Price-to-Earnings Ratio Explained: Trailing, Forward and CAPE

The price-to-earnings ratio (P/E) is a stock’s share price divided by its earnings per share. It tells you how many dollars investors are paying today for each dollar of a company’s annual profit.

In short: P/E = share price ÷ earnings per share. A P/E of 20 means the market price is 20 times the company’s annual earnings per share. Trailing P/E uses the last twelve months of reported earnings. Forward P/E uses analysts’ estimates for the next twelve months. The cyclically adjusted P/E (CAPE), popularized by Robert Shiller, divides an index’s price by its average inflation-adjusted earnings over the previous ten years. A high P/E means investors are paying more per dollar of current profit, usually because they expect growth, lower risk, or both. It does not, on its own, mean a stock is overpriced.

The basic arithmetic

Take Maple Widgets Inc., a hypothetical company:

Share price: $45.00
Net income over the last four quarters: $90 million
Weighted average diluted shares: 40 million
So trailing earnings per share = $90 million ÷ 40 million = $2.25

Trailing P/E = $45.00 ÷ $2.25 = 20.0

You get the same answer at the company level: market capitalization of $45 × 40 million = $1.8 billion, divided by $90 million of net income = 20.0.

Forward P/E

Suppose analysts on average expect Maple Widgets to earn $2.50 per share over the next twelve months. Forward P/E = $45.00 ÷ $2.50 = 18.0.

Forward P/E is lower whenever earnings are expected to grow. It is also built on a forecast, and forecasts are often wrong. Trailing P/E uses reported numbers but looks backward. Neither is better in every case; know which one a quote is using.

Earnings yield: P/E turned upside down

Earnings yield = EPS ÷ price = 1 ÷ P/E. For Maple Widgets: $2.25 ÷ $45.00 = 5.0%.

This framing makes comparison easier. A P/E of 25 is an earnings yield of 4%; a P/E of 10 is 10%. It shows how much profit each dollar of share price is buying. Note that earnings yield is not what you receive in cash. The company keeps part of its earnings and pays out part as dividends, so the dividend yield is usually lower.

Why P/E ratios differ so much

Two companies with the same EPS can deserve very different P/Es. The main drivers are:

Growth

If a company’s earnings are expected to grow quickly, today’s earnings understate what investors are buying. Here is a simple illustration with two hypothetical companies, each earning $1.00 per share today:

	Company A (grows 15% a year)	Company B (grows 3% a year)
EPS today	$1.00	$1.00
EPS in year 5	$1.00 × 1.15^5 = $2.01	$1.00 × 1.03^5 = $1.16
EPS in year 10	$1.00 × 1.15^10 = $4.05	$1.00 × 1.03^10 = $1.34
At a price of $30: P/E on today’s EPS	30	30
P/E on year-10 EPS	7.4	22.4

At the same price, Company A looks expensive on today’s earnings and cheap on year-10 earnings, if the growth actually arrives. That “if” is where most of the risk lives.

Risk and stability

Investors pay more for earnings they consider reliable. A business with steady demand and little debt usually trades at a higher P/E than one whose profits swing with the economy.

Interest rates

When safe interest rates rise, the return investors demand from stocks tends to rise too, which pushes P/Es down, all else equal. When rates fall, the opposite. This is one reason P/Es across the whole market move over time.

Sector and accounting

Sectors differ in growth, capital needs and cyclicality, so typical P/Es differ too. Banks, energy producers and carmakers often trade at lower P/Es than software or consumer brands. Accounting also matters: a company that spends heavily on R&D expenses it immediately, which lowers reported earnings and raises its P/E compared with a company that capitalizes more of its spending.

Cyclical earnings

For cyclical companies, the P/E can be misleading at both ends. At the top of a cycle, earnings are high and the P/E looks low, just before profits fall. At the bottom, earnings are depressed and the P/E looks high, sometimes just before profits recover.

CAPE: smoothing out the cycle

The cyclical problem is what the cyclically adjusted price-to-earnings ratio tries to fix. John Campbell and Robert Shiller used ten-year average earnings in their research in the late 1980s, building on an idea Benjamin Graham and David Dodd described in the 1930s. Shiller later popularized the measure in his book Irrational Exuberance (first edition 2000). It is also called the Shiller P/E or P/E10.

CAPE = real (inflation-adjusted) price of an index ÷ average real earnings over the previous 10 years

A simplified worked example

Suppose an index’s real earnings per unit over the last ten years were 80, 85, 60, 70, 90, 95, 100, 105, 110 and 105. The average is 900 ÷ 10 = 90. If the index’s real price today is 2,700:

Ordinary P/E on the latest year’s earnings: 2,700 ÷ 105 = 25.7
CAPE: 2,700 ÷ 90 = 30.0

The low earnings year (60) is still pulling the average down, so CAPE is higher than the one-year P/E. In a recession year, when one-year earnings collapse, the opposite happens: the ordinary P/E spikes while CAPE barely moves.

What Shiller’s data shows

Robert Shiller publishes monthly US stock market data going back to 1871, based on the S&P Composite and, in recent decades, the S&P 500. In the version of the dataset on his Yale web page, which runs through September 2023:

Reading	CAPE	Month
Highest	about 44.2	December 1999
Lowest	about 4.8	December 1920
Before the 1929 crash	about 32.6	September 1929
Low after the 2008-2009 crisis	about 13.3	March 2009
Average, January 1881 to September 2023	about 17.4	(median about 16.5)

These figures come from our own calculation on that file; later updates of the data may differ slightly. We are not quoting a current reading here, because CAPE changes every month with prices and earnings.

What CAPE does and doesn’t tell you

Campbell and Shiller’s research, and later work, found that high starting CAPE levels have tended to be followed by lower average returns over the following ten years or so, and low levels by higher returns. That is a statement about long horizons and about the average relationship. It has a wide spread of outcomes around it.

What it does not do is time the market. CAPE was above its long-run average for most of the period from the 1990s onward, and investors who sold every time it looked high would have missed long stretches of gains. Critics also point out that changes in accounting rules, in how much companies pay out versus buy back, and in sector mix can make today’s CAPE hard to compare with readings from a century ago. Some researchers, including Shiller himself in later work, have proposed adjustments, such as comparing earnings yields with bond yields.

What can go wrong

Negative or tiny earnings

If a company lost money, its P/E is meaningless (or reported as “N/A”). If it barely made money, the P/E can be in the hundreds, which says more about the denominator than about valuation. For these companies, analysts often look at price-to-sales or cash flow measures instead.

Adjusted versus GAAP earnings

The “E” in a quoted P/E might be GAAP net income or a company’s adjusted figure that removes certain costs. The adjusted number is usually higher, which makes the P/E look lower. Check which one is being used; our guide to how earnings reports work explains the difference.

One-time items

A big one-time gain inflates earnings and shrinks the P/E for a year. A big one-time charge does the opposite. Look at the income statement for anything unusual before relying on a single year.

Comparing across countries and sectors

A market dominated by banks and resource companies, as the S&P/TSX Composite has historically been, will usually show a lower overall P/E than one weighted towards technology. That difference reflects composition, not necessarily cheapness.

Treating “low P/E” as “cheap”

A low P/E can mean a bargain, or it can mean the market expects earnings to fall. Some of the companies with the lowest P/Es are low for good reasons. P/E is a starting question for valuation, not a conclusion. For the other tools, see how to value a stock.

The short version

P/E tells you how much you are paying for each dollar of current or expected earnings. Trailing P/E uses past earnings, forward P/E uses forecasts, and CAPE uses a ten-year inflation-adjusted average to smooth out the business cycle. P/Es differ for real reasons: growth, risk, interest rates, sector, and accounting. A high or low number is the beginning of a question about why, not an answer about whether to buy or sell.