EPS (Earnings Per Share) Explained

Picture this. You open Moneycontrol on Monday morning, see a stock's EPS jump 40 per cent over last quarter, watch the P/E ratio drop, and decide the stock looks cheap. A month later you find out the jump came entirely from a one-time land sale; the operating business was flat. That moment of feeling tricked by a number that looked like good news is what this article is here to prevent.

Open Moneycontrol, Screener or Tickertape, look up Infosys or HDFC Bank, and the EPS number sits right next to the share price. A higher EPS instinctively feels like a stronger company. A growing EPS feels like a winning one.

That instinct is half right. EPS is the foundation of almost every other valuation number you will ever use, starting with the P/E ratio — the share price divided by EPS, a quick measure of how many years of current earnings you are paying for a share. But the figure on the screen is the headline, not the story.

Two companies with the same EPS can be very different businesses. The same EPS for the same company can mean two different things in two different quarters. Reading EPS well is mostly about knowing what the number leaves out.

What EPS actually measures

The formula has two parts. On top sits the profit left for ordinary shareholders after tax — the technical phrase is "net profit attributable to equity shareholders," meaning profit after subtracting any fixed payouts owed to preference shareholders. On the bottom sits the average number of equity shares the company had during the period.

Take Infosys as a worked example. If the company reports a net profit of around ₹26,000 crore for the year and has roughly 414 crore equity shares, the EPS works out to about ₹63 per share.

That number is the profit the business made for every share you own. If you hold one share of Infosys, the company effectively earned ₹63 on your behalf for that year. Whether you ever see that ₹63 in cash is a different question; it might be paid as a dividend, ploughed back into the business, or used to buy back shares.

This is the most important thing to internalise. EPS is an accounting figure, not a cash figure. It reflects profit as reported under the accounting rules, with all the judgement calls those rules allow. The cash actually generated in the same year can be quite different.

The denominator is a "weighted average" because share counts change inside a single year. Picture a company that starts the year with 10 crore shares and issues 2 crore fresh shares on 1 October. The new shares only existed for three months of the year, so the average for EPS is 10 + (2 × 3/12), which equals 10.5 crore. EPS for that year is calculated on 10.5 crore, not on the full 12 crore.

EPS is also a per-share number, which means it can move for two completely different reasons. Either the profit changed, or the share count changed. A buyback shrinks the share count and lifts EPS even if profit is flat. A fresh equity issue, or an ESOP — employee stock option plan, shares granted to employees that they may exercise later — expands the share count and dilutes EPS even if profit is rising.

Indian companies are required under Ind AS 33 — the Indian Accounting Standard for EPS, the local version of the international IAS 33 — to report both basic and diluted EPS on the face of the P&L (profit and loss statement). Both numbers show up in the quarterly results uploaded to BSE and NSE on results day, as required under SEBI's Listing Regulation 33, which sets the standard format for every listed company's financial results.

Basic EPS versus diluted EPS

Every Indian listed company reports two EPS numbers. Basic EPS uses only the shares that actually exist today. Diluted EPS also adds the shares that would come into existence if every outstanding ESOP, convertible bond, warrant and similar instrument were exercised.

For profitable companies that carry dilutive instruments, diluted EPS is usually equal to or lower than basic EPS. The gap between the two tells you how much future dilution the existing share base is sitting on. (The accounting rules leave out anti-dilutive items, so for loss-making companies the relationship is not always one-way — a useful nuance, not a daily worry.)

For old, simple companies like ITC or Hindustan Unilever, the two numbers are nearly identical because there are very few convertible instruments outstanding. The gap is so small it does not change any decision.

For new-age listings like Zomato, Paytm, Nykaa and Policybazaar, the gap can be meaningful. These businesses ran for years on heavy ESOP grants to attract talent, and a chunk of those options will eventually vest into real shares. Diluted EPS is the more honest figure for valuing them. (You can confirm the exact ESOP pool size in any of these annual reports under the "share-based payments" note.)

The same applies to companies that have raised money through FCCBs — foreign currency convertible bonds, a loan in dollars that can later be turned into shares — or compulsorily convertible preference shares. Those instruments are not equity today but will be tomorrow, and serious analysts work with the diluted number from the start.

One more wrinkle is the weighted average we met earlier. If a company issues fresh shares mid-year, EPS is not calculated on the new total; it uses the time-weighted average over the period.

Take a QIP — a Qualified Institutional Placement, the standard route a listed company uses to raise fresh equity from big institutional investors — done on 1 October that doubles the share count. Because the new shares only existed for three months of that financial year, the denominator only goes up by three months' worth of new shares. The full impact on EPS only lands the following year, when those shares are outstanding for all twelve months.

EPS growth bands across Indian sectors

The first thing every Indian investor needs to internalise is that the absolute EPS number on its own means nothing. A ₹500 EPS on a tiny share base is no better than a ₹20 EPS on a large one. What matters is the trend in EPS over time, and whether that trend is sustainable.

Different kinds of businesses grow EPS at very different rates, and the rate itself carries information about quality and risk. A consumer staples giant — an FMCG (Fast-Moving Consumer Goods) company like Nestle India — will not double its EPS in three years, and you should be suspicious if it ever claims to. A small-cap chemicals company in an upcycle might.

The right anchor is the long-run growth rate across a full cycle, not the latest quarter. Quarterly EPS bounces around with seasonality, commodity prices, tax rate changes and one-off items. The trailing five-year and ten-year EPS CAGR — the compounded annual growth rate, i.e. the average yearly growth over the whole period — strips most of that noise out.

Compare a stock to its peer group, not to an unrelated one. A 14 per cent EPS CAGR on HDFC Bank is healthy. The same 14 per cent on a mid-cap specialty chemicals business that historically grew at 25 per cent is a slowdown.

As a long-cycle rule of thumb, the Nifty 50's underlying earnings have compounded at low double digits over multi-year periods — not in every single year, and not to be confused with the index's price return. Treat it as a sanity check, not a target. A stock claiming to grow EPS at three times this rough rate had better have a very specific reason behind it.

Use the sector band as your starting point. A stock growing well above its band deserves a hard look at whether the growth is real or borrowed from the future. A stock growing well below its band is either entering a tough patch or being mis-classified by the market.

When EPS lies to you

The most expensive single mistake retail investors make with EPS is treating the reported number as a clean signal of business health. Higher EPS means better quarter. Lower EPS means worse quarter. Both halves are wrong often enough to lose serious money.

EPS can rise even when the underlying business is shrinking. It can fall even when the business is thriving. The accounting rules give management enough room to move the number in either direction without breaking any law.

One-off items are the most common distortion. A company that sells a piece of land, wins a tax refund, settles a legal dispute or revalues an asset can show a big jump in net profit for a single quarter that has nothing to do with the operating business. The headline EPS spikes, the P/E looks suddenly cheap, and the stock rallies on what is essentially a non-recurring number.

Whole industries make this trap concrete. Commodity-linked businesses — metals, oil and gas, refiners — show big EPS swings whenever a single asset is sold, a deferred-tax adjustment lands, or a global commodity price moves. Reading EPS for an ONGC or a Vedanta-style commodity company without separating recurring operating profit from one-off gains is the fastest way to misjudge the business; the company's own MD&A (the management discussion section of the annual report) is the right place to verify the exact split.

Share buybacks are the other big driver. When TCS or Wipro buys back shares, the share count drops and EPS rises mechanically even if profit is flat. That is fine if you understand the source. It becomes misleading when investors treat the EPS jump as evidence of operating strength.

The opposite trap is just as costly. A growing company that issues fresh equity to fund expansion dilutes its share base and can show flat or falling EPS for a year or two, even as the business itself is getting bigger. The next two years usually show the catch-up as the new capital starts producing.

Neither rule is universal. Some big EPS jumps are real and repeatable, and some EPS dilution is wasteful capital raising that never earns its keep. The printed EPS alone never tells you which case you are looking at — the number is the starting question, not the answer.

The right question is not "did EPS go up" but "did operating EPS grow in line with revenue, on a stable share base, with cash flow to back it up." That second question is what fundamental analysis actually trains you to ask.

How to actually use EPS

Once the framework is clear, the practical use of EPS becomes simple. It is a building block, not a verdict. Walk through these five checks every time you look at an EPS number.

1. 1
   Pair EPS with the share price. Compute the P/E ratio (price ÷ EPS). A ₹63 EPS at a ₹1,500 share price is a P/E of about 24. EPS without price is half a sentence — the ratio is what makes a large-cap IT company comparable to a small-cap chemical one.
2. 2
   Look at five-to-ten-year EPS history. One quarter tells you almost nothing; a decade tells you whether the business compounds. Most data sites plot the long-run line. A line that climbs steadily through demonetisation, Covid and the 2022 commodity shock is making a real claim.
3. 3
   Compare diluted EPS against basic EPS. The gap between the two is the dilution risk you are taking on. For a clean old company they will look identical; for a heavy-ESOP business they will not. Use the diluted figure for any valuation work.
4. 4
   Strip out exceptional items. Indian quarterly results separate exceptional items on the face of the P&L. Read that line; if a chunk of the profit came from a land sale or a tax refund, knock it out before calculating "real" EPS.
5. 5
   Cross-check against cash flow. Real profit eventually shows up as real cash. If EPS is rising for years but operating cash flow is flat or falling, the EPS line is being held up by accounting choices, not by money in the bank.

A practical Saturday-morning workflow: pick five well-known stocks from sectors you understand. Note the trailing EPS, the five-year EPS CAGR and the share count today versus five years ago for each. Then note the latest quarterly EPS with and without exceptional items, and the diluted EPS versus the basic EPS. The pattern across those numbers tells you more about each business than any single headline ever can.

Repeat that exercise once a quarter and EPS stops being a number you skim. It becomes the spine of your read on the business, and the foundation under every other ratio you ever calculate from it.