Industry average financial ratios

There is no such thing as a strong sector. Across the sixteen metrics with a published distribution on the 2026-09-07 build, healthcare holds the highest median gross margin and the highest median current ratio of the eleven US sectors, and the lowest median on every profitability and return measure in the table.

The tables below are the reference: industry average financial ratios, stated as medians rather than means because a mean of a ratio across a sector is moved by a handful of extreme values. Everything is the median of the sector cohort on the 2026-09-07 build, one row per company, drawn from companies whose filed statements are on record. Percentages are percentages, ratios are ratios, and the units follow the metric.

Median by sector, basic materials through financial services

Median by sector, healthcare through utilities

The three findings in that grid

One metric is nearly sector-independent, and it is the fraud screen. The Beneish M-Score medians run from -2.71 in energy to -2.41 in financial services, a spread of 0.30 across all eleven sectors. Every other metric here spreads by much more, and the margins spread by tens of points. Comparing spreads across different units is loose, so here is the tight version of the claim: the registry cites Beneish's own cutoff at -1.78, and every one of the eleven sector medians sits between -2.71 and -2.41, so no sector's typical company is anywhere near the line. A cross-sector threshold on the M-Score is defensible in a way that a cross-sector threshold on a margin never is, and the grid is why.

The sector that leads on one measure rarely leads on another. Technology has the highest median price to earnings, price to book and EV to EBITDA, and the ninth-highest median return on assets of eleven. Financial services has the highest median net margin at 17.12 and the eighth-highest median return on assets at 1.18. Consumer defensive has the highest median return on assets at 3.34 and a median net margin of 2.71, seventh of eleven. Reading down a column tells you about how an industry books its costs and finances itself. It does not rank the industries.

Healthcare is the outlier and it is not close. Median operating margin -33.98, median net margin -29.71, median return on equity -38.97. Those are not distressed businesses. They are companies with a research programme and a small revenue line, and a ratio with that denominator is arithmetic rather than information.

The sample behind each cell, which is not the sector

A sector row is not a sector. It is the companies in that sector for which this build holds that figure, after the not-applicable exclusions. The two differ by a lot.

Financial services appears with 778 companies in the return on assets cohort and 79 in the return on invested capital cohort. The difference is not data quality. Invested capital is equity plus debt less cash, and for a bank both debt and cash are operating positions, so the figure is withheld for banks, capital markets firms, insurers and mortgage trusts before the cohort is counted. What is left is 79 financial companies for which the concept applies, and their median return on invested capital is -0.05.

The same effect explains the gross margin row: 61 financial services companies against 589 in the net margin row, because banks, insurers and property trusts present no cost of revenue.

Utilities never exceeds 104 companies in any row and rests on 37 in the gross margin row. The two smallest cells in the whole grid are financial services on EV to EBITDA at 35 companies and utilities on gross margin at 37. Every cell clears the thirty-company minimum this build applies, and the smallest of them are thin enough that one unusual filer moves the median.

The four cells with no number

Real estate has no gross margin row, no current ratio row and no Altman Z-Score row. Financial services has no Altman Z-Score row. In each case the endpoint returns the same sentence rather than a figure:

Fewer than 30 comparable companies have this figure on file, so a percentile would be noise. We are not showing one.

The refusals are downstream of the applicability rule. A property trust presents an unclassified balance sheet, so there is no current ratio to compute for most of the sector, and what remains is too small a group to publish a percentile against.

An empty cell in this grid is a deliberate answer. It is not a gap waiting to be filled.

What this leaves out

One build, dated 2026-09-07. Every figure moves. The distribution endpoint stamps each response with its build date and marks itself stale after 35 days, and /coverage carries the live coverage figures.

A median is one number about a group and it hides everything about the shape of that group. Several of these distributions are strongly skewed, and for those the median and the quartiles tell different stories. The full breakpoints, including the fifth and ninety-fifth percentile, are in the endpoint behind this page.

Sector is the coarse grouping this build publishes. Within healthcare a device manufacturer and a pre-revenue biotech sit in the same cohort, and within financial services a bank and an asset manager do. The finer industry level would separate them and is not published here.

Every observation was clamped to the first and ninety-ninth percentile before the breakpoints were computed. Values are clamped, never dropped, so a cohort size is the cohort that produced the number beside it.

The three market-priced metrics, price to earnings, price to book and EV to EBITDA, combine a price with filed fundamentals. Every price in this system was older than a week when this build ran.

A percentile describes where a figure sat among comparable companies. It is not a view on any company or any sector, and nothing here says any of these numbers should be used as a threshold.