KARACHI: Two companies can report identical profits and be wildly different businesses — one earning its crore on a mountain of shareholders' money, the other on a molehill. Return on equity (ROE) and return on assets (ROA) are the ratios that expose the difference, by dividing profit by the resources used to produce it. They are the nearest thing accounting offers to an efficiency grade, and — used carefully, with their tricks understood — among the most informative numbers an investor can compute from a PSX annual report.
The two definitions
Return on equity = profit after tax ÷ shareholders' equity. It answers the owner's question: for every rupee I have invested and left in this company, how many rupees of profit did it generate this year?
Return on assets = profit after tax ÷ total assets. It answers the manager's question: for every rupee of resources the business controls — however financed — how much profit came out? (Some analysts use profit before interest and tax in the ROA numerator, since assets are funded by lenders too; be consistent when comparing, and note which version a data source uses.)
Continue with the illustrative company used across this series, the fictional Example Mills Ltd. Its profit after tax was Rs 1,050 million; its equity Rs 5,500 million; its total assets Rs 11,300 million. So:
- ROE = 1,050 ÷ 5,500 ≈ 19.1 per cent
- ROA = 1,050 ÷ 11,300 ≈ 9.3 per cent
Both figures use year-end balances; careful analysts average opening and closing balances, since profit is earned across the whole year on a base that itself grew. Either way, the interpretation is parallel: Example Mills turned each rupee of shareholder funds into about 19 paisas of annual profit, and each rupee of total assets into about 9 paisas. Whether 19 per cent is good depends on the alternatives and the risk — a benchmark worth keeping in mind is the return available on near-riskless government securities at the time, which company returns should beat over a cycle to justify equity risk. Check the current State Bank policy rate and treasury bill yields for that comparison rather than relying on any figure printed in an article.
Why ROE always exceeds ROA here — the leverage bridge
The gap between the two ratios is not a coincidence; it is arithmetic. Assets are funded by equity plus debt, so assets exceed equity, and the same profit divided by the smaller equity base yields the bigger percentage. Debt is the lever between them: Example Mills' assets are 11,300 ÷ 5,500 ≈ 2.05 times its equity, and 9.3 per cent × 2.05 ≈ 19.1 per cent. A company with no debt would show ROE nearly equal to ROA; a highly geared company can display a dazzling ROE on a mediocre ROA, because lenders supplied most of the resources while shareholders claim all the residual profit. This is why the pair must be read together: ROE without ROA can be a leverage report disguised as a performance report. The balance sheet anatomy behind the lever is covered in our guide to reading a balance sheet.
The DuPont breakdown: three engines of ROE
A century-old decomposition, from the DuPont company, splits ROE into three multiplicative parts:
ROE = net profit margin × asset turnover × financial leverage
For Example Mills: margin = 1,050 ÷ revenue 12,000 = 8.75 per cent; turnover = 12,000 ÷ 11,300 ≈ 1.06; leverage = 2.05. Multiply: 8.75% × 1.06 × 2.05 ≈ 19.1 per cent — the same ROE, now explained. The power of the breakdown is diagnostic. A retailer earns thin margins at high turnover; a utility earns fat margins at low turnover; a geared property business leans on the third lever. When a company's ROE jumps, DuPont tells you which engine fired: genuinely better margins (operational improvement), faster asset use (efficiency), or more leverage (more risk). Only the first two are unambiguously good news, and a rising ROE driven purely by the leverage term should be read alongside the interest cover and gearing checks from the income statement and balance sheet guides.
What distorts the ratios
Both ratios inherit the weaknesses of their inputs, and Pakistani accounts offer several classic distortions. Revaluation reserves: when a company revalues its land and buildings upward — common among older Pakistani industrials — equity and assets swell without any profit being earned, mechanically depressing ROE and ROA even though the business did not worsen; a company that never revalues will look artificially efficient by comparison. Accumulated losses: a company whose equity has been eroded to a sliver can post astronomical ROE on trivial profits — 200 per cent of almost nothing — which is arithmetic, not excellence. One-off items: a gain on selling property inflates one year's profit and both ratios; strip such items out, or use multi-year averages, before concluding anything. Industry structure: banks operate with leverage that would be alarming anywhere else, so their ROE/ROA relationship follows banking norms and should be compared only within the sector. Inflation: in high-inflation years, profits measured in today's rupees are divided by asset bases recorded at yesterday's costs, flattering both ratios across the whole market.
Sensible benchmarks and honest comparisons
Use the ratios comparatively and across time, never as absolute verdicts. Compare a company's ROE with its own five-year history (is the engine strengthening?), with direct competitors in the same sector (same accounting weather), and with its cost of equity in a rough sense — a business persistently earning ROE below what government paper yields is destroying shareholder value on a risk-adjusted view, whatever its profits say. Sustained ROE in the high teens or above, funded by margins and turnover rather than escalating leverage, across a full economic cycle, is the signature of a genuinely strong franchise; few companies anywhere manage it, which is precisely why the signature is valuable. ROA, meanwhile, is the cleaner cross-check of operating efficiency between companies with different financing choices, and the better starting ratio when leverage differs wildly between the firms compared.
The dividend connection
ROE also disciplines dividend analysis, tying this series together. A company can only grow its equity — and, sustainably, its dividends — at roughly the rate of ROE multiplied by its retention ratio: Example Mills, with ROE of 19.1 per cent and retention of 60 per cent (from the payout analysis in the previous guide), can grow its equity base at about 11–12 per cent a year without outside funding. A board paying out 90 per cent of profits while promising 20 per cent growth is making a promise arithmetic cannot keep. Conversely, retention only deserves applause when ROE is high: shareholders should want profits paid out, not compounded internally, when the company's ROE trails what they could earn elsewhere. The payout question — examined in our guide to the dividend payout ratio and dividend cover — is really the ROE question wearing different clothes.
Used this way — as a pair, decomposed, trended and cross-checked — ROE and ROA convert three dense financial statements into a one-line character reference for a business: how hard its money works, and how honestly that work is financed. This article is educational content, not financial advice; ratios inform judgement, they do not replace it.
A worked DuPont comparison: same ROE, different companies
Two fictional firms both report ROE of 20 per cent. MarginStar earns a 20 per cent net margin with asset turnover of 0.8 and leverage of 1.25 — profit from pricing power, conservatively financed. LeverageStar earns a 4 per cent margin with turnover of 1.25 and leverage of 4.0 — profit from volume and borrowed money. The headline ratio pronounces them equals; the decomposition shows they are opposites. MarginStar's ROE survives a recession and a rate shock; LeverageStar's ROE evaporates if margins slip one point or lenders raise rates, because its thin margin must service a heavy structure. Pakistani investors meet both archetypes constantly — the conservatively financed consumer franchise and the geared cyclical trading at the same "attractive" ROE — and DuPont is the five-minute test that tells them apart. The full ratio toolkit continues in the next guide in this series, on debt ratios and interest cover.
ROCE: the cousin worth knowing
A close relative refines ROA for companies with significant debt. Return on capital employed divides operating profit (before interest and tax) by capital employed — equity plus non-current liabilities, the long-term money in the business. By using pre-interest profit and long-term capital, ROCE measures the return the whole enterprise generates for all its long-term funders, making it the natural benchmark against the cost of that funding: a company earning ROCE persistently below its borrowing rate is shrinking its lenders' and owners' combined wealth with every year of operation, however pleasant its ROE looks after leverage. Analysts comparing two companies with different debt levels often start with ROCE, then let ROE show what the financing choice added — or subtracted — for shareholders specifically.
ROE and the price you pay
Finally, connect the ratio to valuation. A high ROE justifies a high price-to-book multiple: investors rationally pay several times book value for equity that compounds at 25 per cent, and close to book — or below — for equity compounding at 6 per cent while government paper yields more. This is why the P/B ratio cannot be judged without ROE beside it, and why screens that hunt "low P/B" alone fill portfolios with low-ROE companies whose cheapness is merely accuracy. The mature workflow runs the other way: find businesses whose ROE is high, durable and honestly financed; then ask whether the market price already assumes it. Ratios begin that investigation better than any other tool — and end it, properly, never.
One more caution completes the toolkit: both ratios are backward-looking averages over a whole year, and they smooth the very cycles Pakistani investors live through. A cement company's ROE at a demand peak and at a trough can differ by twenty points, and quoting either as "its ROE" misleads. The remedy is the one applied throughout this series: average across a full cycle, compare within the sector, and treat any single year as a sample, not a character reference. Companies that hold respectable ROE through the trough — funded by margin rather than leverage — are the ones whose compounding survives contact with Pakistan's economic weather, and finding them is most of what fundamental investing here consists of.
For a practical routine, compute both ratios for any company you seriously follow, once a year, from the audited accounts: profit after tax over average equity, profit after tax over average assets, and the DuPont split when ROE moves sharply. File the three numbers beside the year's EPS and dividend, and in three years you will possess a compact dossier that distinguishes — better than price charts ever will — the companies whose returns are structural from those whose returns were weather. Pakistan's market contains both kinds in every sector, often at similar prices; the ratios are how a patient investor tells them apart before the market does.
That is the standard to hold every glossy results headline to: not whether profit rose, but whether the resources behind the profit — shareholders' rupees and the assets they command — were made to work harder, and whether the way they were made to work can survive the next turn of Pakistan's reliably dramatic economic cycle. Few analytical habits repay an investor more reliably, year after year, than insisting on that answer before applauding the profit figure itself — because profit is an event, but return on resources is a character trait, and portfolios are built, in the end, out of character traits rather than headlines.
Key takeaway: ROE measures profit per rupee of shareholders' funds and ROA per rupee of all assets; the gap between them is leverage, so read them together, decompose ROE into margin, turnover and gearing, and trust trends within a sector more than any single year's headline.
