What Is a Good P/E Ratio?

A price-to-earnings ratio of 20 is expensive for a bank and cheap for a software company. That is the whole answer, and it is why the question has no single number attached to it. What it does have is a set of ranges, and those we can measure.

Eleven sectors, and the gap between them

Below is the trailing P/E of every company SageRanks tracks that reports one, grouped by sector and sorted from cheapest to dearest. The middle column is the median. The two either side are the quartiles: a quarter of the sector sits below the first, a quarter above the third.

The distance from one end to the other is the point. Financial services sit at a median of 14.8 and industrials at 25.1, so the same P/E of 20 puts a bank in the expensive half of its sector and an industrial firm in the cheap half. Any rule of thumb that ignores which sector a company is in will be wrong most of the time, in whichever direction the sector happens to lean.

The spread matters more than the middle

Almost every published table gives one number per sector. That number hides the more useful fact.

Look at technology. Its median is 25.0, which sounds precise. But its quartiles are 17.6 and 48.9, meaning half the sector is spread across a range nearly three times as wide at the top as at the bottom. Financial services, by contrast, run from 11.9 to 19.1. Two sectors, similar company counts, and utterly different dispersion.

So a technology company on a P/E of 30 is entirely unremarkable, sitting comfortably inside the middle half of its own sector, while a bank on 30 is an outlier that deserves an explanation. Neither of those readings is available from a median alone, which is why we publish the quartiles next to it.

The same sector, three regions

Sector is not the only thing that moves the number. Here is the median P/E for five large sectors split by region, with the company count in each cell.

Technology is remarkably flat: 24.7 in North America, 24.2 in Europe, 26.6 in Asia-Pacific. The market prices software and semiconductors at much the same multiple wherever they are listed.

Financial services are not flat at all. European banks and insurers sit at a median of 13.3 against 18.7 in North America, a gap of about 40 per cent in a sector where the businesses are broadly comparable. Whether that is a discount worth having or a discount that is deserved is exactly the sort of question a screen cannot answer for you.

One caution on this table. Our North American coverage is 156 companies, so once it is split eleven ways the cells hold 14 to 29 companies each. Those medians are indicative, not precise, and we would not build an argument on a difference of one or two points between them.

Why these numbers differ from ones you have seen

If you have looked up sector P/E ratios before, our figures will not match, and the reason is worth understanding because it applies to every such table you will read.

Siblis Research publishes sector P/E ratios for US large caps, and as of 30 June 2026 they had information technology at 39.21, financials at 17.28 and real estate at 42.70. We have technology at 25.0, financial services at 14.8 and real estate at 15.4. Those are not competing measurements of the same thing.

Three differences account for it. Theirs covers the 500 largest US companies; ours covers 844 companies across 30 indices in three regions. Theirs is an aggregate for the sector, which the largest constituents dominate; ours is the median of individual companies, where a trillion-dollar company counts once and so does a small one. And the dates differ by a month.

Real estate shows how far that can push a number: 42.70 in an aggregate of US large caps against a global median of 15.4. Property earnings are heavily reduced by depreciation charges, so a handful of constituents with very small earnings can lift an aggregate enormously while barely moving a median. Neither figure is wrong. They answer different questions, and a table that does not tell you which question it answers is not much use.

Cheap is not the same as good

The sectors at the top of the first table are cheap for reasons, and the reasons are visible in the other numbers we hold.

Real estate has the lowest price-to-book of any sector at a median of 0.97, meaning the typical property company trades slightly below the accounting value of what it owns. It also has the lowest return on equity, at 6.8 per cent. Technology sits at the opposite corner: a price-to-book of 5.14 and a return on equity of 20.5 per cent. The market is not being irrational in paying more for the second; it is paying for the returns.

This is the trap in every low-P/E screen, including the ones we run. Sorting by cheapness reliably surfaces banks, property and energy, and it does so whether or not those industries are actually attractive right now. A low multiple is the market's opinion, not its mistake, and the work of deciding which it is in a given case is yours.

What we do with this

Two things, both of which this measurement is the reason for.

First, every ranking on SageRanks is relative within its index rather than against an absolute threshold. A Graham first place in one market means the least expensive of that particular bunch, and comparing a rank across two markets with different sector mixes will mislead you.

Second, we built sector rankings precisely so that like can be compared with like. Ranking a bank against other banks tells you something; ranking it against a semiconductor designer mostly tells you that banks have different balance sheets.

Sources