The P/E ratio, and why ratios matter
The answer
The P/E ratio is the price of 1 share divided by the profit that 1 share earned. It tells you how many years of the company's current profit you are paying for.
That number means nothing on its own. A P/E is a question, not an answer, and the question is always the same: what went into the E?
Why this costs you money
You open a stock screener. You sort by lowest P/E. One company shows a P/E of 6 while the rest of its industry sits near 22. It looks like the market has missed something. You buy it.
What you have usually bought is one of 3 things.
A one-off profit. The company sold a piece of land, or a subsidiary, or won an old tax dispute. That money is real, it sits in the profit line, and it will not happen again. The E was inflated for 1 year only.
A peak profit. The company makes steel, or sugar, or chemicals. Its profit swings with a cycle. At the top of the cycle profit is enormous, so the P/E looks tiny. The lowest P/E in a cyclical business is usually the most dangerous moment to buy it.
A profit the market does not believe. The market expects the profit to fall. Sometimes the market is wrong. Very often it is early.
Look at the same stock over 2 years.
| Year 1 | Year 2 | |
|---|---|---|
| Share price | ₹120 | ₹120 |
| Profit per share | ₹20 (₹14 normal + ₹6 land sale) | ₹14 |
| P/E | 6 | 8.6 |
The price did not move. The company did nothing wrong. The P/E rose 43% because the E fell back to normal. You were never paying 6 times profit. You were paying 8.6 times, and you did not know it.
The reverse mistake costs just as much. Somebody refuses a company at a P/E of 45 because "45 is expensive", without asking whether the E is depressed, clean, or growing 30% a year. The number 45 does not contain that information.
How it works
A company earns ₹100 crore of profit. Is that good? You cannot answer. It is superb for a business worth ₹500 crore and poor for one worth ₹50,000 crore.
A ratio divides out size. That is why ratios exist. Once profit is divided by the price you paid for it, a small company and a large one can be compared directly. Every ratio in this cluster does that same job in a different place.
The P/E is built from 2 numbers.
P = the price of 1 share. E = earnings per share, which is net profit divided by the share count.
Market capitalisation divided by total net profit gives the identical number.
Turn it upside down and you get the earnings yield: 1 divided by the P/E, written as a percentage. A P/E of 20 is an earnings yield of 5%. That direction is often more useful, because 5% can be compared with a bank deposit or a government bond. "20 times" cannot.
Now the part that matters. There is no single E.
Trailing E is the profit actually reported over the last 12 months. It is a fact, and it is history, and history includes events that will not repeat.
Forward E is the profit analysts expect over the next 12 months. It is a forecast, and it is wrong most often at exactly the turning points where it matters.
Two further choices sit inside the E. Consolidated profit includes the subsidiaries; standalone is the parent alone, and for a group with many subsidiaries the 2 differ widely. Diluted earnings per share assumes every share option and convertible turns into a share; basic does not. Diluted is the more honest number, because those shares are coming.
So "the P/E is 18" is a number assembled from at least 4 separate choices, and nobody tells you which ones were made.
What it tells you, and what it does not
A P/E tells you what the market expects. A high P/E means buyers expect profit to grow, or expect the profit to be unusually safe, or both. A low P/E means buyers expect profit to fall, or expect it to be unreliable, or both.
That removes the most common error. A high P/E is not an opinion about whether a stock is expensive. It is a record of what the crowd expects. Your job is to judge whether the expectation is reasonable, not whether the number is big.
A P/E does not tell you whether the market is right. It does not work for a loss-making company, because dividing by a negative number produces nothing meaningful. It becomes unstable when profit is near zero, because a tiny denominator produces an enormous ratio.
It also says nothing about debt. The E is calculated after interest is paid, but the P covers only the equity, not the borrowings. Two companies with the same P/E can carry completely different risk.
And it does not travel across industries. A software company with no factories and a steel company with 3 plants will never carry the same multiple, and neither is mispriced because of it.
The decision rule
Never read a P/E alone. Every P/E gets a second question, and the second question is where the information is.
1. What is in the E? Trailing or forward. Consolidated or standalone. Does it contain an exceptional item, or a large "other income" line that has nothing to do with the main business?
2. Compared to what? Two comparisons are legitimate: the company against its own P/E range over 10 years, and the company against others doing the same thing. Comparing it to the index is not a comparison.
3. What has to be true for this to be correct? Say the assumption out loud. "This price assumes profit roughly triples in 6 years." Now you can decide whether you believe it.
Most low P/E ratios do not survive question 1.
Try this now
Five minutes, your own holdings.
- Open your broker app or portfolio page. Write down the P/E shown for every stock you hold, as a list.
- Take the holding with the lowest P/E. Open its financials tab, or the company's latest results filing on the NSE, BSE or SEC website.
- Answer 3 questions about that number. Is it trailing or forward? Most Indian apps show trailing; most US sites show both, labelled TTM and forward. Is it consolidated or standalone? What was net profit for the last 4 quarters?
- Find 2 lines in the profit and loss statement: "other income" and "exceptional items". Add them. Divide the total by net profit.
- Repeat steps 2 to 4 for the holding with the highest P/E.
What you should see. Three things, and any one is worth the 5 minutes.
You will probably find you did not know which E your app was using. Most people have never checked once.
On at least 1 holding, "other income" and "exceptional items" will be a meaningful share of profit — often 10% or more. That part of the E is not the business. It is interest on cash, or a sale, or a writeback.
And the gap between your lowest and highest P/E is probably explained by industry, not by value. If the low one is a lender or a commodity producer and the high one sells a branded product, you have not found a cheap stock and an expensive stock. You have found 2 different businesses.
Three real cases
1. The Nifty 50, March 2020 and February 2021 (India) — the denominator moved, not the numerator In late March 2020, after the fastest fall in the index's history, the trailing P/E of the Nifty 50 dropped to roughly 17. Within a year it had risen above 40, a level it had never reached before. A casual reader concluded Indian shares had become 2 times more expensive. That is not what happened. Prices recovered — but the reported profits of the previous 4 quarters had collapsed during the lockdown, so the E shrank at the same time. The ratio moved mostly because of its bottom half. Most people watch only the top half.
2. Nokia, 2007 to 2013 (Finland, listed in the United States) — cheap all the way down In the fourth quarter of 2007 Nokia held about 40% of the global mobile phone market and about 51% of the smartphone market. Every year from 2008 onward a value investor could look at Nokia and see a low P/E on trailing profits. Revenue fell from about €51 billion in 2007 to about €12.7 billion in 2013. Market value fell from roughly €110 billion in 2007 to about €6.3 billion by mid-2012, a loss of more than 90%. The P/E was low at every point on the way down, because the E fell faster than the P. A low P/E on a shrinking business is not a discount.
3. One97 Communications (Paytm), November 2021 (India) — the P/E that did not exist Paytm's initial public offering raised about ₹18,300 crore at a valuation near $20 billion. It listed on 18 November 2021 at ₹1,950 on the NSE, below the issue price, and closed the first day down more than 27% at about ₹1,560 — the largest listing-day fall in Indian market history at the time. The company was making losses, so there was no E. There was no P/E to check and no earnings yield to compare with a bond. This is the honest limit of the tool. When there is no profit, the P/E cannot protect you, and something else has to do the work.
The question that resolves it
A novice looks at a P/E and asks: is this number high or low?
An expert looks at the same P/E and asks: what is inside the E, and what is the P assuming?
The first question takes 1 second and is worth almost nothing. The second takes 5 minutes and is most of fundamental analysis.
What would make this wrong
If low P/E ratios reliably identified good investments, buying the cheapest tenth of the market every year would beat the market every year.
It does not. In the United States, low-multiple stocks underperformed high-growth stocks for most of the period between 2009 and 2020 — a stretch long enough to end careers. Any claim that "low P/E wins" has to survive that decade, and it does not survive it cleanly.
The honest limits: the P/E is useless for a loss-making company and unstable for a barely profitable one. It is shaped by accounting choices that differ between countries. It ignores debt entirely. And almost every serious mistake made with this ratio is a comparison mistake, not an arithmetic mistake.
In India
Indian companies report quarterly and must file results with the exchanges under SEBI's listing regulations, usually within 45 days of the quarter end. Those filings are free on the NSE and BSE websites. That is where the E comes from, and where you should check it.
Consolidated versus standalone matters more here. Indian groups often hold their operating businesses inside subsidiaries. A holding company's standalone profit may be almost entirely dividend income received from those subsidiaries. Using the wrong statement does not make the P/E slightly wrong. It makes it wrong by a multiple.
Promoter-controlled companies are the norm. A promoter is a founding shareholder with a controlling stake. Where one family controls the board, related-party transactions, royalty payments and the timing of recognised profit deserve more attention than in a widely held company. The E is more likely to reflect a decision.
Lenders cannot be read this way. For a bank or a non-banking financial company, profit depends on how much has been set aside for loans that may not be repaid. That provision is a management judgement. A lender's P/E prices a judgement, not a fact.
NSE Indices publishes daily P/E, P/B and dividend yield for the Nifty 50 and other indices, free, with history. It is a trailing consolidated figure and one of the most useful free datasets an Indian investor has.
In the United States
US companies file form 10-Q each quarter and form 10-K each year with the Securities and Exchange Commission. Those carry audited numbers under US Generally Accepted Accounting Principles, called GAAP.
Then there is a second set of numbers. Adjusted, or non-GAAP, earnings. Almost every large US company publishes an earnings figure that excludes items management considers unrepresentative. The most common exclusion is stock-based compensation — shares issued to employees as pay. That is a genuine cost to shareholders, because it creates more shares. Adjusted earnings always produce a lower and more flattering P/E than GAAP earnings.
Buybacks. US companies return cash mainly by buying their own shares and cancelling them. The share count falls, so earnings per share rises even when total profit does not. A company with flat profit and a steady buyback shows growing EPS every year while the business stands still.
Consensus estimates are published and widely quoted, so the forward P/E is easy to find and compare. That is a genuine advantage of the US market.
The cyclically adjusted P/E, or CAPE, developed by Robert Shiller, divides the price of the S&P 500 by average inflation-adjusted earnings of the last 10 years. It exists precisely because a 1-year E is unstable. It is a useful long-horizon measure and a poor timing tool.
Where they differ, and what that tells you
Indian shares have traded at persistently higher multiples than most other emerging markets for many years. This is not a temporary distortion and it deserves honest treatment. Part of it is explained by higher expected growth, a large domestic base buying every month through systematic investment plans, and an index dominated by high-return consumer and services businesses rather than state-owned commodity producers. Part of it may simply be that Indian shares are expensive. Nobody can separate the 2 parts with confidence.
What follows is practical: compare an Indian P/E with Indian history and Indian peers, not with a global average. Importing an American rule about what multiple is normal will make you sell good businesses too early for a decade.
American earnings per share is flattered in a way Indian EPS historically was not. Large US companies buy back shares continuously, lifting EPS mechanically. Indian companies have used buybacks far less. For most of the last decade an Indian buyback also carried a distribution tax paid by the company under section 115QA, which made it less attractive than in the United States; from 1 October 2024 buyback proceeds became taxable in the shareholder's hands as a deemed dividend instead.
That gives you a checking habit. For a US company, before believing EPS growth, look at the share count over 5 years. If the count fell 20%, a fifth of the EPS growth was arithmetic. For an Indian company the count is usually stable or slowly rising, so EPS growth is more likely to be real — but check for preferential allotments and convertibles, which move the count the other way.
Carry this
- A ratio divides out size, so 2 companies of different sizes can be compared.
- A P/E is a question, not an answer. The question is: what is in the E?
- Check 3 things before using one: trailing or forward, consolidated or standalone, and how much of the profit was a one-off.