Earnings per share (EPS) is the net profit or loss accruing to equity-holders per outstanding share. It is a popular measure of the performance of a company and a factor in the valuation of its shares. Companies disclose their EPS in their financial statements.

Both the components that enter into the calculation of EPS -- earnings and the number of outstanding shares -- require careful definition.

'Earnings' for this purpose is the amount over which equity-holders have a claim. Therefore, it is calculated by deducting preference dividends and taxes attributable to them from the net profit and loss (including extraordinary items) for the reporting period.

As for the number of shares, it may happen that the number of shares outstanding changes through buy-backs or issues during the period for which earnings is reported.

In that case, EPS is calculated by taking a weighted average of the shares outstanding, with the weight for each share being the proportion of the reporting period for which it has been outstanding.

Bonus issues, stock-splits and reverse stock-splits (consolidation of shares) change the number of outstanding shares without changing the resources available to the firm. Therefore, companies adjust the number of equity shares outstanding for those periods for bonus issues, stock-splits and reverse stock-splits while calculating the EPS.

Let us take an example. Suppose that for a firm, the net profit attributable to equity share holders for 2005 was Rs 10,00,000 and the number of outstanding shares 50,000. The EPS for 2005 would then be Rs 20 in the financial statements of 2005.

Suppose that in 2006, the net profit attributable to equity share holders was Rs. 15,00,000 and the number of shares outstanding at the beginning of the period was 50,000. On June 1, 2006, bonus shares were issued in the ratio of 1:1. After the issue of bonus shares, the number of outstanding shares increased to (50,000�2), or 1,00,000.

In the financial statement for 2006, the EPS for 2005 would be readjusted to Rs 10 (Rs 10,00,000/1,00,000) to ensure comparability, even though it was reported at Rs 20 in financial statements for the year 2005.

Usually rights issues have a bonus element too, because shares are issued at a price which is lower than the market price (fair value) prevailing at the time of the issue of shares. Therefore, while computing EPS, the number of shares outstanding before the rights issue also needs to be adjusted for the bonus element in the rights issue.

The measure of earnings per share that we have been discussing so far is known as the 'Basic EPS.' While it adjusts for a change in the number of outstanding shares during the reporting period, it does not take into account the possible dilution of equity arising from instruments like convertible bonds or stock options which can be converted into equity shares.

Dilution is a reduction in EPS or an increase in loss per share resulting from the assumption that convertible instruments are converted, that options or warrants are exercised, or that ordinary shares are issued upon the satisfaction of specified conditions. 'Diluted EPS' is a variant of the EPS which tries to adjust for this dilution too. For the purpose of computing 'Diluted EPS,' an enterprise assumes that all dilutive potential equity shares have been converted into equity shares at the beginning of the period, or if issued later, the date of the issue of potential equity shares.

However, the conversion of potential equity instruments into equity shares need not always lead to a dilution of equity. Anti-dilution is said to happen for conversions which lead to an increase in EPS or a reduction in loss per share.

For these conversions, no adjustment to basic EPS for dilution is required.

EPS is not a good measure of performance because it does not consider the opportunity cost of capital and can be manipulated by short-term actions.

Let us take an example. Assume that a company has 20,000 outstanding shares and earnings available to shareholders is Rs 200,000. The EPS is (Rs 2,00,000/ 20,000), or Rs 10. Assume that the company borrows Rs 10,00,000 at an interest rate of 8 per cent to buy back 10,000 shares. Assuming an income tax rate of 40 per cent, the earnings available to shareholders after the shares are bought back will be [Rs 2,00,000 - (1.00 - 0.40) x Rs. 80,000] or 1.52,000. Accordingly, EPS will be reported at [Rs 1,52,000/10,000] or Rs 15.20.

What this calculation misses is the increase in the cost of equity that has taken place because of the company's decision to substitute equity by debt. Since equity-holders are the residual claimants, almost all variability in the firm's earnings have to be absorbed by them.

Increasing the proportion of debt in the capital structure of a firm (in technical terms increasing the leverage or the gearing ratio) reduces the equity base, which faces this variability of earnings, thereby increasing the riskiness and hence the cost of equity even while increasing its expected returns.

Indeed, financial economics tells us that (in the first approximation) these two effects cancel out exactly and the return on invested capital (ROIC) and the cost of capital do not change with a change in the capital structure. Accordingly, economic profit (often called EVA) too does not change with the change in the capital structure. Yet, the example above shows that the EPS can increase in such circumstances.

Therefore, it is not correct to conclude that the increase in EPS always reflects better performance by the company.

For example, if a company finances a new project totally by debt, EPS will increase if the project returns are higher than the after-tax cost of debt, even if the project earns a return lower than the cost of capital (WACC) of the company. Although the EPS increases, such a project destroys value.

Another important shortcoming of EPS is that it does not relate EPS with the invested capital. For example, two companies may have same EPS, even if one company has lower invested capital as compared to the other.

Despite these shortcomings, it remains the case that EPS increases with increase in ROIC and growth if other things remain the same. Therefore, it will be incorrect to say that EPS is a totally useless measure. However, while interpreting increase in EPS we should keep in mind that EPS can be manipulated and that increase in EPS may be coupled with increase in the risk of equity investment.

The writer is professor of finance and control at  IIM-K.