Original data · Recomputed hourly

US market valuation snapshot

Owl Labs is a data and screening service, not a registered investment adviser. Every figure below is recomputed from our database when the page is served and cached for up to one hour, so the numbers you see may differ from a screenshot taken yesterday. Sample sizes and as-of dates are printed next to each figure precisely so the numbers can be checked rather than taken on trust.

What exactly does this snapshot measure?

It measures the Owl Labs US coverage universe — the US-listed operating companies we collect financial statements for, of which 678 carried a score on 2026-08-11. It is a coverage universe, not the entire US market, and it is not an index: there is no weighting, no float adjustment and no committee. Three exclusions matter when you read the numbers below.

Two separate counts appear on this page and they do not measure the same thing: 678 companies were scored in the run of 2026-08-11, while 678 companies had a current quote when this page was built. They come from different tables and need notagree: the scoring run covers only the companies flagged as screening candidates and reflects that day’s batch, while the quote count is taken live when this page is built. Grade figures below rest on the first count and valuation figures on the second, and each figure is labelled accordingly.

• Exchange-traded funds are excluded

  A fund’s reported P/E is a weighted average of the companies it holds. Leaving them in would double-count those companies and pull the distribution toward whatever the largest funds hold.

• Loss-making companies drop out of the P/E figures

  A P/E is undefined when earnings are negative, so companies with non-positive trailing earnings are simply absent from the P/E sample — which is why the P/E sample is smaller than the number of companies we hold a quote for.

• Extreme multiples are trimmed

  Trailing P/E above 300x and price-to-book above 50x are excluded. A company that just crossed back into profit can print a four-digit P/E, and a handful of those would visibly move the quartiles without telling you anything about the typical company.

What are the median P/E and price-to-book across US listings?

As of 2026-08-11 21:22 UTC, the median trailing P/E was 24.7x across 534 companies with positive earnings, with a first quartile of 17.2x and a third quartile of 36.3x. The median price-to-book was 3.48x across 579 companies. Half of those 534 companies therefore sit between 17.2x and 36.3x on earnings — a spread wide enough that a single market-wide multiple explains very little about any one company.

A P/E is price divided by earnings per share over the trailing twelve months, so it moves when either the price or the last four quarters of profit change. A price-to-book is price divided by book equity per share, which is informative for banks, insurers and asset-heavy industries and much less so for businesses whose main assets are intangible — research, brands and software rarely appear on a balance sheet, so book equity understates them and the ratio reads high by construction. Companies that have bought back stock for years can even report negative book equity, at which point the ratio stops carrying meaning altogether. Both figures describe where a company sits in a distribution; neither one, on its own, establishes that anything is cheap or expensive.

How many US companies pay a dividend, and what is the median yield?

Of the 678 covered US companies that had a current quote as of 2026-08-11 21:22 UTC, 433 had a positive trailing dividend yield, and among those payers the median yield was 1.69%. 108 companies yielded 3% or more. Note that the median is taken over payers only — including the non-payers would drag it toward zero and describe something different.

Yield is the trailing twelve months of dividends divided by the current price, which means it rises mechanically when a price falls. A yield that stands out from its sector is a prompt to look at the payout ratio and the cash flow behind it, not a conclusion in itself. US companies also return cash through buybacks, which never show up in a yield figure at all, so a low market-wide yield does not by itself say shareholders are receiving less.

How are today’s grades distributed?

On 2026-08-11, 678 US-listed companies were scored and the average composite score was 57.6 out of 100. The grade distribution was 36 A, 209 B, 244 C and 189 D.

The composite score starts from six published financial metrics — P/E, price-to-book, return on equity, operating margin, EV/EBITDA and dividend yield — together with a market-attention axis built from relative trading volume and recent price action, which is market data rather than a reported financial figure. Those seven axes are weighted onto a common scale and then adjusted: a sector adjustment, a momentum and flow overlay and a deduction for serious disclosure risk are applied on top, and the result is capped both by how many of the six metrics we actually hold for that company and by a soft ceiling that keeps any single score below 100. The axis breakdown is shown for every company. A grade is a compact summary of that calculation, not a forecast and not a buy or sell signal; the distribution above is included so you can see that grades are relative to the companies scored that day rather than absolute judgements.

Which sectors carry the highest and lowest median P/E?

Sector medians spread far wider than the market-wide figure. Among the 11 sectors shown below — those with at least 15 covered companies with positive earnings — Information Technology had the highest median trailing P/E at 36.7x and Financials the lowest at 16.1x, as of 2026-08-11 21:22 UTC. This is why comparing a company against its own sector is more informative than comparing it against the market as a whole.

The company count is the number of covered companies in that sector with a usable trailing P/E — that is, positive earnings and a multiple inside the trim — so it is smaller than the number of companies we cover in the sector, and the P/B and yield columns are medians over their own smaller samples again. Median yield is taken over dividend payers only. Sectors with fewer than 15 qualifying companies are omitted because a median over a thin sample is not interpretable. Sector labels are normalised to one naming scheme before grouping, because our sources label the same sector differently (for example Healthcare and Health Care).

How many companies pass the five value screens?

We also run five published value-investing rule sets over the same universe and record which ones each company satisfies. As of 2026-08-11, 671 US-listed companies were evaluated and 27 satisfied at least three of the five rule sets, of which 6 satisfied at least four. The point of the number is its scale: strict mechanical value criteria are satisfied by a small minority of listed companies at any given time.

Some inputs are explicit approximations where the underlying data is not available at statement level — operating income stands in for EBIT, and free cash flow is approximated from operating cash flow. Passing a screen is a statement about published accounting ratios on one date, not an assessment of a business and not a recommendation.

How current are the numbers behind this page?

Prices and market-derived ratios are refreshed every half hour while the market is open (last successful refresh run before this snapshot: 2026-08-11 21:22 UTC), while the scores and grades are rebuilt once per weekday, after the US close. The financial statements underneath move much more slowly, and that is set by regulation rather than by us: a US large accelerated filer must file its annual report on Form 10-K within 60 days of its fiscal year end and its quarterly report on Form 10-Q within 40 days of quarter end, with smaller filer categories allowed more time (US Securities and Exchange Commission, periodic report filing deadlines; verified as of August 2026). Earnings-based ratios use the trailing twelve months, so a company that has just reported is reflected in the next rebuild, and one that has not reported in months is still carrying older figures.

Where can you check these numbers yourself?

Every aggregate on this page is derived from the same tables that drive the tools below, and the calculation behind each individual score is shown on the company pages — which is the point. A statistic you cannot decompose is a statistic you have to take on faith.

Frequently asked questions

Method: all figures are computed at request time from the Owl Labs database and cached for up to one hour. Market data is sourced from Yahoo Finance; company financial statements are sourced from public filings via our data providers, with Korean filings from DART. Exchange-traded funds are excluded, trailing P/E is trimmed to 0–300x and price-to-book to 0–50x, sector labels are normalised to a single naming scheme before grouping, and sector medians require at least 15 qualifying companies. The as-of timestamp is the start of the last quote-refresh run that completed successfully, not the timestamp of the freshest individual row. Filing deadlines cited above are those published by the US Securities and Exchange Commission and were verified as of August 2026. Owl Labs is an information and analytics service and is not a registered investment adviser. Nothing on this page is investment, legal or tax advice, and no figure here is a recommendation to buy or sell any security.