Open finviz, sort the S&P 500 by EV/EBIT ascending, and you get a list. Open Sperio, do the same, and you get a different list. Neither is broken. They are measuring different things, and the difference is the entire reason this screener exists.
The denominator is the whole argument
EV/EBIT is enterprise value divided by operating profit. Everybody agrees on the numerator: market capitalisation, plus debt, minus cash. The disagreement is underneath the line.
Free screeners use TTM — the trailing twelve months. It is the most recent number available, it is unambiguous, and it is what a data provider can compute for every company on the same day. Sperio uses the median operating profit of up to fifteen fiscal years, read from SEC EDGAR filings.
For a business whose profits barely move, the two are nearly identical. For a business with a cycle, they are not close at all — and a screener sorted ascending will systematically put the second kind at the top of the list at exactly the wrong moment.
Why that matters at the top of the list
A cyclical company at the peak of its cycle has a large trailing profit. Divide a normal enterprise value by an unusually large denominator and the multiple looks small. The screener says cheap. What it has actually measured is that last year was good.
The same company at the bottom of its cycle has a small trailing profit, so the multiple looks enormous, and the screener buries it — at the point in the cycle where its earnings power is most understated.
Ranking an index on TTM therefore does not sort by cheapness. It sorts, in part, by where each company happens to sit in its own cycle. A median across fifteen fiscal years does not fix everything, but it removes that particular illusion: a full cycle contains the good years and the bad ones, and the middle of it is a fairer guess at what the business earns in an ordinary year.
What it costs
This is not free. Three honest objections:
It is backward-looking. A fifteen-year median describes a business that existed for fifteen years. For a company that has genuinely changed — a divestiture, a new model, a transformative acquisition — the old years describe something that is gone. The median will understate it, and no amount of arithmetic fixes that.
It needs history. A company listed four years ago has no cycle to average. Sperio declines to rank those rather than averaging what little exists and presenting the result as comparable.
It reads high across the index. Measured on the 315 ranked names on 15 August 2026, the median full-cycle multiple was 38.1x, with a first quartile at 25.3x. Those numbers look alarming next to the TTM multiples published everywhere else. They are not comparable, and treating a 30x full-cycle multiple as if it were a 30x TTM multiple is the single easiest way to misread this screen.
A concrete case
Verizon comes out at 12.9x on the median operating profit of fifteen fiscal years through 2025 — among the cheapest names in the index on this measure.
That is where a screener's job ends. It has found a candidate; it has not formed an opinion. The board that runs afterwards prices the business giving no credit at all to growth and arrives at an intrinsic value of $36.58 a share, against a market price of $49.19 — a margin of safety of −25.6%.
Cheap on a screen and cheap against what a business actually earns are two different statements. The first is a filter. The second is an argument, and you can read every step of it.
What this is not
None of this is a recommendation to buy or sell anything, and none of these numbers predicts a share price. A multiple is a description of what you are paying relative to a profit, nothing more. It is silent about whether the profit will be there next year — which is precisely why the multiple is where the work starts, not where it ends.