YIELDTERMINAL
Build My Free Plan
← Back to all studies

Study · S&P 500 names near 52-week lows

The De-Rating List: Same Screen, Opposite Meaning

136 S&P 500 names sit within 10% of their 52-week low. Splitting them by operating income separates a multiple coming in from earnings going the wrong way, and most lists never do it.

Close of 25 Sep 2026

Published Sep 26, 2026, 6:45 PM ET

Data through Sep 26, 2026

TL;DR

  • 136 S&P 500 names are within 10% of their 52-week low. 81 have operating income higher than the year before, 24 have it lower, and 31 cannot be determined from filings yet.
  • Only the first group is a de-rating. In the second the market is repricing earnings that actually fell, and lists that mix them together are doing the reader no favours.
  • Lower earnings does not mean a shrinking business. $CSGP grew revenue 18.7% last year and still swung to an operating loss, because it spent. That is compressed margin, not contraction.
  • Operating income, not net income. Net carries impairments, interest and tax. Of four widely-listed names with net income down, only one had operating income down too.
  • Every figure is operating income from each company's own two most recent annual periods in its SEC filings, never an estimate feed. Where operating and net income point opposite ways, that disagreement is flagged.

The problem with every list of stocks near their lows

They mix two completely different situations and never separate them.

A stock can be 50% off its high because the market decided to pay less for the same earnings. That is a de-rating, and it is the thing worth looking at. Or it can be 50% off its high because earnings went the wrong way. That is not a de-rating, it is the market repricing a different set of numbers.

Printing both under one heading and calling them quality is the sleight of hand. This list splits them and shows the receipt for each row.

Lower earnings is not a shrinking business

This distinction matters and it is easy to get wrong, so here is the clearest example on the list.

$CSGP, CoStar Group. Revenue went from US$2,736 million to US$3,247 million in its last fiscal year, up 18.7%. Over the same year operating income went from positive US$5 million to negative US$72 million. Both figures are from the FY2025 10-K filed 26 February 2026.

That company is not shrinking. It is growing revenue fast and spending more than the revenue it added. The stock is 67% off its high and earnings went the wrong way, and those are both true, but calling it a business in decline would be false.

So the bucket is named for what it measures. The second group is not "the business shrank". It is "earnings went the wrong way", and the reason differs company by company.

Why operating income and not net income

This is where our own first version of this screen went wrong.

Net income carries impairments, interest expense and tax artefacts. None of those is the operating business. Take four widely-listed names whose net income fell over the last year: only one of them had operating income fall as well. One was dragged down by interest on an acquisition, one by a single large impairment, one by acquisition costs against a prior-year gain.

So the bucket is decided on operating income, net income is printed beside it, and where the two disagree the row is flagged. That disagreement is usually the whole story.

The staleness trap

A company whose fiscal year ends in July or August closes and reports its new year by 8-K several weeks before the 10-K exists. For those weeks, the newest annual figure in SEC XBRL is a full year old while the market already has the new numbers.

A screen that reads calendar-year frames will print that stale year as current and get the direction wrong. Two names flipped bucket on exactly this when we checked our own output against the filings. Anything whose newest annual period ended more than 365 days ago is now excluded as unknown rather than shown, which is why the unknown column is not small.

What the screen says this week

136 names within 10% of their 52-week low. 81 with operating income higher than the year before, 24 with it lower, 31 we will not guess at.

The table below shows the deepest drawdowns in each bucket.

Educational only. Not advice.

Every episode (16)

CompanyBucketFrom 52w highOperating incomeFiscal year usedNotes
$PODD Insulet Corporationde-rating-61.6%+53.4%FY ended 2025-12-31Operating and net point opposite ways. Net income -40.9%.
$INTU Intuitde-rating-60.8%+19.5%FY ended 2026-07-31Net income +18.0%.
$BSX Boston Scientificde-rating-58.4%+38.8%FY ended 2025-12-31Net income not comparable on the same basis.
$APP AppLovinde-rating-58.3%+117.3%FY ended 2025-12-31Net income +111.0%.
$FICO Fair Isaacde-rating-56.8%+26.1%FY ended 2025-09-30Net income +27.1%.
$ROL Rollinsde-rating-54.6%+10.5%FY ended 2025-12-31Net income +12.9%.
$PNR Pentairde-rating-52.9%+6.7%FY ended 2025-12-31Net income +4.5%.
$FIS Fidelity National Information Servicesde-rating-48.8%+1.9%FY ended 2025-12-31Operating and net point opposite ways. Net income -73.7%.
$CSGP CoStar Groupearnings fell-67.0%FY ended 2025-12-31Revenue GREW 18.7% to US$3,247M while operating income went from +US$5M to -US$72M. Percentage suppressed because the prior-year base is near zero. Net income -95.3%.
$FISV Fiservearnings fell-64.4%-1.0%FY ended 2025-12-31Operating and net point opposite ways. Net income +11.1%.
$CHTR Charter Communicationsearnings fell-60.5%-1.6%FY ended 2025-12-31Net income -1.9%.
$LULU Lululemon Athleticaearnings fell-55.2%-11.8%FY ended 2026-02-01Net income -13.0%.
$APTV Aptivearnings fell-49.6%-35.7%FY ended 2025-12-31Net income -90.8%.
$NRG NRG Energyearnings fell-47.2%-23.9%FY ended 2025-12-31Net income -23.2%.
$WYNN Wynn Resortsearnings fell-39.8%-1.3%FY ended 2025-12-31Net income -34.7%.
$COO Cooper Companiesearnings fell-38.4%-3.2%FY ended 2025-10-31Net income -4.4%.

How this was measured

Universe is the S&P 500 constituent list plus large US-listed ADRs. Prices are daily closes from the Yahoo Finance chart API, and every row is computed in a single pass so no row is staler than another; rows not sharing the dominant as-of date are dropped. Drawdown is the last close against the highest high of the trailing year. Earnings are us-gaap OperatingIncomeLoss and NetIncomeLoss read from SEC EDGAR XBRL, taking each company's own two most recent ANNUAL periods by period-end date rather than a calendar-year frame. Three hard gates apply. First, universe coverage must clear 97% or the ranking is abandoned, because a partial universe makes a 'deepest drawdown' list meaningless. Second, a company whose newest annual XBRL period ended more than 365 days ago is excluded as STALE rather than shown, because an off-calendar fiscal year reports by 8-K weeks before its 10-K exists and the newest XBRL annual figure can be a full year behind reality. Third, a percentage change off a near-zero or negative prior-year base is suppressed and only the direction is shown, because the arithmetic is real but the number is noise. An empty units block from the SEC companyconcept endpoint is treated as a fetch failure and retried against companyfacts, never as an absence of earnings.

Sources

Discuss on XAll studies

Educational commentary, not investment advice. See the full disclaimer.