A good current ratio: 2,044 clear 2.0

Graham's second defensive-investor criterion asks for current assets of at least twice current liabilities. On 2026-09-07, 2,044 of the 4,399 US listings holding a current ratio cleared it. That is 46.5 percent of them.

A screen that passes nearly half the market is not selecting for anything. The interesting question is what changed, and the answer is in the sector split.

Counted with POST /api/screener/search on 2026-09-07, filtering current_ratio at or above 2.0. The public endpoint reports at most fifty matching rows, so a query that comes back under fifty is reporting a true count. The pool was split by sector and then by market capitalisation until every part came back under fifty, and the parts were added. Listings in the current pool of 5,257, not deduplicated companies.

Healthcare alone is 689 of the 2,044, a third of everything that clears the line.

A liquidity test that a company with no operations passes trivially

The current ratio is current assets over current liabilities. A clinical-stage biotechnology company has raised cash, holds it as a current asset, and has almost no payables because it buys very little. Its current ratio is enormous and it means nothing about its ability to fund itself, which depends entirely on how long the cash lasts against the burn.

The distribution confirms the shape. Across 3,673 companies with filed statements, excluding the classes where the ratio is not defined:

From GET /api/public/distribution/current_ratio, build dated 2026-09-07, rounded to two decimals, halves away from zero. That table is one build and stays as it was read.

The two market-wide marks are not. They are read from the same endpoint when this page is served and each prints the build it came from: the all-market median is 1.87 (built Oct 21, 2026) and the 75th percentile is 3.69 (built Oct 21, 2026).

The healthcare median of 3.40 is nearly double the market's. The utilities median is 0.80: more than three quarters of that cohort holds less in current assets than it owes within the year, which for a regulated business with predictable receipts and permanent access to the commercial paper market is ordinary rather than alarming. The same sector fails a second published threshold for the same reason, and reads as speculative grade on interest coverage while issuing investment-grade debt.

The largest company in the market has broken the rule for a decade

The same rule, run backwards over what companies actually filed, is published for three companies on this site.

From GET /api/public/research/index, read 2026-09-07. Microsoft's rule broke in 2022 and has stayed broken through 2026. Texas Instruments has never broken it. Apple's ten years, each from its own 10-K:

The ratio has not been above 2.0 once in ten years, and since 2022 it has been below 1.0 in every year. No share price enters any figure in that table. A guardrail written on this line for this company would have been broken on the day it was written.

Apple is not a liquidity risk. It runs negative working capital on purpose, because it collects from customers before it pays suppliers, and a company that can do that is funding itself with its supply chain rather than with its balance sheet. Graham's test reads that as weakness because in 1949 a company with current liabilities approaching its current assets usually was weak.

So the criterion now does close to the opposite of what it was written to do. It passes a large population of cash-holding companies with no operations, and it fails several of the most cash-generative businesses in the market. Both errors have the same cause: the ratio measures a stock of assets rather than the speed at which cash arrives.

What the test is still good for

Two things, and they are narrower than the original claim.

Direction, on one company, over years. A current ratio falling from 2.4 to 1.1 over four filings is a fact about that company that no cross-sectional percentile can tell you. The retro-test above is the useful form of the question: not "does it pass" but "in which years did it stop passing".

Comparison within an industry. A current ratio of 1.2 means something very different in utilities, where the 75th percentile is 1.15, than in healthcare, where the 25th is 1.66. The peer percentile is defensible. The absolute threshold is not.

Where the ratio is not computed at all

617 companies were excluded from the distribution cohort as not applicable: banks, equity REITs, insurers and mortgage REITs. None of them present a classified balance sheet in the form the ratio needs, so there is no honest current asset and current liability line to divide. Real Estate publishes no sector percentiles here at all, because after the REIT exclusions fewer than thirty companies remain and thirty is the smallest cohort this product will compute a percentile from.

The count of 2,044 is a different measurement and includes those companies, because a screener filter matches the stored column before the applicability gate runs. Excluding the Financial Services and Real Estate sectors, 1,879 of the remaining listings clear 2.0. That is a rough correction and not the same as applying the gate, because the gate classifies by filed industry rather than by sector.

What this leaves out

The count of 2,044 is listings in the pool as it stood on 2026-09-07, not deduplicated companies, and it counts stored values including those the product would refuse to display on a company page.

The percentiles in the sector table are cross-sectional, from one build dated 2026-09-07, on US listings at the verified tier, one company per row, winsorised at the 1st and 99th percentile. Six companies were dropped market-wide for a current ratio of exactly zero, on the rule that an exact zero here is a coerced missing value. The two market-wide marks below the table are read when the page is served, so they answer for a later build than the sector rows do and the two will not always agree to the decimal.

The 2.0 threshold is Graham's, from The Intelligent Investor, revised edition 1973, chapter 14. It is a published convention and not a boundary this product estimated.

The retro-test covers only the fiscal years we hold as filed statements from EDGAR, and only three companies are published for this rule. Three companies is an illustration, not a sample. Nothing here is a view on Apple, Microsoft, Texas Instruments or any other company.

The current ratio says nothing about the quality of the current assets. A receivable nobody will pay and a treasury bill count the same in the numerator. The next question after this ratio is which parts of it are cash, and that is the quick ratio.