Earnings Per Share (EPS)

Earnings per share (EPS) is the part of a company's profit that belongs to each ordinary share in issue during a reporting period. It is worked out by dividing the profit available to common shareholders by the weighted average number of those shares outstanding, and it is one of the most quoted figures in an annual report and in quarterly earnings releases. It is also the bottom half of the price-to-earnings ratio.

Listed companies present two versions. Basic EPS uses the shares actually outstanding; diluted EPS also counts the shares that could appear if options, warrants and convertible securities were turned into stock. The rules come from IAS 33 Earnings per Share under IFRS and from ASC Topic 260 under US GAAP, two standards that have been closely aligned since the late 1990s. EPS measures accounting profit, not cash, and it moves with share-count decisions such as buybacks, new issues and stock splits as well as with the business itself.

Think of a company's yearly profit as a large pizza and each share as a ticket that entitles the holder to one slice. EPS tells you how big one slice is. If the pizza grows, every slice gets bigger. If the company prints more tickets while the pizza stays the same, every slice gets smaller, even though nothing about the kitchen has changed.

Diluted EPS asks a careful question: what if everyone holding a promise to receive a ticket later, such as an employee with share options, cashed it in today? The slice would shrink a little. Reporting both figures lets a reader see the slice size now and the slice size if all those promises were honoured.

The standard calculation is:

• Net income is the profit for the period after tax, as shown in the income statement.
• Preferred dividends are subtracted because that slice of profit belongs to holders of preferred stock, not to ordinary shareholders.
• The weighted average share count adjusts for shares issued or bought back part-way through the period, according to how long each share was outstanding.

Worked example: a company earns net income of $120 million and pays $20 million of preferred dividends. It had 48 million shares for the first half of the year and issued 4 million more at mid-year, so the weighted average is $48 + 4 \times 0.5 = 50$ million shares. Basic EPS is $(120 - 20) \div 50 = 2.00$, or $2.00 a share.

Share splits and bonus issues bring in no money, so the share count for every period shown is restated as if the split had always applied.

Diluted EPS adds in every security that could become an ordinary share and would lower EPS if it did. Both standards prescribe the same main methods:

• The treasury stock method for options and warrants. It assumes holders exercise them and the company uses the exercise money to buy back shares at the average market price; only the net extra shares are added. With 5 million options at a $20 strike and an average price of $25, the $100 million of proceeds buys back 4 million shares, so 1 million shares are added.
• The if-converted method for convertible instruments. A convertible bond is treated as converted at the start of the period: the shares it would create are added to the denominator and the after-tax interest saved is added back to earnings.
• Contingently issuable shares, such as performance share awards, are included once the conditions for issuing them have been met.

Continuing the example, adding the 1 million option shares gives diluted EPS of $100 \div 51 \approx 1.96$, or about $1.96. Securities that would raise EPS (anti-dilutive ones) are left out, and when a company makes a loss, potential shares are ignored because they would make the loss per share look smaller.

ASC 260 requires public US companies to show basic and diluted EPS on the face of the income statement, and IAS 33 applies the same requirement to entities whose ordinary shares are publicly traded or are being registered for trading. Both figures fall within the audited financial statements.

Many companies also publish "adjusted" or "core" EPS that strips out items such as restructuring charges, impairments or acquisition costs. In the United States these are non-GAAP measures: the Securities and Exchange Commission requires them to be reconciled to the GAAP figure under Regulation G and, in filings, Item 10(e) of Regulation S-K, which also bars giving them more prominence than the GAAP number. Analysts' consensus forecasts usually refer to adjusted EPS, which is why a company can "beat estimates" while its reported EPS falls.

EPS feeds directly into share-based measures used in valuation: the P/E ratio, the dividend payout ratio (a company's dividend per share divided by EPS) and the earnings yield. EPS growth is a common target in executive pay plans and the main focus for buyers of a growth stock.

The figure has well-known weaknesses:

• Share buybacks raise EPS by shrinking the denominator, even when total profit is flat. Critics see buybacks aimed mainly at lifting EPS as a sign of short-termism, favouring near-term results over investment that would build long-term value.
• Accounting choices on depreciation, provisions and revenue timing change net income without changing cash flow.
• One-off gains and losses can swing a single year sharply, which is why analysts also look at measures such as EBITDA.
• EPS cannot be compared between companies in isolation, because the number of shares each has issued is arbitrary; it only makes sense alongside the share price, which is what the P/E ratio and market capitalization provide.

Read with the balance sheet, the price-to-book ratio and the cash-flow statement, EPS remains the standard per-share summary of what a company earned for its shareholders.

• IFRS Foundation: IAS 33 Earnings per Share