Price-to-Earnings Ratio (P/E)

Starter — for Never bought a share

The P/E ratio tells you how many dollars investors are paying for each dollar of yearly profit the company makes.

What it is

A share is a slice of a business. The business earns a profit; your slice of that profit is called earnings per share, or EPS. Divide the share price by that profit and you get the P/E — the price you pay per dollar of profit. A P/E of 25 means the market is paying $25 for every $1 the company earns in a year.

Why anyone cares

The P/E is a rough speedometer for expectations. A low number usually means the market expects little growth or sees risk; a high number means the market expects the profit to grow a lot. Neither is automatically good. A high P/E is only justified if the profit really does grow; a low P/E is only cheap if the profit holds up.

How to use it without being fooled

Only compare P/E ratios between companies that do similar work, because different industries earn profits at different rates. And always ask which year of profit you are dividing by — last year's actual, or this year's forecast. They can give very different answers.

What to take away

• P/E = price per share ÷ profit per share. It is a price tag on profit.
• A high P/E is a promise about growth; a low P/E is a question about risk.
• Compare it only against similar businesses, and check which year of profit it uses.

Medium — for Invested before, reads the news

P/E is the market's price for one dollar of annual earnings; it compresses growth expectations, risk and accounting quality into a single number.

Trailing, forward and everything between

Trailing P/E uses the last twelve months of reported earnings — real, audited, but backward-looking. Forward P/E uses consensus estimates for the next twelve months, which are neither audited nor guaranteed and are usually flattering because companies guide analysts gently downward. The gap between the two is itself information: a trailing P/E of 30 collapsing to a forward P/E of 18 says the market expects a step change in profit, and that expectation is the thing you are buying.

The arithmetic, and the two numbers that decide it

P/E = share price ÷ diluted EPS. "Diluted" matters: it assumes employee options and convertible securities are exercised, so it is the honest share count. Comparing a basic-EPS P/E with a diluted-EPS P/E is a silent apples-to-oranges error. Use the same share count the company uses in its own income statement, and read the footnotes if the count changed materially.

Earnings yield — P/E standing on its head

Invert the ratio and you get E/P, the earnings yield: profit as a percentage of price. A P/E of 20 is a 5% earnings yield. That is the number to compare with a bond yield or a savings rate, because it is expressed in the same units. It is also the honest way to describe an expensive market: a 40× P/E is a 2.5% earnings yield, which is a statement you can argue with.

What the ratio does not see

P/E says nothing about debt. Two companies can trade at 20× with one carrying no borrowings and the other carrying five times its profit in debt; the equity of the levered one is riskier, and the ratio will not tell you. It also says nothing about capital intensity: a business that must reinvest half its profit to stand still deserves a lower multiple than one that does not.

What to take away

• Label every P/E as trailing or forward; the two can differ by a third.
• Earnings yield (E/P) is the same fact expressed as a return, so it can be compared with rates.
• P/E is blind to debt and to capital intensity — check both before trusting the comparison.
• A negative P/E is not cheap. It is undefined.

High — for Works with this number already

P/E is really a price-to-accounting-earnings ratio: its denominator is a management-and-standard-setters construct, so the work is deciding how much of it is cash and how much is accrual.

The denominator is an opinion

Net income is the output of accrual accounting: revenue recognized when earned rather than when collected, expenses matched to the periods they relate to, estimates for bad debts, warranties, impairments and useful lives. Two firms with identical cash flows can report materially different earnings. That is not fraud — it is the standard — but it means a P/E comparison is partly a comparison of accounting policy. Normalizing to cash (free cash flow, or EBIT) is what removes the policy difference.

Trailing vs forward, and the anchoring problem

Forward multiples embed analyst estimates, and estimates are sticky: they anchor on management guidance, get revised slowly, and cluster around round growth rates. The practical consequence is that the cheapest-looking forward P/Es are frequently the ones with the largest downward revision risk, and the most expensive are sometimes mid-upgrade. If you use a forward P/E, use it as a claim about revision direction, not as a settled fact.

Cyclicals, peaks and the trough illusion

For a cyclical — semis, chemicals, autos, homebuilders, banks — the P/E inverts its own meaning. At the top of the cycle earnings are peak and the P/E looks low; at the bottom earnings are depressed and the P/E looks enormous or goes negative. The peak-cycle P/E is the trap: it is giving you a cheap signal exactly when the denominator is about to fall. The professional approach is to normalize earnings across a full cycle (or use an average of several years, or a margin-based normalized EPS) before applying a multiple.

The adjusted-EPS gap

Most companies publish "adjusted" or "non-GAAP" EPS that excludes amortization, stock-based compensation, restructuring, impairments and "one-time" items that recur annually. The GAAP–adjusted gap has widened over decades. Adjusted EPS is useful when the excluded items genuinely do not affect value (a genuine one-off) and dangerous when they are a chronic cost (stock-based compensation is a real expense paid in dilution). Track the gap over time: a company whose adjustments grow every year is telling you something the headline number is hiding.

Where the ratio is the wrong tool

P/E fails for loss-makers, for pre-profit growth companies, for heavily levered or distressed businesses (because the E captures interest and the equity value is an option on the enterprise), and for anything where capital structure is the main variable. Those cases want EV/EBITDA, EV/sales, price-to-book, or a discounted cash flow — not a bigger P/E. A useful test before quoting any multiple: write down what the denominator would have to be for the number to be justified, then decide whether that figure is reachable. If the answer is “earnings would have to triple and stay there”, the multiple is not a valuation, it is a growth assumption wearing a ratio's clothes.

What to take away

• Net income is an accrual construct; P/E differences are partly accounting-policy differences.
• A low P/E on a cyclical at peak earnings is a warning, not a bargain — normalize the denominator.
• Watch the GAAP-to-adjusted gap over time; growth in it is the signal.
• P/E is invalid for loss-makers and misleading when leverage or capital intensity dominates — switch to an enterprise multiple.
• Earnings yield is the comparable-units form; use it whenever you compare a multiple against a rate.