Pre-print setup: CVS closed at $104.42 on August 4, up 31.6% year to date, 2.3% over the trailing 30 days and 64.6% over the trailing twelve months, against a 52-week closing range of $61.45 to $110.60. The S&P 500 was up 13.0% year to date. This was among the best-performing large-cap names in the market going into the print, not a washed-out one.
• Reaction session (August 5, the print day, reported before the open): opened at $93.72, a gap of 10.2%, traded a range of $93.50 to $100.55, and closed at $99.12, down 5.1% or $5.30. The S&P 500 closed down 0.2% on the same session.
• Volume: 18.5M shares against a 7.5M 30-day average, or 2.5x normal.
• Intraday shape: the low of $93.50 came early and the close of $99.12 recovered 6.0% off it. The session gave back roughly half the opening gap.
• Positioning outcome: the close still leaves the stock up 24.9% year to date and 14.1% above the $86.86 close that followed the first-quarter print.

The mechanics are unusually legible. The press release crossed before the open with a 40% earnings beat, a $9B revenue raise, a $0.60 earnings raise and a $2B cash flow raise, and the stock traded up on it. The call began at 8:00 a.m. ET, the CFO delivered the 2027 commentary in prepared remarks, took the first question with the $8.44 figure, and the stock opened 10.2% lower. Nothing in the printed quarter caused the decline. The decline was caused by a passage of voluntary disclosure about a year the company was not required to discuss.

What repriced was the terminal value of the recovery, not the recovery. A market that had taken the stock up 64.6% in twelve months was underwriting the first-half run-rate as a base for 2027 and beyond. Management said that base contains $0.54 it will not carry, that Caremark loses members next year, and that the best case it would name for 2027 was the number already in models. The 6.0% recovery off the intraday low is the market working out that a floor of $8.44 with no buyback assumed, on a balance sheet delevering this fast, is a different object from a ceiling of $8.44.

The setup detail that constrains the rating is the one that also constrained it in May, in a more expensive form. The stock entered this print having compounded 64.6% over a year on a recovery whose reported margin improvement is, on our bridge, roughly a quarter as large as it looks. A 5.1% drawdown against that does not create value; it removes a little froth.

Street Perspective

Debate: Is $8.44 a floor or a ceiling?

Bull view: management called it a floor twice, and this is the same team that has beaten and raised its 2026 guide in consecutive quarters by $0.30 and then $0.60. The figure assumes no share repurchase beyond dilution offset, on a balance sheet that just delevered to roughly 3.5x with $10.6B of first-half operating cash flow, so capital return is pure upside to it. The stated guidance philosophy is explicitly designed to produce outperformance against credible targets.

Bear view: a company that volunteers a number equal to consensus, three months before it has to, in the same breath as two new headwinds, is managing expectations down. The $7.46 baseline is management conceding that $0.54 of the 2026 guide is non-recurring, and the largest single component of this quarter's beat relates to a plan year of a business that no longer exists. Off the reported base, $8.44 is 5.5% growth.

Our take: both readings are defensible and the buyback point is the strongest thing the bulls have. Our resolution is that $8.44 is probably a floor for the reported number and roughly a fair estimate of the underlying one, which is the same thing said twice: the growth is real and smaller than 2026's headline implies. The decisive fact is that management chose to say this early, and companies with upside they intend to deliver do not usually pre-empt their own third-quarter call to anchor the following year at consensus.

Debate: Has the Aetna recovery already happened?

Bull view: more than $2B of year-over-year adjusted operating income improvement in six months, medical membership stabilising sequentially for the first time since the exchange exit, days claims payable now above the prior-year level, the full-year medical benefit ratio guide cut 75 basis points with the range halved, and bids for 2027 written assuming elevated trend continues. Every one of those is a leading indicator, not a lagging one.

Bear view: roughly 140 basis points of the 250 basis point margin improvement is 2025-plan-year items, and the largest piece cannot recur because the business generating it was exited. The full-year guide asks for about $95M of further year-over-year improvement across the back half after $2,166M in the front half. Management describes the destination as "over the next couple of years" where a quarter ago it named 2028 explicitly.

Our take: the operational evidence is the strongest it has been, and it is genuinely stronger than the accounting evidence. Membership stabilisation and a rebuilt days-claims-payable position are hard to manufacture and both moved the right way. What the bears have correctly identified is that the reported earnings power of this segment has been overstated by legacy-year items for two consecutive quarters, and the second-half guide is management agreeing. We would characterise the recovery as real, roughly 60% complete on management's own target-margin framing, and materially less far along than the reported margin suggests.

Debate: Does the Caremark membership decline break the mid-teens CAGR?

Bull view: the membership decline is chosen, not suffered. CVS is declining business that does not clear its underwriting standard after a selling season that brought in over $6B of new sales the year before, and it is offset by a specialty pharmacy franchise entering a strong generic year. Health Services adjusted operating income grew 10.0% this quarter with gross margin flat, which is the first evidence in this cycle that scale is holding.

Bear view: the segment generates the largest share of enterprise adjusted operating income, its claims volume was flat for a year and grew 0.9% this quarter, and management has now guided its membership down for the following year while refusing to size a simultaneous 340B headwind. A retention rate reverting to industry norms from above them is share loss regardless of the reason, and the 2027 offsets have to cover a bigger headwind than the ones that consumed this year's outperformance.

Our take: the bears are describing the more probable path. The disciplined-underwriting explanation is credible and would be more credible if the same call had not also cited client market exits as a second, entirely involuntary cause. A pharmacy benefit manager losing lives into a year with a growing external headwind is not a business that carries an 18% to 22% 2028. If the mid-teens CAGR holds, it holds because Aetna reaches target margins and retail keeps compounding, with Caremark contributing less than the 2028 framework originally assumed. That is a materially more concentrated bet than the one on offer in May.

Model Update Needed

Valuation impact: at the August 5 close of $99.12 and the second-quarter diluted share count of 1,287M, market capitalisation is roughly $127.6B. Total debt is $61.4B ($59,452M long term plus $1,958M current) against $11,329M of cash and equivalents, for net debt of about $50.1B and an enterprise value near $177.6B, with management stating leverage of approximately 3.5x. The stock trades at 12.4x the $8.00 guidance midpoint, 12.1x our $8.20 estimate, 11.7x management's $8.44 floor for 2027 and 11.5x our $8.65, and 14.7x 2025 adjusted EPS of $6.75. Dividends paid of $1,725M in the first half annualise to roughly 2.7% of market capitalisation, still with no repurchase alongside. We move our twelve-month fair value to $104 from $92, equivalent to 13.0x the 2026 guidance midpoint and 12.3x management's 2027 floor, about 4.9% above the August 5 close. The increase is entirely the higher earnings base and the newly disclosed 2027 anchor; the multiple we are willing to pay is unchanged, because the structural question that sets it has not moved and the segment carrying the largest profit pool has just guided its membership down.

Thesis Scorecard Post-Earnings

Scored against the standing thesis carried since our initiation, not against a fresh set of pillars.

Grading last quarter's commitments

Overall: the operating thesis is materially stronger than it was in May and the investment thesis is modestly weaker. Three of four bull pillars were confirmed, one of them decisively upgraded, and the bear point on cash conversion was largely retired. Against that, the two bear points concerning earnings quality and Health Services erosion both escalated a tag, the valuation pillar has been consumed by a 14.1% share-price move, and management volunteered a 2027 anchor equal to consensus while introducing two headwinds it declined to size. This was a good quarter for the company and a neutral-to-negative one for the stock, which is precisely what the tape said.

Action: Maintaining Hold with a conviction of 6 out of 10, up from 5, and a twelve-month fair value of $104, up from $92. The higher conviction reflects a narrower range of outcomes rather than a better one: management has now put a floor under 2027, removed the balance-sheet question, and told the market where the earnings quality is weak. After Q1 we named two conditions for an upgrade. The first, a second-quarter medical benefit ratio without unusual prior-year development, was not met: the ratio was excellent and roughly $500M of it was prior-year items. The second, two consecutive quarters of stable Health Services gross margin, is one quarter complete and was immediately followed by guidance for lower Caremark membership in 2027. Neither trigger fired, and the stock is 7.7% above the fair value we set in May. We would move to Outperform on a third-quarter medical benefit ratio inside the tightened full-year range with prior-year development below $200M, or on a capital-return announcement that converts the deleveraging into per-share compounding. We would move to Underperform if the 2027 headwinds and tailwinds detail at the third-quarter call places $8.44 at the top of the range rather than the bottom, which would make the mid-teens 2028 arithmetically unreachable.