CVS Health

IThe business

CVS Health is an American health company with four faces. It runs Aetna, an insurer that covers an estimated more than 37 million people. It runs CVS Caremark, a pharmacy benefit manager, or PBM, with about 87 million plan members. It runs about 9,000 retail drugstores and more than 1,000 walk-in and primary care clinics. And since 2023 it has owned two care businesses, Oak Street Health and Signify Health.² At the December 2025 investor day it said it has about 185 million connections with consumers and that 85% of Americans live within 10 miles of a CVS Pharmacy.⁶ It has more than 300,000 colleagues.⁶

The way I think about it: a person with a prescription touches the company three times. Aetna may be the insurer. Caremark may be the PBM that decides which drugs the plan covers and at what price. A CVS Pharmacy may be where the pill is picked up. In theory the company earns something at each step and sees the whole picture. In practice each step is regulated by a different part of government, and each is under a different kind of pressure.

Where the money is made

The 2025 10-K reports four segments. Health Care Benefits (Aetna) had revenue of $143.4B and adjusted operating income of $2.9B, 2.1% of revenue. Health Services (Caremark and the care businesses) had revenue of $190.4B and adjusted operating income of $7.2B, 3.8%. Pharmacy & Consumer Wellness had revenue of $139.4B and $6.0B, 4.3%. Corporate costs took $1.7B.² Add them and you get $14.4B of adjusted operating income on $402.1B of revenue after $71.6B of sales between the segments is removed C (calculated).² Reported operating income was only $4.7B, because of a $5.7B goodwill write-down on the care businesses and $1.2B of charges for two court decisions about past conduct.²

Look at what changed in three years. The two segments that touch the most daily commerce, Caremark and the pharmacies, earned $13.3B in 2023 and $13.2B in 2025 C (calculated).² Almost all of the group's profit swing came from one place, Aetna. That is the first thing to take from this page: when someone says CVS had a bad stretch, they mean Aetna.

Aetna

Aetna took in $134.7B of premiums in 2025: $103.4B from government programs and $31.4B from commercial plans.² Government means mostly Medicare, including Medicare Advantage, in which the government pays a private insurer a fixed amount per person, adjusted for how sick that person is. It also includes Medicaid and the standalone Medicare drug plans. At 30 June 2026 Aetna had 26.0M medical members, of whom 4.2M were in Medicare Advantage, 1.2M in Medicare Supplement, 2.3M in Medicaid, and 18.3M in commercial plans, 15.8M of them in self-funded plans where the employer carries the claims risk.⁵ Another 3.9M had a standalone Medicare drug plan, down from 4.9M at the end of 2024.⁵²

The number that tells you how Aetna is doing is the medical benefit ratio, or MBR: the share of premiums paid out as medical costs. It was 84.8% in 2021, 83.8% in 2022, 86.2% in 2023, 92.5% in 2024 and 91.2% in 2025.⁹² In the first half of 2026 it was 86.0%, with 84.6% in the first quarter and 87.4% in the second.³¹³ The 2025 figure includes two premium deficiency reserves, $448M on the individual exchange product in the first quarter and $471M on group Medicare Advantage in the second, which are charges for contracts CVS expected to lose money on.² Aetna has since left the individual exchange business, and its insured commercial membership fell from 3.4M at the end of 2025 to 2.5M by June 2026.⁵

Each point of MBR on $135B of premiums is about $1.35B of operating income, or about $0.80 a share after tax at 24% E (estimate). The gap between 2025's 91.2% and the 86.0% of the first half of 2026 is five points, though the first half is seasonally low. That is why a medical-cost surprise in either direction matters more than anything else on this page.

Caremark

Caremark is the PBM. It negotiates with drug makers for rebates and discounts, builds the list of covered drugs, runs a network of pharmacies, and fills prescriptions by mail and through specialty pharmacies. Health Services had $190.4B of revenue in 2025 and processed 1,900.7M claims on a 30-day-equivalent basis.² Its adjusted operating income of $7.2B works out to about $3.76 a claim C (calculated).² That is a small take on each prescription, and it is the take that the many laws and lawsuits about PBMs are aimed at. The 10-K says the company shares a larger portion of rebates with clients and offers pricing guarantees that cannot always be met.² The competitors it names are Express Scripts at Cigna, Optum Rx at UnitedHealth, Prime Therapeutics and MedImpact.²

Claims processed were 2,344.3M in 2023, 1,917.6M in 2024 and 1,900.7M in 2025.² That is an 18% fall in one year. I did not trace the cause in the filings I opened, so I note it and move on. In the first half of 2026 claims were flat at 937.7M, but adjusted operating income rose 1.4% to $3.2B, helped by purchasing economics and drug mix and held back by price concessions to clients.³

The drugstores

Pharmacy & Consumer Wellness had $139.4B of revenue in 2025. Of that, pharmacy was $115.5B, up 14.7%, front-store goods $21.5B, down 0.3%, and other $2.4B.² Front-store sales were $22.5B in 2023, so they are down 4% in two years while pharmacy rose 25% C (calculated).² Prescriptions filled rose from 1,649.1M in 2023 to 1,808.8M in 2025, up 9.7% C (calculated).² In the second quarter of 2026 prescriptions rose 4.3% and same-store sales 2.6%.³ Adjusted operating income was $6.0B in 2025, a margin of 4.3% against 5.1% in 2023, because the mix moved to prescriptions, which earn less per dollar.²

The shrink is the other story. The store count was 9,962 at the end of 2020 and 8,979 at the end of 2025.¹⁰² CVS closed 318 stores in 2023, 299 in 2024 and 243 in 2025, and opened or acquired 39, 39 and 87.² The 2025 additions include stores and prescription files bought from Rite Aid for $465M.² I read that as a company choosing to run fewer, busier stores: prescriptions per store, a rough measure, rose from about 175,000 in 2023 to about 201,000 in 2025 C (calculated).² The pharmacy has 1,870 locations inside Target stores, under a lease agreement with $2.3B of obligations.²

Oak Street and Signify

These are the two care businesses CVS bought in 2023, when its cash spent on acquisitions was $16.6B C (calculated).¹² Oak Street runs primary-care centers for older people and takes on the full medical bill for each patient in exchange for a fixed payment from Medicare Advantage plans or the government. At the end of 2025 it had 246 centers in 27 states and cared for about 500,000 patients that year.² Signify sends clinicians to people's homes for health evaluations, more than 3.5M of them in 2025, 85% of them for Medicare Advantage plans.²

It has not gone well. In the third quarter of 2025 CVS wrote down $5.7B of goodwill on what it calls the Health Care Delivery unit. The 10-K gives the reasons: persistent high usage of care, a change of management, a decision to open fewer new centers from 2026 and to close some existing ones, and lower values for similar companies.² It also recorded an $83M charge for the closures, sold its Medicare Shared Savings Program business and wound down another government program, for a $288M loss.² The goal CVS stated at the investor day is to drive the unit "towards breakeven", which tells you the unit was not at breakeven.⁶ In the second quarter of 2026 it reported a "modest improvement" and no figure.³ I can't judge how much the unit loses, because CVS does not report it separately.

IIThe moat

A durable advantage is a cost or a habit that a rival with money and talent can't copy in a reasonable time. CVS has three real ones and two that depend on the government. I try to separate them.

The pattern is that the advantages are largest where the profit is smallest per unit, and weakest where the profit is most exposed. Caremark's scale is real. It earns about $3.76 a claim. Aetna's data and membership are real, but the profit it makes on each member depends on the government rate and on the care costs of an ageing population. Neither of those is within CVS's control.

Evidence it is widening

• Prescriptions filled rose 5.4% in 2025 and 4.3% in the second quarter of 2026, while the store count fell 10% from its peak C (calculated).²³
• Health Care Benefits adjusted operating income rose from $0.3B in 2024 to $2.9B in 2025 and $5.5B in the first half of 2026 alone.²³
• CVS raised its 2026 adjusted EPS guide twice, from $7.00 to $7.20 in February to $7.90 to $8.10 in August.⁴¹³³
• Health Services adjusted operating income rose 10.0% in the second quarter of 2026, with claims flat.³

Evidence it is narrowing

• Adjusted EPS peaked at $8.74 in 2023 and was $5.42 in 2024. Adjusted EPS in 2024 was lower than in 2016, when it was $5.84.¹⁴¹⁵
• The share of Aetna's Medicare Advantage members in plans rated at least four stars fell from 88% to 81%, which reduces the bonus payment in the following year.²
• The 2027 Medicare Advantage rate is up 2.48% for the industry before the risk-score trend, against 5.06% for 2026.⁵²
• State laws in Arkansas, Louisiana, North Dakota and Oklahoma have tried to limit PBM practices, and the 10-Q notes that states have considered banning pharmacy licences for pharmacies affiliated with a PBM.²⁵
• Medicare drug plan membership fell from 4.9M to 3.9M in eighteen months.²⁵

On balance I think the position is wide in drug purchasing, medium in retail, and fragile in Medicare insurance. I give credit for the first, and I count the second as an asset that is costing more to hold. For the third I assume it can earn an ordinary margin in a good year and a poor one in a bad year, which is why my bear case is as low as it is.

IIIThe competitive advantage

The honest comparison for CVS is not one company. Its insurer competes with UnitedHealth, Elevance, Cigna and Humana. Its PBM competes with Express Scripts at Cigna and Optum Rx at UnitedHealth. Its stores compete with other chains and with mail order. The table sets CVS next to the four large insurers on one basis: 2025, as filed, with operating income before any adjustment.

Three things stand out. First, CVS has the lowest margin of the five on a reported basis, 1.2%, and still the lowest on an adjusted basis, 3.6% against reported margins of 2.1% to 4.2% at the others, though the bases differ C (calculated).²⁷ Second, CVS is the only one that owns drugstores, which have lower margins than insurance. Third, the price does not describe a company with a margin gap to close. On 2026 guidance the stock is on 11.1 times earnings. Elevance, Cigna and UnitedHealth are on 13 to 30 times 2025 reported earnings, though those are not on the same basis C (calculated).⁷

Where CVS wins

• It is the only large insurer that also owns a retail pharmacy chain, a PBM and primary-care clinics, so it can try to steer a member through all three.²
• Prescription share. Prescriptions filled rose 9.7% in two years against a store base that fell 4%.²
• Medicare drug plans. It is a leading standalone plan provider, even after membership fell.²⁵
• Cost of capital for the PBM. It needs little capital: capital spending was $2.8B in 2025, 0.7% of revenue, and 77% of it went to technology and strategic projects.²

Where it loses

• Margin and consistency. Adjusted EPS was $8.74 in 2023 and $5.42 a year later.¹⁴
• Value-based care. It paid for Oak Street and Signify in 2023 and wrote $5.7B off the care unit's goodwill in 2025.²
• Balance sheet. Debt is $61.4B, and only $2.7B of its $11.3B of cash sits outside regulated subsidiaries.⁵ Rivals with less debt can ride out a bad year more easily.
• Retail. Front-store sales fell from $22.5B to $21.5B in two years, and the 10-K flags weaker spending and a shift to grocery and digital retailers.²

IVManagement

David Joyner became chief executive in 2024; the proxy reports two chief executives for that year. He has been with the company for nearly 40 years, and said so at the 2025 investor day.⁶ On 1 January 2026 he also became chair of the board, after Roger Farah, who had been executive chair from October 2024, retired from the board.⁸ The senior team around him is mostly new. Brian Newman joined as chief financial officer in April 2025 and replaced Thomas Cowhey on 12 May 2025.⁸ Steven Nelson became president of Aetna in November 2024; he ran UnitedHealthcare from 2017 to 2019.⁸ Prem Shah, who had been chief pharmacy officer, became group president in November 2024.⁸

The turnover is a fact I weigh both ways. A new team can reset a business that had lost its way, and in 2025 it did several things a settled team might not have done: it wrote down the care unit, cut new clinics, left the individual exchange, and put Omnicare, its long-term care pharmacy unit, into Chapter 11, which removed it from the accounts.² The other side is that the people who now promise a mid-teens earnings growth rate through 2028 have run the business for under two years, and the people who set the 2024 numbers are gone.

Pay, and what it pays them to do

Joyner's reported pay for 2025 was $21.2M: a salary of $1.5M, stock awards of $14.5M and a cash bonus of $4.7M, with $0.5M of other items. For 2024 it was $17.8M, of which $12.0M was stock options.⁸ The cash bonus has three parts: 60% adjusted operating income, 20% an adjusted ratio of selling and administrative cost to revenue, and 20% customer satisfaction.⁸ The long-term stock awards pay out on three years of cumulative adjusted EPS for 70% and a strategic scorecard for 30%.⁸

So follow the money. Adjusted operating income and adjusted EPS leave out amortization of acquired intangibles, goodwill write-downs, restructuring and the litigation charges CVS chooses to separate.² The $5.7B written off the care unit in 2025 did not touch the bonus measure. Neither did $1.2B of litigation charges for two court decisions. The pay design rewards growth in adjusted profit. It does not reward cash left after capital, return on the money spent on acquisitions, or the size of the balance sheet. Some protections exist: no upward modifier on the stock awards when the share price return is negative, a holding period on shares from vested awards, and policies against hedging and pledging.⁸ I'd rather see a measure of return on capital in the plan, because the biggest mistakes at CVS in the last ten years were purchases.

Their own money

Joyner owned 57,977 shares outright at 16 March 2026, about $5.1M at $88.51, plus rights to 556,877 more through options and share units that were exercisable or vesting within 60 days C (calculated).⁸¹ That is a quarter of one year's pay in shares he has bought or been given and kept. It is a fair stake, not a founder's. Each director and officer holds less than 1% of the company, and the largest holders are index managers: Vanguard 9.14%, BlackRock 8.30% and Capital World Investors 5.50%.⁸

One holding is worth following. Larry Robbins of Glenview Capital Management sat on the CVS board until 13 August 2026.¹⁷ Glenview's 13F filings show 8,196,799 CVS shares at 31 December 2025 and 31 March 2026, and 4,824,799 at 30 June 2026, a cut of 41%, valued at $499M, 9.1% of Glenview's reported US stock portfolio.¹⁸¹⁹ That is a holding, not a stated view. I did not find anything Robbins has written or said about CVS in the documents I opened, and I make no claim about why he left. I note it because the 13F shows the position was already smaller at 30 June, six weeks before the resignation, and because both filings are public.

What they did with the money, over a full cycle

I take 2016 to 2025 as the cycle. It covers the Aetna deal, a pandemic, two big care acquisitions and an insurance cost shock. Over those ten years CVS generated $123.3B of operating cash. It spent $25.0B on capital, $62.7B on acquisitions, $26.6B on dividends and $17.4B on buybacks.¹² Reported net income over the same ten years was $52.2B C (calculated).¹²

Three things stand out. First, the dividends and buybacks together, $43.9B, equal 84% of the ten-year net income. It paid out most of what it earned on paper, and bought companies with borrowing and new shares. Second, the acquisitions were almost all financed by debt and shares. Debt rose by $39B and the share count by 18%, and adjusted EPS went from $5.84 in 2016 to $6.75 in 2025, a gain of 16% in nine years C (calculated).¹²¹⁵⁴ The price is $88.51 against $83.95 on 2 November 2016, up 5%, with dividends on top.¹ Third, buybacks were opportunistic. In 2023 and 2024 CVS bought 62.5M shares for $5.0B, an average of about $80; the stock is above that now.²

The record is fair, not good. Aetna widened the reach and brought a lot of profit for a few years. The 2023 care purchases have already been marked down by a third of the amount paid, if I take the $5.7B against the $16.6B of cash spent C (calculated).²¹² I put management's capital allocation in the bottom half of what I see on this site, and the current team's job is to show the next ten years are different. A dividend that is 38% of my owner-cash estimate is reasonable, and no buyback since 2024 is the right call for a company with a BBB rating.

VFinancials

CVS's accounts are large, and four features make the reported earnings noisy. First, 2018 and 2025 each include a goodwill write-down of more than $5.7B. Second, the company reports an adjusted profit that leaves out amortization of acquired intangibles, $2.0B in 2025, and a changing list of charges. Third, a large share of the revenue is drug cost passed through at a thin margin. Fourth, an insurer's profit depends on estimates of claims it has not yet paid, which can move up or down by billions. I try to see through each of them.

Margins

The operating margin on reported figures ran between 1.2% and 5.8% and ended at 1.2% in 2025 C (calculated).¹² Adjusted, it was 3.6% in 2025 C (calculated).² The low margin is a feature of the model: the PBM and the pharmacies pass drug costs through, so a 4% margin on $400B is $16B.

Earnings quality

Adjusted EPS and GAAP EPS tell different stories. Over the ten years the adjusted figures sum to $71.40 a share and the reported figures to $42.11, a gap of $29.29 C (calculated).¹²¹⁵⁴ Part of that gap is amortization of intangibles, which is not a cash cost. The rest is real: $12.5B of goodwill write-downs, restructuring in 2024, an opioid accrual that built over several years and $1.2B of litigation charges in 2025.² The opioid obligation alone was $3.4B at 30 June 2026 and is paid over years.⁵

A second point is the season. Medicare drug plans are paid evenly but their costs come later in the year, so Aetna earns most of its profit early. In 2025 adjusted EPS was $4.06 in the first half and $2.69 in the second. In 2026 the first half was $5.16, and the midpoint of guidance implies about $2.84 for the second half, only $1.42 a quarter.³⁴ If you annualize a good first half you will be wrong by a lot.

Reserve releases

Insurers book medical costs for claims they have not yet paid, and later adjust the estimate. In the first half of 2026 CVS's estimate for prior years came down by $1.2B, because claims cost less than expected, against $1.9B in the first half of 2025.⁵ So the release was smaller this year than last. The 10-Q says it was lower and that the improvement in margin came from the underlying business and the absence of $902M of deficiency reserves taken in 2025.⁵ Days claims payable fell 1.2 days in the quarter to 41.7.³ I cannot tell from filings whether the reserves are cautious or thin. A reserve that is short would bring a charge in a later quarter, which is what happened in 2024.

From profit to cash

Operating cash flow was $10.6B in 2025 against reported net income of $1.8B.² The gap is depreciation and amortization of $4.6B, the goodwill write-down of $5.7B, stock pay of $0.5B and a working-capital swing. In the first half of 2026 operating cash flow was $10.6B against net income of $6.0B, helped by a $2.2B inflow from working capital that may reverse.³ CVS guides to at least $11.5B for the year.³ Capital spending is modest at $2.8B in 2025, so reported free cash flow was $7.8B, and $7.3B after stock pay C (calculated).²

What an owner could take out

Here is the number I care most about. If CVS stopped growing and simply kept everything in good order, how much could it pay its owners each year? I start from the middle of 2026 adjusted income guidance and take off the things the adjusted figure leaves out but a shareholder still pays for. Stock pay is already in adjusted income, so I do not add it back.

The answer is about $8.9B, or $6.93 a share E (estimate). That is 13% below the adjusted $8.00. I can't pin it down more tightly, because CVS does not tell us how much of its profit has to stay in the insurance subsidiaries, and I don't know what other court payments will fall due. If I took the whole of the stock pay out of operating cash flow instead, I would get a figure close to the 2025 free cash flow after stock pay, $5.72 a share on 2025 diluted shares C (calculated). The honest range for steady owner cash is $5.7 to $7.0 a share.

The balance sheet

At 30 June 2026 CVS held $11.3B of cash, $2.7B of it by the parent or unrestricted subsidiaries, and $2.6B of short-term and $33.2B of long-term investments, mostly in the insurers.⁵ Debt was $61.4B, of which $2.0B was current.⁵ Goodwill was $85.5B, more than the total shareholders' equity of $79.9B, so the book value has a lot of acquired value in it.⁵ Operating leases add $14.9B of liabilities, but their rent is already in operating costs.⁵ In August 2025 CVS sold ten-year notes at 5.45%.² The ratings were BBB at Fitch, Baa3 at Moody's and BBB at S&P at the date of the 10-K, with negative outlooks at Fitch and S&P.²

In my valuation I do not subtract net debt from an enterprise value, because the owner cash I value is already after interest. The balance sheet matters in a different way. A BBB company with $61.4B of debt and $2.7B of free cash at the top cannot take a second year like 2024 and also keep paying a $3.4B dividend and invest. That is the real cost of the bad year, and it is why the bear case in my valuation is low.

VIValuation

Start with what would go wrong

Before any numbers, I ask what would most likely make this a bad investment from here, and work back from that. For CVS the problem is not a lack of customers. It is a thin margin on a very large base, set by a regulator and by medical costs that move without warning. Here are the three ways it goes badly wrong, in the order I think most likely.

1. Aetna's medical costs run ahead of its premiums again. Each point of medical benefit ratio is about $1.35B of operating income, or $0.80 a share E (estimate). Three points, a smaller move than the 6.3 points between 2023 and 2024, would take about $2.39 off a share, nearly 30% of the $8.00 guide.²³ The 2027 Medicare Advantage rate rises only 2.48% before the estimate of rising risk scores, and CMS is auditing every Medicare Advantage contract.⁵² Medical costs have not stopped rising.
2. The PBM is forced to change. Caremark earned $7.2B in 2025, 44% of the three segments' total before corporate costs C (calculated).² State laws that bar a PBM from owning a pharmacy, rebate lawsuits and clients demanding a larger share would hit that profit without a matching saving elsewhere.⁵² A 25% cut to it takes $1.8B off operating income, about $1.06 a share E (estimate).
3. A second bad year meets a BBB balance sheet. With $61.4B of debt, negative outlooks at two rating agencies and $2.7B of cash outside regulated subsidiaries, a repeat of 2024 would force a choice between the $3.4B dividend, the debt and the care businesses.⁵² The opioid obligation is $3.4B and falls due over years whether or not profits are good.⁵

I can't put honest odds on the second, because it depends on legislation and court rulings I can't predict, so I don't try. I can show what the first does to the value, and I do below.

What the price implies

CVS guides to adjusted EPS of $7.90 to $8.10 for 2026, up 17% to 20% on $6.75.³⁴ The guide has been raised from $7.00 to $7.20 in February, to $7.30 to $7.50 in May, and to the current range in August.⁴¹³³ My base case lands near the guided figure for 2026 and does not assume the third quarter beats it. So I am not more cautious than the company about this year. I am more cautious about 2028.

Over ten years the price went from $83.95 to $88.51, up 5%, while adjusted EPS went from $5.84 to $8.00 on guidance, up 37% C (calculated).¹¹⁵³ The market has paid between 8.3 and 13.5 times each year's adjusted earnings at the year end, never more than 13.5 times, against 11.1 times now C (calculated).^{11215202122144} So the price is not unusually low compared with the company's own history. It is in the middle of it. What is unusual is the earnings it sits on, which are 19% above the 2025 level and 8% below the 2023 peak of $8.74.

Reverse DCF: what the price needs

Held at my owner-cash conversion, the price needs adjusted earnings per share to grow only about 0.2% a year for ten years from the $8.00 guide C (calculated). If earnings stayed flat at $8.00 for ten years and then grew at 2% for ever, the model gives $87 a share, almost exactly the price. If they fell back to the $6.75 of 2025 and stayed there, it gives $74; at $5.50, near the 2024 level, $60 E (estimate). So the price does not describe a business that is expected to grow. It describes one expected to hold this year's profit for ten years, or to fall back and recover. That is a different claim from "cheap", and it is the claim the market is making.

There are two other ways to read the same price. One is that it equals my base cash flows at a discount rate of about 10.7%. That is the return the price offers if my base is right: 5.8 points above the Treasury and 5.3 points above what CVS paid to borrow for ten years.²³² The other is the three cases below, in which the same price returns 6.9%, 10.7% and 13.2% a year E (estimate).

The discount rate

I discount at 9.0%. It is the 10-year Treasury at 4.94% plus about four points.²³ I use more than the 8.0% I used for Walmart, because CVS has a lower credit rating, a business that depends on decisions by Washington, and a record in the last three years that includes a 38% fall in adjusted EPS in one year. The company's own ten-year borrowing cost of 5.45% is below the return on its stock by definition. Four points above a risk-free rate is, to me, the least I ask for a business whose profit can drop by a third.

Bear, base and bull

My base case is CVS recovering to the level of its best recent years and no further. Adjusted EPS goes from $8.00 to $8.29 in 2027 and $8.60 in 2028, which is about where it was in 2022 and 2023, then grows 4% a year, fading to 2%. That requires adjusted operating income of about $17.7B in 2028, against $14.4B in 2025 and about $16.7B this year E (estimate). It is a case in which management hits roughly half of its target growth and no more, which is my allowance for the company's record of missing.⁶

The bull case is management's own. In December 2025 CVS said it expected a mid-teens compound growth rate in adjusted EPS through 2028, from the $6.75 of 2025.⁶ Fifteen a year takes $6.75 to $10.27 in 2028. I use $10.30 and let growth run at 7% a year after, fading to 2%. That is 1.34 times my base earnings in 2036 and is worth $144. The bear case has 2028 adjusted EPS of $5.50, close to 2024's $5.42, and almost no growth after. It is worth $63.

What I can't judge

Some inputs I can't know, and I would rather say so than pretend. I don't know how much capital the insurers must keep as premiums grow; I put in $0.4B and it could be twice that. I don't know whether the reserves for unpaid claims are cautious or thin, and only later quarters will show. I don't know what the care businesses lose, because CVS does not report them separately, and I don't know how the PBM laws will be decided in the courts. I can't forecast Medicare Advantage rates beyond 2027. I take the whole range to be wider than my three values suggest, and I would not argue with someone who put bear at $50 or bull at $170.

Multiply it out

My base needs several things at once. Aetna holds a ratio near 89% for a full year without a big reserve release: I would give that 80%. Caremark profit holds near $7B through 2028: 80%. The stores' profit holds near $6B as the count falls: 85%. The care businesses reach about breakeven: 75%. No new charge or ruling on litigation or Medicare audits that takes more than $0.3B a year: 80%. Multiplied, that is about 33%, one chance in three that all five hold. Management's target needs all of them to go better than my base, not just hold: 40%, 55%, 70%, 50% and 70% give 5%. Those odds are guesses and I do not use them in the values. They are why my base is only the middle and not the likely outcome.

Quality and price are two questions

How good is the business? Mixed. The PBM and the drugstores are large, steady and low-margin, and together earned $13.2B in 2025. The insurer has earned between $0.3B and $5.6B in the last three years. The care businesses have destroyed value so far. How good is the price? Fair to good on my numbers: it sits 20% below my base value, 40% above the bear, and returns 10.7% on my base against the 9.0% I require. The risk is that the quality is lower than my base assumes. A cheap-looking price on a business with an unstable earnings line is not cheap if the line does not recover. Both of those are true at once.

It helps to compare with the next best use of the same dollar. A ten-year Treasury pays 4.94% with no risk to the dollars promised.²³ CVS returns about 10.7% a year on my base, 5.8 points more, and 6.9% on my bear, which is 2.0 points more than the Treasury even when the shares are worth less than $88.51 at a 9% rate. On this site, Walmart returns 5.5% on its base case at its rate of 8.0% and Costco 6.2% at 8.5%; Berkshire Hathaway and JPMorgan Chase both sit 15% to 16% above their base values, while CVS sits 20% below. CVS offers the higher return of these on my numbers, and also the widest doubt about whether the numbers are right.

The margin of safety

My margin-of-safety price is $77, 30% below my base value. At $77 my bear case returns 7.7% a year, below the 9.0% I require but above the Treasury, and the base returns 12.0% C (calculated). A 30% cushion is what I use for a business whose earnings can fall by a third. The price is 15% above it, and the stock last closed below $77 on 22 April 2026, at $76.43, with a low of $70.08 on 27 March 2026.¹

VIIThe case for the price

The price is below my base value, so the question here is not whether the bull case can reach it. The price already assumes less than I do. The question is what is true in the bull case, what is true in the market's, and what each side has to show.

What most people miss

First, the bad news was one segment. Caremark and the drugstores earned $13.3B in 2023 and $13.2B in 2025. Aetna fell from $5.6B to $0.3B in 2024 and has since climbed to $2.9B in 2025 and $5.5B in the first half of 2026 alone C (calculated).²³ Second, the recovery is not coming from bigger reserve releases. The estimate for prior years came down by $1.2B in the first half of 2026 against $1.9B a year earlier, and the 10-Q lists lower favourable development among the things that held profit back.⁵ Third, the stores shrank and the prescriptions did not: 243 stores closed in 2025, yet prescriptions filled rose 5.4%.² Against that, only $2.7B of $11.3B of cash sits outside regulated subsidiaries, which limits what the headline cash can do for an owner.⁵

The bull case

The bull case, as I understand it: Aetna's margin was broken by a spike in medical costs, CVS has repriced and exited the worst business, and it is returning to target margins. Caremark is the steady earner. The drugstores are consolidating, with Rite Aid's files, into fewer and busier stores. The care businesses stop losing money. The group then earns mid-teens growth on an 11-times multiple.

Management's version has numbers. At the 9 December 2025 investor day CVS committed to a mid-teens compound growth rate in adjusted EPS through 2028, resting on four things: a return to target margins at Aetna, winning and keeping business at Caremark at attractive margins, sustained earnings at the pharmacy, and a path to breakeven in the care businesses.⁶ Chief financial officer Brian Newman said the company's guidance philosophy is to give credible expectations with room to beat them.⁶ Since then it has done that: the 2026 guide has risen from $7.00 to $7.20 to $7.90 to $8.10.⁴³

I found no named investor with a public, current case for CVS that I can attribute. The one holder I could check was Larry Robbins of Glenview Capital Management, a CVS director until 13 August 2026. Glenview held 8,196,799 shares at the end of March and 4,824,799 at the end of June.¹⁷¹⁸¹⁹ That is a holding, not a thesis, and it was reduced.

Against reality

Much of the bull case is true now. The first half is the best in years, the medical benefit ratio is much lower, and the guide has risen twice.³ So I tested what happens if CVS does what it says. Adjusted EPS of $10.30 in 2028 needs adjusted operating income of about $20.7B, against $14.4B in 2025 and about $16.7B this year E (estimate).² The gap of about $3.9B from this year is nearly three points of Aetna's medical benefit ratio, or equivalent gains in Caremark and the care businesses E (estimate). If CVS does that and keeps growing 7%, my bull case is worth $144, 63% above the price. So yes, if they hit their targets the price is too low.

How far is today from that? One good half-year is not three good years. The same company reached $8.74 in 2023 and $5.42 in 2024. The first half of 2026 is seasonally strong, and the guide implies a second half of about $2.84 against $5.16 in the first.³ The 2027 Medicare Advantage rate is lower than 2026's and the care businesses are not yet at breakeven.⁵ That is why my base is $8.60 in 2028 and not $10.30. The case for the market is the same facts read the other way. If 2026 is the best year for a long time, then flat earnings, which is what the price assumes, is a fair estimate.

What a 10x would take. Ten times $88.51 is $885 a share, about $1.13T of market value on today's share count C (calculated). At 20 times earnings that needs about $57B of net income, 5.5 times the guided 2026 adjusted income; at 13 times, $87B. UnitedHealth, the largest US insurer, earned $12.1B in 2025 on $447.6B of revenue.⁷ At CVS's adjusted net margin of about 2.5% on $414B of revenue, $57B of income needs about $2.3T of revenue, 5.7 times the 2025 figure C (calculated). A tenfold move in ten years would be a rate of 26% a year. CVS itself is up 5% in ten years. For companies of this size a tenfold rise in a decade has almost never happened, and I see no route for it here. A doubling is possible: my bull case is 1.6 times the price.

Where I come down

The split of profit moved me: the loss was in one segment, not three, and the recovery in that segment is not coming from bigger reserve releases. The price did not move me much, because it is at the middle of its ten-year range of multiples, and management's target did not either, because the same team style missed in 2024. What would raise my values most is two more quarters of an annual medical benefit ratio near 89% with Caremark's profit holding after the legal and rate decisions of early 2027.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest.

• Medicare audits and rates. CMS said in May 2025 it would audit every Medicare Advantage contract each payment year, with an expedited plan for 2018 to 2024. A court vacated the 2023 audit rule in September 2025 and CMS appealed. The Department of Justice has sent civil demands about how CVS submits diagnosis codes.²
• Drug-plan economics. Standalone Medicare drug plan membership fell from 4.9M to 3.9M, and the Inflation Reduction Act changed Part D economics enough to lift Aetna's 2025 revenue.²⁵
• PBM law. Arkansas, Louisiana, North Dakota and Oklahoma have passed laws aimed at PBM practices, with lawsuits about whether federal law overrides them. Rebate suits by states and others are now combined in a New Jersey federal court. The 10-K named the Federal Trade Commission among the plaintiffs in these suits; the 10-Q does not.²⁵
• Legal cash. The opioid accrual is $3.4B. Omnicare and the company were found liable in a False Claims Act case in 2025, and Omnicare then filed for Chapter 11.⁵²
• Retail. The 10-K warns of weaker consumer spending, theft at stores and a shift to grocery and digital sellers. Front-store sales fell from $22.5B to $21.5B in two years.²
• Care businesses. Oak Street and Signify depend on Medicare Advantage payments. A cut to risk adjustment or a change to home visits would hit both and the insurer.²
• Rating and funding. Negative outlooks at Fitch and S&P. A cut raises borrowing cost on $61.4B of debt.²⁵

What would raise my base value: the next two reported years showing a medical benefit ratio near 89% with Caremark's profit intact. What would lower it: another year like 2024 at Aetna, a court ruling that stops PBMs from owning pharmacies, or a downgrade that forces a dividend cut.²⁵