The widest discount in a value screen is almost always a warning rather than an opportunity. That is not a slogan, it is what the list looked like on 23 August 2026.
Sperio values a business from its filed accounts and a discount rate, with no credit of any kind for future growth. Rank the companies the board has valued by how far the market price sits below that number, and the thirteen widest gaps run from +338% to +50%.
Then apply two lines of arithmetic to that list. Is revenue larger than it was three years ago? Is earnings per share larger than a year ago?
Six of the thirteen fail the first line — two of them with revenue shrinking more than 14% a year. Three more are flat, under 1.5% a year, which for this purpose is the same answer. Of the four still standing, two fail the second: earnings per share down 89% in one case, down 2% in the other. A no-growth valuation applied to a business in decline produces a large and meaningless margin of safety, because the model assumes flat earnings and the earnings are not flat — they are falling.
Eleven of the thirteen widest discounts in the index are out on those two lines. The two that survive are Cigna and Universal Health Services. This piece is about the cheaper of the two, on a trailing price-to-earnings ratio of 7.6 against 12.4 — and the one the board argued about least.
The arithmetic
On the most recent annual accounts on file, the no-growth valuation came to $267.05 a share, with the upper end of the range at $393. The market closed at $177.26 on Friday 21 August 2026. That is a 50.7% discount to a number that assumes the company never grows again.
On the screener's own basis — enterprise value against the median operating profit of up to fifteen fiscal years — it stood at 13.5x, roughly the twenty-first cheapest of the 309 names with enough filed history to rank.
The trailing price-to-earnings ratio was 7.6.
Why "no growth" is the conservative assumption here
This is the part that separates this name from the eleven the two lines removed.
Universal Health Services grew revenue from $13.40 billion in FY2022 to $17.36 billion in FY2025 — 9.0% a year, compounded, for three years. Earnings per share went from $9.23 to $23.42 over the same period, and rose 36.5% in the most recent year alone.
Some of that is operating leverage and some is arithmetic: the share count fell from 73.8 million to 64.5 million, a 12.7% reduction in three years. Net income rose 30% in the last year; earnings per share rose 36.5%. The difference is the buyback.
Return on equity was 20.5%, and debt stood at 0.71 times equity — modest for a capital-intensive operator.
So the valuation is not being generous. It is refusing to count a 9% growth rate, a 36% earnings increase and a shrinking share count, and the business still comes out worth half again what the market paid. That is the strongest form this method's argument can take: the conclusion does not depend on the thing the method is worst at estimating.
The case against
Three objections, and none of them are decorative.
Capital expenditure eats the cash. Operating cash flow was $1.86 billion in FY2025. Capital expenditure was $1.04 billion — 56% of it. What is left is a free cash flow margin of 4.7%, thin for a business earning a 20% return on equity. Hospitals are not asset-light, and a valuation built on earnings power will always read richer than one built on cash that actually reaches the owner.
And that cash flow is lumpy. Free cash flow across the last four filed years ran $262m, $525m, $1,123m, $825m. A single year is not a trend, in either direction, and anyone anchoring on FY2024 is anchoring on the best of the four.
The acquisition is into a worse business. The Talkspace purchase takes a wide-moat hospital operator into telehealth, a market with low barriers, heavy competition and none of the pricing power that produces UHS's gross margins. Integration risk here is not a formality.
To which add the obvious: hospital economics are a policy variable. Reimbursement rates are set in a political process, and no discounted cash flow has a view on the next one.
What would change the number
A sustained fall in the reimbursement rate, which would reset earnings power rather than dent it. A capital expenditure cycle that stops converting into revenue growth — the current spending is defensible precisely because revenue is rising 9% a year, and that argument dies with the growth. Or a price in the $260s, where the market and the no-growth arithmetic would finally agree.
What this measurement does not cover
The board had valued 158 of the 502 priced companies in the index when this ranking was taken — 141 of them that same day. So the thirteen are the widest discounts among the names that have been valued, not a proven claim about all five hundred. The remainder are being worked through, and when the coverage is complete this piece will say so without the qualifier.
Every figure here comes from a single pipeline run dated 23 August 2026 against accounts filed with the SEC, and from those filings directly. The price used is the close of Friday 21 August 2026. Nothing in this piece is a recommendation to buy or sell any security.