KARACHI: When results season arrives, one number escapes the financial pages and travels everywhere: earnings per share. "EPS up 34 per cent." "EPS misses estimates." "EPS of Rs 42.15." It is the unit in which corporate success is publicly measured, the denominator of the price-to-earnings ratio, the numerator of dividend cover, and the figure on which a thousand bonus calculations and headline judgements quietly depend. Yet for a number so quoted, EPS is remarkably easy to move without improving the business at all. Understanding exactly how it is constructed — what goes into the profit, what counts as a share, and what can inflate or deflate either — is basic hygiene for anyone who reads a results announcement.
The basic formula
Basic earnings per share is: profit attributable to ordinary shareholders ÷ weighted average number of ordinary shares outstanding. Both halves deserve a closer look. The numerator is not the company's total profit: it is profit after tax, minus anything belonging to others — the share of profits attributable to minority interests in subsidiaries, and dividends owed to preference shareholders, whose claims rank ahead of yours. What remains belongs to the ordinary shareholders collectively. The denominator is not simply the shares in issue on the last day of the year: it is a weighted average over the period, so shares issued halfway through the year count for half a year. That weighting matters enormously in a market where companies issue shares frequently — through rights issues, bonus issues and conversions — because a naive year-end count would distort the per-share result. EPS is the finishing line of the income statement; our guide to reading an income statement traces the path from revenue down to the profit from which EPS is drawn.
A worked example with invented numbers
Take a fictional company, Bolan Beverages Limited, invented for illustration with round numbers. Bolan reports profit after tax of Rs 2.4 billion. A subsidiary it owns 80 per cent of contributed part of that profit, and Rs 200 million belongs to the minority partners; there are no preference shares. Profit attributable to Bolan's ordinary shareholders is therefore Rs 2.2 billion. Bolan had 200 million shares outstanding for the first half of the year, then issued 40 million more in a rights issue, giving a weighted average of 200 million × ½ + 240 million × ½ = 220 million shares. Basic EPS is Rs 2.2 billion ÷ 220 million = Rs 10 per share. Every number in that little calculation carries a lesson. Compare Bolan with a rival reporting the same Rs 2.4 billion profit: if the rival has no minorities and half the share count, its EPS is dramatically higher — not because its business is better, but because its profit is divided among fewer claims and fewer shares. EPS is a per-share measure of a particular company's structure, not a universal score; cross-company comparisons need the ratios built on top of it, starting with the P/E discussed below.
Basic versus diluted EPS
Some claims on future shares are already lurking in a company's capital structure: convertible bonds and preference shares that can turn into ordinary shares, employee share options, warrants attached to earlier issues. If all of them converted, today's shareholders would own a thinner slice, and each share's claim on profit would shrink. Diluted EPS performs that thought experiment: it recalculates EPS as if the diluting instruments had converted — adjusting the share count up, and the profit up too where conversion would save interest or preference dividends — and reports the lower, more conservative figure. Accounting standards (IAS 33, which governs EPS presentation) require listed companies to show both numbers, and the gap between them is itself information. A small gap says the capital structure is clean. A wide gap says a meaningful part of the company's future is already promised to others — an important footnote when valuing the shares, and a standard check in careful readings of results, alongside the one-off items and "underlying" EPS figures management sometimes promotes instead. Treat management-adjusted EPS the way you treat any edited highlight reel: interesting, but start from the audited standard version.
Bonus shares, rights issues and why old EPS gets restated
Pakistani investors meet a distinctive EPS complication constantly: the bonus issue. When a company issues free shares to existing holders — one for every five held, say — no new money or value enters the business, yet the share count jumps. If EPS were simply recalculated on the new count, it would appear to collapse even though nothing economic changed. The standards therefore require prior-period EPS to be restated as if the bonus shares had always existed, so this year's EPS and last year's sit on the same share base and the growth rate means something. Rights issues — new shares sold to existing holders, usually at a discount — are handled with a subtler adjustment reflecting the bonus element hidden in the discount. Our explainer on bonus and right shares on the PSX works through both mechanics in full. The practical takeaway for anyone charting a company's EPS history from old annual reports or data websites: check whether the figures you are comparing were restated after later share issues. Many an apparent EPS "decline" in a long series is just an unrestated bonus issue, and many a triumph is a changed share count wearing a growth costume.
How EPS gets managed — buybacks, one-offs and timing
Because EPS is the scoreboard, it attracts games. The most respectable-looking is the share buyback: spend corporate cash repurchasing shares, the denominator falls, and EPS rises even if total profit is flat. A buyback below intrinsic value genuinely benefits remaining shareholders; a buyback at any price merely to polish EPS can destroy value while improving the headline — the arithmetic cannot tell the difference, so the reader must, by asking what else the cash could have done and at what price the shares were retired. Then come the numerator games: classifying recurring costs as "exceptional" so that underlying EPS excludes them, year after year; timing asset sales into a weak quarter; capitalising costs that were once expensed, which lifts current profit at the expense of future depreciation. None of these changes cash by a single rupee, which is why the seasoned reader always triangulates EPS against free cash flow per share: two series that should rhyme over a cycle, and whose persistent divergence tells you the profit figure is being curated. A final quiet lever is the tax rate and minority structure — consolidating or deconsolidating a subsidiary can move attributable profit without any factory selling a single extra unit.
Executive pay ties the whole subject together, because in many companies management bonuses and option grants are calibrated to EPS targets. That alignment has a wholesome face — managers rewarded for growing the owners' per-share claim — and a distorting one: every lever described in this section, from buybacks timed for the measurement date to costs classified away, pays its operator personally when EPS is the scoreboard. Governance codes and remuneration committees exist partly to police exactly this, by setting targets on cash-backed or multi-year measures. The individual investor cannot redesign the bonus plan, but can read it: the annual report discloses the performance conditions attached to executive awards, and a board that pays its executives on adjusted EPS, while reporting statutory EPS that keeps disappointing, has told you something no ratio can.
EPS at work: P/E, growth and payout
For all its flaws, EPS earns its ubiquity because three essential tools are built from it. Divide price by EPS and you have the P/E ratio, the market's most quoted price tag, whose growth-adjusted cousin — the PEG ratio — divides P/E itself by the EPS growth rate, so the credibility of the growth rate in this year's results feeds straight into valuation. Divide the dividend per share by EPS and you have the payout ratio, the quick test of how much of each year's profit is being handed back to shareholders. And EPS growth, compounded over years and honestly measured, remains the single best statistical summary of whether a company's per-share economics are actually improving — superior to total profit growth precisely because it refuses to let management buy growth by endlessly issuing shares. The tests to apply are consistency (growth across a full cycle, not one rebound year), quality (growth matched by cash flow), and provenance (growth earned by the business rather than manufactured by the share count or the classification pen). A company passing all three, year after year, is compounding in the only sense that matters to the people who own one share of it.
Trailing, forward and annualised: which EPS is being quoted?
The letters "EPS" on a screen may refer to at least three different numbers, and confusing them manufactures fake bargains and fake disasters. Trailing EPS sums the last four reported quarters — factual, audited in part, but backward-looking and possibly loaded with a one-off gain from nine months ago. Forward EPS is an analyst consensus forecast for the coming year — relevant to what you are really buying, but an opinion, and one that herds: forecasts cluster, drift down slowly after bad news, and are famously optimistic at market peaks. Annualised EPS multiplies a single quarter by four — the most dangerous of the three in a seasonal economy, since it would annualise a sugar company's crushing-season quarter or a fertiliser maker's peak-application quarter into a fantasy run-rate. Pakistani results also arrive on both standalone and consolidated bases, which differ wherever subsidiaries and minorities are significant, exactly as the worked example showed. The rules of engagement are simple: know which EPS your P/E divides by, compare like with like across companies and years, and when a quoted multiple looks impossibly cheap, suspect the denominator before you celebrate — you have usually found a trailing figure swollen by a non-recurring item, or an annualised figure built on the season's best quarter.
Finally, EPS sits at the centre of a expectations game worth understanding. Professional forecasts aggregate into a consensus, and prices move less on whether EPS rose than on whether it beat the number the market had already pencilled in — a company can report record EPS and fall, because the whisper number was higher still, or report a decline and rally, because worse had been feared. For the long-term holder this theatre is mostly noise, but it explains price behaviour around results that baffles beginners, and it reinforces the guide's central discipline: the trend of honestly measured, cash-backed EPS over years is information; the quarterly dance around expectations is commentary. Read the dance if you enjoy it. Invest on the trend.
One habit ties the whole guide together for the reader of results announcements: always read EPS in a sentence, never as a fragment. The sentence has four clauses — EPS of so much, up or down so much on a restated base, on a share count that changed in this way, with this much of the movement from one-off items and this much from operations. Any results story that gives you fewer clauses than that is a summary written by somebody with an interest in which clauses you never see. The full sentence is always available in the accounts themselves, a few pages behind the headline — and the few minutes it takes to assemble it are among the best-compensated minutes in an investor's calendar, season after season, company after company.
This article is educational content explaining how earnings per share is calculated and used; it is not financial advice, and the company and figures in its example are fictional illustrations rather than data about any real business.
Key takeaway: EPS divides the profit attributable to ordinary shareholders by the weighted average share count — a per-share scoreboard that powers the P/E and payout ratios, but one that buybacks, one-offs and share issues can move without the business improving at all.
