The P/E Ratio, Explained: What It Tells You and What It Hides

The price-to-earnings ratio is a question, not an answer. That's the whole article in one line, and if you take nothing else away, take that. Most people treat the P/E like a price tag, a number that tells you whether a stock is cheap or expensive. It isn't. It's the market's opinion about the future, divided down into a single figure, and a single figure can hide an enormous amount.

Here's the mechanical part first, because it's genuinely simple. P/E is the price of one share divided by the earnings that share produced over the last year. Same thing at the company level: take the entire market cap and divide it by net income.^[1]A P/E of 20 means you're paying twenty dollars for every dollar of annual earnings. That's it. The math is the easy part. The hard part is that the number lies to you constantly if you read it at face value.

The formula, and why it feels like cheapness

Let's ground it. Two ways to compute the same ratio, and a worked example so the number stops being abstract.

P/E ratio = Price per share / Earnings per share (EPS)

  equivalently:

P/E ratio = Market cap / Net income (annual)

--- Worked example ---

Stock price        = $100 per share
Net income         = $5 billion
Shares outstanding = 1 billion

EPS  = $5,000,000,000 / 1,000,000,000 = $5.00
P/E  = $100 / $5.00                   = 20x

Read it out loud:
  "I'm paying $100 for a slice of the company
   that earned me $5 last year. Twenty dollars
   of price for one dollar of earnings."

Flip the ratio upside down and you get the earnings yield, which is EPS divided by price. A P/E of 20 is an earnings yield of 5 percent. That framing is useful because it lets you compare a stock to a bond. If a 10-year Treasury pays 4.5 percent and a stock yields 5 percent on earnings, you're being paid almost nothing extra to take on a lot more risk. The earnings yield is the same number wearing a different hat, and it's a better hat for a lot of decisions.

So why does the P/E feellike a measure of cheapness? Because the math invites it. A low number looks like a bargain and a high number looks expensive, the same way a cheaper price per pound looks like the better deal at the grocery store. The problem is that earnings aren't pounds of ground beef. They're an accounting output that can be distorted, they describe the past, and they're about to change. The denominator is doing all the work, and the denominator is slippery.

What the number hides

Here's where I get opinionated. The P/E is not wrong so much as it's incomplete in ways that bite.^[7] Four things it hides.

One: earnings can be massaged or distorted by one-time stuff. Net income is the bottom line after a long chain of accounting choices, and those choices matter. A company takes a giant one-time write-off, earnings crater for a quarter, and the trailing P/E spikes to something absurd, not because the business got expensive but because the denominator got temporarily wrecked. The reverse happens too. A company sells a building, books a one-time gain, earnings jump, and the P/E looks deceptively low. Buybacks add another wrinkle. When a company buys back its own shares, the share count shrinks, EPS rises even if total net income didn't move, and the P/E drops without the business getting any better. The ratio moved. The company didn't.

Two: trailing versus forward, and forward is a guess. The trailing P/E uses the last twelve months of actual reported earnings. The forward P/E uses analysts' estimates for the next twelve months.^[2]That difference is bigger than it sounds. Trailing is real but stale. Forward is timely but it's a forecast, and forecasts are reliably optimistic. Analysts as a group tend to start the year too high and walk estimates down as reality arrives. So a stock with a comfortable-looking forward P/E might just have hopeful estimates baked in. When you see a number quoted with no label, assume nothing. Ask which earnings.

Three: a low P/E can be a trap, a high P/E can be a bargain. This is the one that costs people the most money. A low P/E often isn't the market being dumb. It's the market pricing in a decline. Cyclical businesses are the classic example. A homebuilder or a car company at the top of its cycle, with record earnings, can show a single-digit P/E that looks like a screaming buy. It's the opposite. The market knows those earnings are a peak and is refusing to pay up for a number it expects to fall. Buy it and you watch earnings collapse, the denominator shrinks, and the P/E you thought was 7 was actually 20 on normalized earnings. That's a value trap.

The mirror image is the high P/E that's actually reasonable. The price in the numerator doesn't move on last year's earnings. It moves on expected futureearnings. A company growing earnings 40 percent a year deserves a higher multiple than one growing 3 percent, because today's earnings are a small fraction of what next year's will be. A P/E of 40 on a fast grower can be cheaper, in any sense that matters, than a P/E of 12 on a business in slow decline. The multiple is a statement about growth, and you can't read it without knowing the growth.

Four: it's useless across industries and broken for unprofitable companies. A bank, a software company, a utility, and a biotech do not share a natural P/E. Different capital intensity, different growth, different risk, different accounting. Comparing a software company's 35x to a utility's 18x and concluding the utility is cheaper is a category error. And if a company has no earnings, which describes a lot of young growth companies, the P/E is negative or simply not a number. You can't divide a meaningful price by a loss. For those names the ratio doesn't exist, and screening on it throws them out entirely.

Two stocks, same P/E, opposite stories

Here's the worked example that makes the whole thing click. Two companies, both trading at a P/E of 20. Identical multiple. Completely different things to own.

                          GrowthCo        SteadyCo
Price per share           $100            $100
EPS (this year)           $5.00           $5.00
P/E ratio                 20x             20x

Earnings growth rate      +25% / yr       -3% / yr
EPS in 5 years (est.)     ~$15.26         ~$4.29
PEG ratio (P/E / growth)  0.8             negative

The P/E says these are priced the same.
They are not the same investment.

Run GrowthCo forward. If it really compounds earnings at 25 percent, EPS goes from $5 to about $15 in five years. Hold the price still and the P/E quietly falls to roughly 6.5x on those future earnings. You paid 20x for something that becomes a 6.5x business if the growth shows up. SteadyCo goes the other way. Earnings drift down, and that 20x starts looking like a generous price for a business that's slowly getting smaller. Same number on the screen. Opposite bet.

The cousins that fix some of this

The P/E's weaknesses are well known, so people built better tools next to it. None of them replace it. They patch specific holes.

PEG: the P/E adjusted for growth. The PEG ratio is the P/E divided by the earnings growth rate.^[3] A P/E of 20 on a company growing 25 percent gives a PEG of 0.8. A P/E of 20 on a company growing 5 percent gives a PEG of 4.0. The rough rule of thumb, popularized by Peter Lynch, is that a PEG around 1 is fair and below 1 is potentially cheap.^[4] It directly addresses the biggest P/E blind spot, which is that the ratio ignores growth entirely. The catch is that it leans on a growth estimate, and we already covered how trustworthy those are. Garbage growth assumption in, garbage PEG out.

Shiller CAPE: the P/E for the whole market, smoothed. The cyclically adjusted price-to-earnings ratio, or CAPE, was built by Robert Shiller to fix the single-year-earnings problem at the index level.^[5]Instead of one year of earnings, it uses ten years of inflation-adjusted earnings. That smooths out the boom-and-bust in the denominator, so you can't get fooled by a single peak or trough year. People mostly use CAPE to gauge whether the whole market is expensive relative to history, and it's been a decent long-horizon signal even though it's useless for timing anything.

EV/EBITDA: comparing across capital structures. The P/E ignores debt. A company funded mostly with borrowed money and one funded with equity can show similar P/Es while being very different in risk. EV/EBITDA fixes that by using enterprise value, which is market cap plus debt minus cash, over earnings before interest, taxes, depreciation, and amortization.^[6]It lets you compare two companies as if their balance sheets were normalized, which is why acquirers and private-equity folks reach for it. It strips out financing choices and accounting depreciation. It also strips out real costs, so it's not a free lunch, just a different lens.

How to actually use it

I don't want to leave you thinking the P/E is useless. It's not. It's a fast, universal first question, and that's genuinely valuable. Used right, it tells you what the market is assuming so you can decide whether you agree.

The right mental model is reverse-engineering. When you see a P/E, ask what the number is telling you the market believes. A 40x multiple is the market saying it expects strong, durable growth. Your job is to decide whether that growth is real. A 7x multiple on a cyclical at peak earnings is the market saying those earnings won't last. Your job is to decide whether it's right. The ratio frames the bet. You take the other side only when you have a reason to think the crowd's assumption is wrong.

A few rules I'd actually follow:

• Never compare P/Es across industries. A software 30x and a utility 16x are not on the same scale. Compare a company to its own history and to its direct peers, nothing else.
• Always know which earnings.Trailing or forward. If it's forward, find out how good that estimate has historically been for this company before you trust it.
• Treat a suspiciously low P/E as a warning, not a discount. The market is usually pricing something. Find out what before you congratulate yourself on the bargain.
• Pair it with growth. A P/E without a growth rate next to it is half a sentence. The PEG, or just the growth number in your head, finishes it.

So the next time someone tells you a stock is cheap because it trades at 12 times earnings, the right response isn't to nod. It's to ask why. Twelve times what earnings, growing at what rate, against what balance sheet, and what does the market know that's keeping the number that low. That's the whole game. The P/E doesn't end the conversation. It starts it.