Seventy-four cents on the dollar, and no equity left

On the close of Friday 2 October 2026, HCA Healthcare traded at $426.69. Sperio values a business from its filed accounts and a discount rate, giving no credit of any kind to future growth, and arrives at $579.11.

The market is paying 74 cents for each dollar the accounts support without growth. Last week's company was priced at 95 cents; this one is a discount in the plain sense of the word. It is the eleventh cheapest of the 100 companies that clear our two gates, the third on that list with no blocking flag, and not the name our machine put first. It is also a company with no book equity left — it has spent it, and more, buying back its own shares — which is the first thing a careful reader should hold against it, and the reason this piece spends as long on the downside as on the discount.

Why not the widest discount, and why not the machine's pick

The widest discount in the index this week belongs, for the third Sunday running, to a fertiliser company whose net income is down 85%. Behind it sit a laboratory landlord and an office landlord whose net income is down by two thirds. Ranking by discount answers "what looks cheap", and it answers it reliably with whichever business is deteriorating fastest, because a collapsing denominator makes any price look like a bargain.

So the ranking applies two hard gates first. Does the business durably earn more on its capital than the capital costs — return on invested capital clearing ten percent in four fifths of the readable fiscal years and in each of the last three. And are both lines going the right way — revenue and operating income each compounding upward over the full readable history. Of the 454 companies the board has valued, 100 clear both.

The survivors are then ordered by cheapness: enterprise value to the median operating profit of up to fifteen fiscal years. The list this morning, cheapest first:

A. O. Smith was the subject three weeks ago, and the machine has put it first three Sundays running; nothing in its accounts has changed, and the scoring section reports its price. Gilead clears everything this list asks, but its shares are only 7% below their high: there is no gap between the price and the accounts to explain, and the explaining is what this series is for. HCA is the next name with nothing blocking, its shares are 22% below their March high, and on the honest version of our own figures — corrected below — it is cheaper against its no-growth value than either. That is an editorial choice, made for stated reasons, and the machine's own pick is printed next to it so that a reader can disagree.

One name we wanted and could not take: Lowe's, the company in this neighbourhood that the public is searching for this month. It clears both of the machine's gates, but not the two lines we check by hand: revenue in its last fiscal year was 11% below the year three years earlier — partly the sale of its Canadian stores — and earnings per share have fallen two years running. We do not loosen the screen for a famous name.

Which multiple

Twenty point seven times is the ranking's measure: enterprise value divided by the median operating profit of the readable fiscal years — ten of them for HCA, on operating profit as our data source computes it, because HCA files no operating-income line of its own. Every figure in the tables below is a filed line instead.

The plainer measures, on those filed figures: at Friday's price the company costs 14.3 times its last twelve months of earnings per share, and 15.1 times fiscal 2025's. The shares are worth about $92.4bn; add $49.7bn of debt, take away $1.0bn of cash, and a buyer of the whole business pays about $141bn — of which more than a third is debt. That last fact matters more to this piece than any multiple.

What it earns

Return on invested capital, on the ranking's measure: a median of 17% across the readable record, and above ten percent in each of the last three years. Return on equity cannot be computed — there is no equity. In fiscal 2025 the company reported $6.78bn of net income and $7.69bn of free cash flow, about 8.3% of what its shares are worth at Friday's price. Over the twelve months to June 2026 free cash flow was lower, $6.00bn: receivables rose, mostly money owed under the Medicaid programs discussed below, and taxes deferred from 2025 came due.

The accounts, five fiscal years and the trailing twelve months, as filed with the SEC (HCA files no gross-profit line; the second row is revenue less its three large cost lines — salaries, supplies and other operating costs):

Over the ranking's window, fiscal 2022 to 2025, revenue is up 26%, and what is left after the three cost lines up 29% — the margin held at about a fifth. Income before taxes is up 15%, or 35% once the 2022 base is cleaned of $1.3bn of gains from selling hospitals. Net income is up 20%. No impairment in any year.

What a fifth fewer shares does, and does not, prove

Earnings per share went from $19.15 to $28.33 over those three years, up 48%. That number on its own is close to useless, and this is the split that matters: net income up 20%, diluted share count down 19%. Slightly more than half of the per-share growth is the share count. Both halves are real — net income genuinely rose, which separates this from the pattern where a shrinking share count hides a shrinking company — but quoting the 48% without the split would be the flattering choice and the less honest one.

Then the cost. From 2022 to 2025 HCA spent $26.9bn buying back its own shares, against $23.4bn of net income, before dividends. The difference was borrowed: total debt went from $38.1bn at the end of 2022 to $49.7bn at 30 June 2026. HCA retires the shares it buys, so every buyback comes straight out of retained earnings, and stockholders' equity went from −$2.8bn to −$6.6bn over the same stretch. Over the last twelve months it bought back $8.7bn of stock with $6.0bn of free cash flow.

Negative equity here is the accounting consequence of a policy, not a sign of distress. But the policy has a consequence of its own: there is no cushion, and the buybacks now run ahead of the cash the business makes.

What the price is saying

On 9 March 2026 the market paid $545 for a share of this company; on 8 June, $361. Two days carried most of the fall. On 24 April, the day of the first-quarter results, the shares lost 8.8%: patient volumes had come in below plan, adjusted EBITDA a little under what analysts expected, and what covered the gap was Medicaid supplemental money the company had not put in its guidance. On 14 July they lost 6.9%, when the company cut its 2026 guidance ahead of its second quarter. Net income is now expected at $6.30bn to $6.70bn — below the $6.78bn of 2025; earnings per share at $28.70 to $30.50, above 2025's $28.33 only because there will be fewer shares. It put the cost of more uninsured patients, as people lost the coverage they had bought on the insurance exchanges, at about $400m before tax in the second quarter alone, and at $1.0bn to $1.2bn for the year — up from $0.6bn to $0.9bn when it last estimated it, in April. The other listed hospital operators fell with it that day, by between 5.6% and 8.3%: the market read it as a problem for the industry, not for one company. On Friday, $426.69: 22% below the March high, and exactly where it was a year ago.

The filed volumes are less dramatic than the price. In the second quarter, same-facility admissions still grew 2.5% and emergency-room visits 3.6%; the uninsured went from about 7% to about 8% of admissions. Surgery is the soft spot — inpatient operations down 2.3%, outpatient down 3.4%.

The market's case is about who pays the bill. A hospital is paid by insurers, by Medicare, by Medicaid and by patients; when people lose private cover the same patient still comes through the door, and pays less, or nothing. And the accounts give the argument real support. Medicaid supplemental payments — money states route to hospitals on top of ordinary Medicaid rates — rose from about $4.4bn in 2023 to $6.2bn in 2025, 8.2% of revenue, and account for $1.8bn of the $10.6bn of revenue growth over those two years. The second quarter of 2026 included about $400m of such money covering October 2024 to June 2026 — twenty-one months, booked in one quarter — roughly the size of the exchange losses it offset. With it, the quarter's net income rose 2.8% on revenue up 8.7%. And the company's own quarterly filing says the grandfathered programs phase down from 2028. Part of the growth in the table above is government money with a published end date.

The downside, priced — and a cushion withdrawn

A single valuation cannot be argued with. Three can.

The price sits well under the middle rung: a buyer pays $426.69 for earnings the accounts support at $579.11 if they never grow again. The trough case rescales that no-growth value to the worst readable year of operating profit, about three fifths of the median, and it sits 18% below the price. A buyer wrong about everything except the trough is down about a fifth, not protected.

Here is the withdrawal. Our machine's Sunday table this morning shows HCA's trough at $451.95 — above the price — and HCA is one of four names in the whole index that the table credits with that cushion. It is not real. It is the error we found last week on Accenture: for this company the board's headline valuation, $747.05, already carries the franchise credit (the no-growth models come out between $579.11 and $612.68, and the headline is the lower one times a factor of 1.29), and the table treats that headline as the no-growth value, scales the trough from it and applies the credit a second time. The figures above are the corrected ones. The cushion is withdrawn, and so is a draft of this piece we wrote in September on the same uncorrected numbers, which was never published. The fix to the table is ours to make and is not made yet; until it is, we check that line by hand for every name we publish.

The rate sweep

At the cost of capital the ranking used this morning — 10.24%, built on the 5.24% ten-year Treasury as FRED reported it on 2 October — 100 companies clear both gates. Raise it by a hundred basis points and 85 still clear; by two hundred, 77. The table also counts the names whose trough sits above their price: four today, of which HCA is one through the double credit and Paychex another; none at +100 basis points. On the corrected figures HCA has no such cushion at any rate.

The case against, in the board's own words

Sperio runs a devil's advocate whose only job is to attack the thesis. From the board's report dated 4 October 2026, reproduced without softening:

HCA operates with negative equity and a substantial $48.7B debt load, creating significant balance sheet risk if interest rates remain elevated or refinancing costs spike.

And from the same report's bear case:

HCA's reliance on debt-funded share repurchases has resulted in negative equity and a massive $48.7B debt load, leaving little room for error if refinancing costs spike. Additionally, recent downward revisions to long-term guidance suggest structural margin erosion from unfavorable payer mix shifts.

The debt half of that is arithmetic, and it is right: a business with $49.7bn of debt and no equity meets a bad year with interest that does not move. It is worth sizing, though, rather than calling it massive: about 3.1 times the last twelve months of adjusted EBITDA, at an average interest rate that fell to 4.9% this year from 5.1%. That is a leveraged hospital company, not a distressed one — and it is still borrowing, to buy back shares. The payer half is the market's case above, and the accounts cannot settle it — the exchange losses only began this year, and the supplemental money ends on a schedule set by legislators, not by HCA.

What the accounts can say is narrower. At Friday's price a buyer pays $426.69 for earnings the frame values at $579.11 held flat forever — 74 cents on the dollar — and gets the record for nothing. The discount is the market's estimate of the two risks above; the frame cannot tell you whether it is the right size. If both arrive in full, the trough line is the honest floor, and it sits below the price.

Why now

Nothing in the accounts is dated this week — the board's verdict has not moved, and saying so is more honest than manufacturing a reason. What is dated is the next report: third-quarter results on Tuesday 27 October 2026. The number to read is not the beat or the miss. It is how much of the quarter came from supplemental programs, whether the exchange losses grew or held at about $400m a quarter, and whether the company keeps the guidance it cut in July. The consensus already expects a down quarter — $6.66 of adjusted earnings per share, against $6.96 a year earlier, a quarter that had its own supplemental payments in it. And the July guidance itself implies a second half with net income 4% to 15% below the second half of 2025. This is published three weeks before that date, and says so.

Last week's pick, scored

Last Sunday's piece said that this one would re-read Accenture against its new accounts, and score it out loud.

It reported on Thursday 1 October. Fourth-quarter revenue was $18.68bn, up 7% in local currency and above the top of its own guidance; the fiscal year closed at a record $74.18bn, with operating income as filed up 12%. The shares rose 15.8% that day and closed Friday at $198.90, against the $176.11 we published at — up 12.9% in a week. A week is not a result, and that number is reported because it was promised, not because it proves anything. What the release adds to the argument cuts both ways: next year is guided at 3% to 6% revenue growth in local currency, and management said on the call that 2 to 2.5 points of that will come from acquisitions. Still growing, then — but next year, a good part of the growth is bought. Accenture still clears both gates this morning, twelfth on the list.

A. O. Smith, the machine's pick again, closed at $56.75 against $58.65 the week before, down 3.2%. Cognizant, the pick before last, closed at $58.51 against $57.33, up 2.1%, and is still first on the list and still blocked by the two flags we argued with. None of these numbers means anything at a week; all are recorded so that in a year they can.

What this is, and is not

Every figure about HCA comes from accounts filed with the SEC — the 10-K for 2025, the 10-Q for the quarter to 30 June 2026, filed on 28 July 2026, and the company's releases of 27 January, 24 April, 14 July and 24 July 2026 — from a board valuation dated 4 October 2026, and from a ranking run dated 4 October 2026. The price used is the close of Friday 2 October 2026. The cost of capital is built from FRED observations. Analysts' consensus figures and the other hospital operators' moves on 14 July come from press coverage, not filings. HCA files no operating-income line: where this piece says operating profit it means the ranking's own measure, and every figure in the tables is a filed line. One note on the company page on this site: it shows earnings per share computed on a different share count from the diluted figure HCA files ($30.20 against $28.33 for 2025); the table above uses the filing. Accenture's figures come from its fourth-quarter release, filed on 1 October 2026.

The valuation gives no credit to growth beyond what the franchise record earns, and the ranking that produced this list orders the survivors of two gates by their full-cycle operating multiple — a way of comparing them to each other, tested on a small sample, not a forecast of what any of them will do. This week the editor went past the machine's pick, for reasons stated above, and withdrew a cushion the machine printed; both are reported so that a reader can disagree.

The board that produced these figures will run on any company you name. What it will not do is decide for you — nothing here is investment advice.