Company Screener

Every example below uses Micron (MU) as of its June 2026 filing — click MU's row in the table to see all of these inputs yourself. "12-month" always means the four most recently reported quarters added together.

Valuation

• Price / earnings — company's total market value ÷ its 12-month profit.
  $1,097B market value ÷ $50.5B profit = 21.7× — you pay $21.70 for each $1 of yearly profit. Blank when the trailing year was a loss (there is nothing to divide by).
• Price / expected earnings (forward) — share price ÷ what analysts expect per share over the NEXT year.
  $971.66 share price ÷ $154.89 expected earnings per share = 6.3×. Much lower than the 21.7× trailing figure because 43 analysts expect profits to keep exploding. A forecast, not a filing. Nobody sells the history of past forecasts, so since August 2026 we record the consensus for every company each trading day — each company's detail panel charts that record as it accumulates.
• Versus its own past — we compute the price/earnings ratio at every month-end since 2016 (using only the earnings known on each date), then ask: what share of those readings were at or below today's?
  MU today: 21.7× is higher than 69% of its own month-end history → 69%. "Cheaper than usual" keeps companies at 50% or below (below their own median); "near its historic lows" keeps 25% or below. 0% = cheapest it has ever been.
• P/E on typical quarters — one-off events (impairments, legal settlements, asset-sale gains, investment marks) can swing a single quarter so hard that the 12-month profit — and every ratio built on it — lies. The fix: take the MEDIAN of the last four quarters and annualize it (a one-quarter spike barely moves a median), then divide price by that. When the two P/Es disagree sharply, the row shows which quarter distorted it and by how much, and the company's detail panel carries a warning. Named one-off tags exist in filings but only 6-20% of companies use them, so this stays statistical — it catches the event whatever it was called.
• Price / net assets — market value ÷ what the company owns minus what it owes ("book value").
  MU: $1,097B ÷ $101B net assets = 10.9× vs JPMorgan 2.2× vs Lincoln National 0.8×. Below 1× you pay less than the assets on the books are worth — combined with a real return on equity (Lincoln earns 21% on its) that is the classic value setup for asset-heavy businesses. Asset-light businesses run high multiples harmlessly (Visa: 18.5×), so compare within a sector.

Growth

• Revenue growing at least X% — last 12 months of revenue vs the 12 months before that.
  ($90.3B − $33.8B) ÷ $33.8B = +167% → MU passes the "100%" setting.
• Profit growing at least X% — same arithmetic on profit.
  ($50.5B − $6.22B) ÷ $6.22B = +711%. This is why the "last year's profit at least" filter exists: a company going from $1M to $50M shows +4,900% — technically true, practically meaningless. Requiring a real base (default $25M) keeps the percentages comparable.
• Growth speeding up — compare this quarter's year-over-year growth with last quarter's.
  Latest quarter revenue +345.7% vs prior quarter +196.3% → +149 points faster → "speeding up".
• Revenue growth, smoothed — revenue rarely rises in a straight line, so this column averages the last few quarters' year-over-year growth with an exponential moving average (3-quarter span: the newest quarter carries ~57% of the weight, older quarters fade quickly). One lumpy quarter barely moves it. The small "trend" line under the number says whether this smoothed pace itself sped up or slowed vs one quarter ago — a steadier read than the raw comparison above, which flips for half of all companies every quarter (measured in experiment 014). The average is taken over the growth percentages, never over raw quarterly revenue — raw quarters are seasonal (comparing Christmas against summer), while year-over-year growth already cancels the seasons.
• Revenue up for N straight quarters — walk backwards from the newest quarter, counting while year-over-year growth stays positive.
  MU: +345.7%, +196.3%, +56.7%, +46.0% … nine positives in a row before a decline stops the count → streak = 9.

Quality

• Profit margin at least X% — 12-month profit ÷ 12-month revenue.
  $50.5B ÷ $90.3B = 55.9% — MU keeps 55.9 cents of every revenue dollar as profit.
• Margin improving — this year's margin minus last year's.
  55.9% now vs 18.4% a year ago ($6.22B ÷ $33.8B) → +37.5 points → improving.
• Profitable for N straight quarters — count backwards while each quarter's profit is above zero.
  MU: 10 straight (the last of the memory-downturn losses, in late 2023, broke the prior streak).
• Turned profitable — the prior 12-month total was a loss, the current 12-month total is a profit. These are ranked by the dollar swing, since a percentage from a negative base has no meaning.
• Net debt — total borrowings minus cash and short-term investments, from the latest balance sheet.
  MU: $5.7B debt − $25.0B cash = net cash of $19.3B (shown green). Charter, for contrast: $94.0B debt − $0.5B cash = $93.5B net debt ≈ 7.4 years of operating profit to repay.
• Cash minus debt, % of price — the same net debt figure as a share of the market value, so companies of any size are comparable.
  Alphabet: +3.4% (a $144B cash fortress, but small next to its $4.2T price); Charter: −508% — a buyer takes on five times the sticker price in debt. Blank for banks and insurers, whose cash backs deposits and policy reserves, not shareholders.
• Debt vs equity — total borrowings ÷ what shareholders own. CHTR: $94.0B ÷ $17.0B = 5.5× (heavily leveraged) vs MU: 0.01× (essentially unleveraged).
• Full price / operating profit — the debt-aware price tag an acquirer pays: (market value + debt − cash) ÷ yearly operating profit.
  A company with heavy debt is more expensive than its price/earnings suggests, because a buyer inherits the debt. The "Debt at most" filter uses years-of-operating-profit-to-repay.

Cash & balance sheet

• Free cash flow — cash from operations minus cash spent on buildings & equipment, from the cash flow statement (the third statement in every filing, and the hardest to dress up).
  MU: $51.4B came in from operations − $25.3B spent building fabs = $26.2B actually left over. Note: quarterly filings report cash flow cumulatively (3, 6, 9 months), so we derive each quarter by subtraction — the detail panel's automatic check verifies the four quarters re-add to the 12-month total.
• Free cash flow yield & growth — the size-neutral cash columns. Yield: 12-month spare cash ÷ market value — MU: $26.2B ÷ $1,097B = 2.4%; a $300M company with $30M of spare cash scores 10% and outranks every giant. Growth: newest four quarters vs the four before, so cash compounding shows up as a percentage regardless of company size. The dollar column stays for scale context — these two exist so sorting doesn't just rank the biggest.
• Building spend, % of revenue — capital spending as a share of revenue: how heavily a company is investing in assets right now.
  The AI build-out in one column: Meta 39%, Microsoft 35%, Alphabet 30%, MU 28% of every revenue dollar is going into data centers and fabs — and Oracle 83%, driving its free cash flow to −$23.7B despite fat profits. Visa, asset-light, spends 3.5%. When a heavy builder's number falls, the harvest begins: free cash flow jumps above profit and the debt gets repaid — exactly the MU story above.
• Cash backing the profit — cash from operations ÷ reported profit.
  MU: $51.4B ÷ $50.5B = 102% — every reported dollar was actually collected in cash. Profit is an accounting opinion, cash is a fact: a company persistently far below 100% is reporting profits it isn't collecting. Well above 100% usually just reflects heavy depreciation (Charter: 335%).
• Return on shareholders' money — 12-month profit ÷ shareholders' equity: how hard the owners' money works. MU: $50.5B ÷ $101B = 50% — the test of whether big investments were worth making.
• Profit vs interest bill — yearly operating profit ÷ yearly interest expense.
  Charter: $12.7B ÷ $5.1B = 2.5× (the debt is a real burden) vs MU: 254× (irrelevant). Below 2× a company is working mostly for its lenders.
• Borrowings trend — the "paying off a massive investment" signal: total borrowings now vs four quarterly balance sheets ago.
  MU: $15.5B → $5.7B = −63% in a year, four falling quarters in a row, funded by that free cash flow. The detail panel's borrowings chart shows this crossing visually.
• Share count — diluted shares vs a year ago. Apple: −1.6% ($82B of buybacks) — each remaining share owns more. Young companies often grow the count +3-5% yearly through stock pay, quietly diluting owners (stock pay is shown in each company's detail panel).
• Short-term bills covered — assets that turn into cash within a year ÷ bills due within a year. MU: 3.4× is comfortable; below 1× means borrowing just to pay bills (though subscription businesses run below 1× harmlessly, because customers pay them up front).

Company

• US only / went public / employees — straight lookups: country and IPO year from the NASDAQ listing directory, employee counts from Yahoo Finance (companies state headcount in prose, not in tagged filings).
• Exclude banks & lenders — drops the Finance sector. Their "revenue" is interest income under different accounting, which makes revenue growth and margin columns misleading for them (profit columns are still valid).
• Closest peers — every company's detail panel opens with a side-by-side table of its most similar companies, found by blending how alike the two companies describe their own business (from their profile text), how closely their share prices moved over the last year, and whether they share an industry label. The blend matters because industry labels alone mislead: APLD files as "Technology — EDP Services" while IREN and CIFR file under "Finance", yet all three run bitcoin/AI data centers — description and price co-movement catch that. Hover "why it's listed" for the exact numbers behind each match.