The EPS formula: basic and diluted
EPS = (net income − preferred dividends) ÷ weighted average shares outstanding
Northfield Tools, a fictional industrial company, earned $150 million of net income last year with 50 million shares outstanding on average, so its basic EPS is 150 ÷ 50 = $3.00. It also has options and restricted stock that could become 2 million more shares, so its diluted EPS, the figure to use, is 150 ÷ 52 = $2.88. Diluted EPS is always the lower number and the more honest one, because the extra shares are a real claim on future profits even before they are issued. Companies report both on the face of the income statement.
"Weighted average" matters when the share count changes during the year: a buyback in the last month barely moves this year's EPS but lifts next year's. For a quarter, companies report EPS for the three months and the year to date; compare the quarter with the same quarter a year earlier, since most businesses are seasonal.
Why the level of EPS means little on its own
A company can double its EPS overnight by halving its share count through a reverse split, and halve it by splitting the stock two for one, without a dollar of profit changing. So $3 of EPS is neither good nor bad. It becomes meaningful only against something:
- Against the price. EPS ÷ price is the earnings yield; price ÷ EPS is the P/E. Northfield's $2.88 against a $60 share is a 4.8% earnings yield and a P/E of 20.8. The good P/E ratio article covers what that multiple should be.
- Against last year. EPS growth, and its consistency, is what the market rewards or punishes.
- Against expectations. On results day the question is not whether EPS was high but whether it beat or missed what analysts expected, which the earnings surprise article explains.
- Against cash. EPS is an accounting figure; whether operating cash flow keeps pace with net income decides whether it is real.
What good EPS growth looks like: a worked example
Suppose Northfield's EPS goes from $3.00 to $3.57 the following year, growth of 19%. That is a good number, but the source of the growth matters as much as its size. Break it into its parts:
| Driver | Change | Effect on EPS |
|---|---|---|
| Revenue | +9% | ×1.09 |
| Net margin | 10.0% → 10.6% | ×1.06 |
| Net income | +15.5% | ×1.155 |
| Share count | −3% (buybacks) | ÷0.97 |
| EPS | +19.1% | 1.155 ÷ 0.97 = 1.191 |
About nine points of the growth came from selling more, six from earning more on each sale, and about four from spreading the profit over fewer shares. Each has a different quality. Revenue growth is the most durable if it reflects demand. Margin expansion is welcome but has a ceiling, and it reverses when input costs rise. Buyback-driven growth is real for continuing shareholders, but it costs cash, and if the company borrowed to fund it the EPS growth came with added risk. A company whose EPS grows 19% on flat revenue, through buybacks and cost cuts alone, is a different investment from one whose EPS grows 19% on 15% revenue growth, even though the headline is identical.
Consistency counts too. Five years of 10–15% growth says more about a business than one year of 40% followed by a decline, and the earnings analysis guide shows how to read the quarterly pattern.
GAAP vs adjusted EPS
Most companies report two EPS figures: the GAAP number required by accounting rules, and an "adjusted" or "non-GAAP" number that excludes items management considers unrepresentative, typically stock-based compensation, amortisation of acquired intangibles, restructuring charges and legal settlements. Analysts' estimates and the headline "beat" usually refer to the adjusted figure. The adjustments can be reasonable, but they are chosen by the company, and the gap between the two numbers is worth watching. Stock-based compensation in particular is a real cost: shares given to employees dilute everyone else, and excluding it makes a company that pays its staff in stock look more profitable than one that pays in cash. Every release includes a reconciliation table from GAAP to adjusted; read it once a year at least.
EPS and the price: the growth that is already expected
Here is the part most discussions of a good EPS leave out. A share price is not a reward for past earnings; it is a forecast of future ones. If Northfield trades at a P/E of 20.8 while the market as a whole trades at 16, the premium is the market's bet that Northfield's EPS will grow faster than average for years. Deliver that growth and the price holds; deliver merely good growth and the price can fall, because good was already priced in. The forward vs trailing P/E article shows how the gap between the two multiples encodes next year's expected growth, and the implied growth rate article extends the idea over a decade.
So a good EPS, for an investor, is one whose growth exceeds what the price requires, produced in a way that can continue. VALYSTR's The Bar works out the earnings a share price needs a company to deliver and sets them against the consensus, which is the comparison the question really calls for.
EPS traps
- One-off gains. A disposal, a tax settlement or a revaluation can lift a year's EPS and vanish the next. The notes disclose them.
- A lower tax rate. EPS growth that comes from the tax line rather than the operating line is not repeatable once the rate normalises.
- Borrowed buybacks. Rising EPS alongside rising debt and a flat operating profit is leverage, not growth.
- Dilution creep. If the diluted share count rises every year, stock-based pay is quietly taking the growth; compare total net income growth with EPS growth.
- Negative EPS. A loss-making company has no meaningful P/E; look at revenue growth, gross margin and cash burn instead.
- Earnings that cash does not confirm. When operating cash flow trails net income for several quarters, read the 10-Q for the working-capital build behind it.
See the earnings a price requires
The Bar translates a share price into the earnings and growth a company must deliver, and sets them against what analysts expect.
Frequently asked questions
What is a good EPS for a stock?
There is no good level of EPS on its own, because the figure depends on how many shares a company has. What matters is EPS relative to the price (the earnings yield or P/E), its growth rate and consistency, whether the growth comes from revenue, margins or buybacks, and whether operating cash flow confirms it. A good EPS is one that grows faster than the price already assumes.
Is a higher EPS always better?
No. A higher EPS can simply mean fewer shares, a one-off gain or a lower tax rate. Two companies with identical profits and different share counts have different EPS and identical value. Judge EPS by its growth, its quality and its relationship to the share price rather than by its size.
What is a good EPS growth rate?
Sustained growth of about 10–15% a year is strong for an established company, and faster is possible for younger ones, but the rate the market requires depends on the valuation: a stock on a high P/E needs high growth just to hold its price. Growth that comes from rising revenue is more durable than growth from margin gains or buybacks alone.
What is the difference between basic and diluted EPS?
Basic EPS divides net income by the shares actually outstanding. Diluted EPS also counts shares that could be created from options, restricted stock and convertible securities, so it is lower and more conservative. Use diluted EPS for valuation and growth comparisons, because those potential shares are real claims on future profits.
Why does a stock fall when EPS grows?
Because the price already assumed growth, often more than was delivered. A share trading at a premium multiple is a forecast of fast EPS growth; if growth is merely good, or guidance for the next period is cautious, the forecast is revised down and the price falls even though the reported number rose. Compare the growth with what the valuation requires, not with zero.