TL;DR: The P/E ratio divides a stock’s share price by its earnings per share. It tells you how many dollars investors are paying for each dollar of annual profit. It is useful, it is universal, and it is easy to misuse – especially at cyclical turning points, for companies with negative earnings, and when comparing across sectors with very different growth rates.
The formula, in one line
The math is trivial:
P/E = Share Price / Earnings Per Share (EPS)
Or, equivalently, market capitalization divided by total net earnings. Both give the same number for a given company at a given moment (Corporate Finance Institute).
The intuition: a P/E of 20 means investors are paying $20 today for each $1 the company earns per year at the current rate. If earnings never grew and were paid out entirely as dividends, you would need 20 years to recoup your purchase price. In practice earnings grow (or shrink), so the ratio is really a shorthand for how much of that expected growth is already baked into the price.
Trailing P/E vs. forward P/E
There are two flavors, and confusing them is the single most common mistake:
- Trailing P/E (TTM) uses the last four reported quarters of EPS. It is a fact – real earnings that already happened.
- Forward P/E uses analysts’ consensus estimate of the next four quarters. It is a forecast – and consensus estimates are usually too optimistic entering a downturn and too pessimistic leaving one.
Rule of thumb: for stable, profitable businesses in a normal environment, forward P/E is more useful because valuation is about the future. For cyclicals, banks, or anything near an earnings inflection, cross-check both – and look at what assumptions are driving the forecast.
A worked example
Three fictional companies, three very different stories, and yet two of them have the same P/E:
| Company (illustrative) | Share price | Diluted EPS | P/E | Why the number is what it is |
|---|---|---|---|---|
| Big Software Co. | $400.00 | $8.00 | 50.0 | growth priced in |
| Mature Utility Co. | $60.00 | $5.00 | 12.0 | slow, stable growth |
| Deep Cyclical Co. | $25.00 | $0.50 | 50.0 | at the trough of its cycle |
The takeaway: the P/E is a starting question, not an answer. “50x” means one thing for a fast-growing software business and something completely different for a cyclical company whose earnings are temporarily depressed.
What’s “high” and “low” – historically
The long-run average trailing P/E for the S&P 500 is roughly 16, and the median is closer to 15 (multpl.com). At the close on September 4, 2026, the index trades at 26.4x trailing earnings – well above that historical center, but not at the extremes. Here is the full picture since 2000:
Two things jump out. First, the 2009 spike to 71x was not a bubble in prices – it was a collapse in the denominator. GAAP earnings briefly cratered during the financial crisis, so the ratio blew up mechanically even as stock prices were near their lows. Any single-period P/E built on distressed earnings will lie to you. Second, the pandemic year 2020 and 2021 pushed the ratio to 36x – a genuine multiple expansion driven by zero rates, not an earnings collapse.
This is why professional investors also watch the Shiller CAPE – the cyclically-adjusted P/E, which uses ten years of inflation-adjusted earnings in the denominator to smooth through booms and busts. Its long-run mean is 17.4 and its median is 16.1. Today it sits at 41.4, one of the most stretched readings in its 150-year history:
Note the peer: the only period in the modern record with a higher CAPE was the terminal phase of the dot-com bubble, which peaked at 44.2 in December 1999 (multpl.com Shiller PE). That is not a prediction – CAPE has been elevated for most of the last decade and stocks have kept rising – but it is a fact worth knowing when you read that “stocks are cheap.”
Sector context matters more than the market P/E
Different industries earn very different multiples for very good reasons. Software businesses have high gross margins and reinvest most of their cash into growth, so the market pays up for future earnings. Regional banks and integrated oil companies earn commodity-like returns on capital, so the market pays down for present earnings. Here is a snapshot from January 2026:
| Sector / Industry | Trailing P/E | Forward P/E |
|---|---|---|
| Advertising | 44.4 | 52.9 |
| Aerospace/Defense | 92.8 | 45.9 |
| Software (System & Application) | 79.2 | 34.1 |
| Drugs (Pharmaceutical) | 55.7 | 24.2 |
| Oil/Gas (Integrated) | 16.2 | 21.9 |
| Banks (Regional) | 33.6 | 11.0 |
| Total U.S. Market | 57.9 | 27.7 |
Comparing Nvidia’s forward P/E to JPMorgan’s forward P/E is almost meaningless – they are competing in different physics. Comparing Nvidia to AMD, or JPMorgan to Bank of America, is meaningful. Rule of thumb: always compare a company’s P/E to (1) its own history, (2) its direct sector peers, and (3) the growth rate of its earnings. A 30x P/E on a business growing earnings 25% a year is arguably cheaper than a 12x P/E on a business shrinking 5% a year.
When the P/E ratio flat-out lies to you
The ratio breaks in five recurring situations:
1. Negative earnings
You cannot divide by a negative or zero denominator in any useful way. Early-stage growth companies, restructurings, and companies during a bad year all fall off the P/E map. Analysts either use forward estimates, price-to-sales, EV/EBITDA, or normalized earnings instead.
2. Peak-cycle earnings on cyclicals
This is the classic value trap. A steel company or a semiconductor equipment maker at the peak of its cycle can print record earnings, which makes the P/E look absurdly low – right before earnings collapse and the “cheap” stock halves. Peter Lynch used to warn students: “You buy cyclicals when the P/E is high and sell them when the P/E is low.” That inversion is not a paradox; it is a description of how the ratio behaves at the top and bottom of a cycle.
3. GAAP vs. adjusted (non-GAAP) earnings
The EPS you see on a stock screener is usually GAAP. The EPS management highlights in the press release is usually adjusted – stripping out stock-based compensation, restructuring charges, one-time write-downs, and sometimes items that recur every quarter. Both numbers matter, but they are not the same number. Always check which EPS is in the denominator you are looking at. The SEC has repeatedly reminded issuers that non-GAAP measures should not be given “greater prominence” than the GAAP equivalent (SEC Compliance and Disclosure Interpretations, Non-GAAP Financial Measures).
4. Buybacks and share-count manipulation
EPS can rise even when net income falls, if the share count shrinks fast enough. Aggressive buybacks funded with debt inflate EPS mechanically. The P/E responds by falling – the stock looks cheaper. But leverage went up, not value. Always look at the EPS trajectory alongside net income and diluted share count.
5. Cross-country comparisons
A 12x P/E in emerging Europe is not the same as a 12x P/E in the United States. Different tax regimes, accounting standards, interest rates, and cost of capital all move the “fair” multiple. P/E is a within-market tool first, cross-market tool a distant second.
What to remember
- P/E is a compression device: it collapses price, growth, risk, and accounting choices into a single number. That is useful for a first look and dangerous for a final decision.
- Trailing tells you what happened; forward tells you what someone thinks will happen. Neither is “right.” Look at both.
- Context beats absolutes. 20x may be expensive for a bank and cheap for a software leader.
- The market P/E is one input, not a market call. It has been “high” for a decade.
- Sanity-check by asking what the denominator would look like in a normal year – not this year, not the peak, not the trough. Do that habitually and half your P/E mistakes disappear.
What to learn next
Once P/E feels natural, the next building blocks are: (1) PEG ratio (P/E divided by earnings growth), which puts P/E and growth in the same frame; (2) EV/EBITDA, which is capital-structure-neutral and works for companies with different debt levels; (3) free-cash-flow yield, which sidesteps accounting choices; and (4) the Shiller CAPE covered above. Each is a different lens on the same underlying question – what am I paying for a claim on this company’s future cash flows.
Sources
- S&P 500 P/E Ratio – multpl.com, current and historical trailing P/E.
- S&P 500 P/E Ratio by year – multpl.com, 2000-2026 series used in the chart.
- Shiller PE Ratio – multpl.com, based on Prof. Robert Shiller’s dataset (Yale).
- P/E Ratios by Industry – Aswath Damodaran, NYU Stern, January 2026 update.
- Corporate Finance Institute – Price-Earnings Ratio, formula and trailing vs. forward.
- SEC C&DI on Non-GAAP Financial Measures, GAAP vs. adjusted EPS presentation.
Disclosure: This article is for informational purposes only and is not investment advice.