KARACHI: Companies rarely die suddenly. Long before a Pakistani manufacturer defaults on its markup payments or a once-famous PSX name is suspended from trading, the vital signs deteriorate in its financial statements: working capital thins, retained profits stop growing, debt quietly outruns the asset base. In 1968, a young finance professor named Edward Altman asked whether those vital signs could be combined into a single score that predicted bankruptcy in advance — and the answer, tested on failed and surviving American manufacturers, became one of the most durable tools in credit analysis. The Altman Z-score will not tell you which share to buy, and it was never designed for banks or brand-new companies. But as an early-warning gauge for the industrial companies that fill much of the PSX, it teaches investors to read distress the way lenders do: early, numerically, and without sentiment.
The five ingredients
The original Z-score blends five ratios, each probing a different organ of the company. Working capital ÷ total assets measures liquidity: the cushion of short-term resources over short-term claims, scaled by size. Shrinking cushions precede most failures. Retained earnings ÷ total assets measures cumulative profitability and, implicitly, age and dividend discipline — a company that has banked profits for years carries a reservoir; one that has eaten losses carries a hole. EBIT ÷ total assets measures the earning power of the asset base before financing and tax effects — the engine's raw output. Market value of equity ÷ total liabilities asks how far the company's market value could fall before its debts exceed its assets; it is the only ingredient that listens to the stock market itself. Sales ÷ total assets measures asset turnover — how hard the asset base works to generate revenue.
Altman did not choose the weights by intuition. Using a statistical technique (multiple discriminant analysis) on his sample of bankrupt and healthy firms, he derived coefficients that best separated the two groups: Z = 1.2×(working capital/assets) + 1.4×(retained earnings/assets) + 3.3×(EBIT/assets) + 0.6×(market value of equity/liabilities) + 1.0×(sales/assets). The heavy weight on EBIT/assets tells its own story: genuine earning power over the asset base was the single strongest discriminator between the living and the doomed. The individual ratios will be familiar from guides such as working capital explained and debt ratios explained; Altman's contribution was to fuse them into one verdict and calibrate where safety ends.
The three zones
Scores sort companies into three bands. Above 2.99, the "safe zone": firms in the original study rarely failed. Below 1.81, the "distress zone": firms with these profiles failed at high rates within the following couple of years. Between the two lies the "grey zone", where the model shrugs — some firms recover, some slide, and the trend of the score matters more than its level. That architecture deserves respect rather than mockery: Altman built honest uncertainty into the output, a trait more financial models should copy. In his original tests the model correctly classified the great majority of firms one year before failure, and a strong majority two years out — early enough, in other words, for a lender to tighten terms or an investor to walk away while a market for the shares still existed.
A purely illustrative walk-through shows the arithmetic. Imagine a fictional PSX manufacturer, Example Textiles, with total assets of Rs 10 billion. Working capital is Rs 500 million (ratio 0.05), retained earnings Rs 1 billion (0.10), EBIT Rs 600 million (0.06), market value of equity Rs 2 billion against total liabilities of Rs 7 billion (about 0.29), and sales of Rs 9 billion (0.90). Multiply through: 1.2×0.05 = 0.06; 1.4×0.10 = 0.14; 3.3×0.06 = 0.198; 0.6×0.29 = 0.174; 1.0×0.90 = 0.90. The Z-score totals roughly 1.47 — deep in the distress zone. The diagnosis writes itself: the company turns its assets at a decent clip (0.90 contributes most of the score) but earns too little on them, has banked almost no retained profits, and its market value covers barely a quarter of its liabilities. Now improve only EBIT to Rs 1.5 billion (0.15) and retained earnings to Rs 3 billion (0.30): the score climbs by about 0.58 to roughly 2.05 — out of distress, into the grey. Profitability and patience, the model insists, are what rescue balance sheets.
Versions for the real world: private firms, non-manufacturers, emerging markets
Altman himself spent decades revising the instrument. Because private companies have no market value of equity, a revised model (often called Z′) substitutes book value of equity and recalibrates the weights and zones. A further variant (Z″) drops the sales-to-assets term — the ingredient most specific to manufacturing — to serve retailers, service companies and firms in emerging markets, and its output can be mapped onto bond-rating equivalents. That last mapping is how the Z-score meets Pakistan's own credit apparatus: the rating agencies PACRA and VIS grade issuers from AAA down to default, and their methodology discussions overlap heavily with Z-score territory — liquidity, profitability, leverage and coverage. The guide to Pakistan credit ratings explained describes that system; think of the Z-score as the do-it-yourself cousin of a rating — cruder, free, and available for companies no agency covers.
The boundaries of the tool matter as much as its reach. The original model was calibrated on manufacturers, and its logic fits asset-heavy PSX sectors — textiles, cement, steel, sugar, chemicals, engineering — far better than it fits banks and insurers, whose balance sheets invert the usual meanings of debt and working capital; financial firms need entirely different distress models, and applying Z to a bank produces confident nonsense. Young companies score badly almost by construction (they have had no time to accumulate retained earnings), and a single restructuring year — a large revaluation surplus, a debt rescheduling — can flatter or depress several inputs at once. Reading the score alongside the narrative of the accounts, and watching it across several years rather than in one snapshot, is the professional habit; a falling Z through the grey zone is a stronger warning than any single low print.
Computing a Z-score from Pakistani accounts: practical notes
A few data-handling points decide whether your score means anything. Use the market value of equity from a normal trading period rather than a single excited day, since the fourth ingredient multiplies whatever price you feed it; a spike during results season can drag a genuinely weak firm out of the distress zone on paper. Take retained earnings from the equity section of the balance sheet, where Pakistani companies present them within reserves and unappropriated profit — and treat revaluation surpluses on property with suspicion in this context, because they inflate the equity and asset bases without representing a single rupee of banked trading profit. Construct EBIT from the profit-and-loss account as operating profit before finance costs and taxation, using the company's own presentation; in years when companies book large "other income" — a property sale, an exchange gain — consider computing the score both with and without it, since a one-off gain can dress a deteriorating core business in a healthy costume for exactly one annual snapshot.
Seasonality deserves a mention for Pakistan's agricultural-linked manufacturers. A sugar mill's balance sheet at the end of the crushing season looks nothing like its balance sheet mid-season: inventories bulge, short-term borrowing peaks, and working capital swings through ranges that a single year-end snapshot captures only accidentally. Where interim accounts are published, computing the score at consistent points across several years — always year-end, say — produces a comparable series, whereas cherry-picking the prettiest quarter produces a brochure. Consistency of method, applied dully year after year, is what turns a classroom formula into an early-warning system for your own watchlist.
Distress is not death: what happens after a bad score
A low score starts processes rather than ending them, and investors should know the script. Lenders typically respond first: banks tighten limits, demand additional security, shorten tenors, or move the account to a recovery unit — steps that show up in the accounts as rising finance costs and reclassified debt long before any default headline. Sponsors then choose among unpalatable options: injecting equity, selling assets or divisions, negotiating a rescheduling with the banking consortium, or entering formal restructuring under Pakistan's corporate rehabilitation framework. Each route has a different consequence for minority shareholders — equity injections and asset sales can genuinely repair the numerator problems the Z-score measures; reschedulings sometimes merely postpone them; dilutive rescues at depressed prices can save the company while largely expropriating existing holders. The score tells you the patient is ill; the sponsor's resources and intentions, the bank group's patience, and the saleability of the assets decide the prognosis. Watch for the concrete signals — a rights issue announcement, an asset-disposal notice, a change of auditors or chief financial officer — in the company's PSX disclosures, because those are the moments the statistical warning converts into an investable fact.
Using the Z-score on PSX companies: a field guide
The inputs are all obtainable from published accounts: working capital and total assets from the balance sheet, retained earnings within equity, EBIT constructible from the profit-and-loss account, liabilities from the balance sheet, sales from the top line, and market value of equity as share price times shares outstanding — the market capitalisation concept set out in market capitalisation explained. Spreadsheets make the five-minute version painless, and screening a watchlist annually will catch the classic Pakistani failure pattern early: a company expanding on short-term borrowing while its retained earnings stagnate, in a sector — sugar, textiles, steel — where the cycle can turn before the debt matures. Pair the score with the qualitative checklist in financial statement red flags: qualified audit opinions, auditor resignations, ballooning receivables and related-party loans are the smoke that often accompanies the Z-score's fire alarm.
Equally important is what a good score does not license. A company deep in the safe zone can still be a poor investment — overpriced, badly governed, or in a dying line of business — because the Z-score predicts survival, not shareholder returns; safe companies bought at silly prices lose money the slow way. And a distress-zone score is not an automatic death sentence: firms with committed sponsors, saleable assets or reschedulable bank debt do climb out, and Pakistan's corporate rehabilitation and restructuring frameworks exist precisely because lenders sometimes prefer a living borrower to a liquidated one. Treat the gauge as a triage instrument — it tells you where to spend your scepticism first — and verify with cash flows, because companies ultimately fail from lack of cash, not lack of ratios.
Finally, remember the data's vintage. Pakistani annual reports arrive months after year-end, and interim accounts are thinner; a Z-score computed on last year's statements describes last year's patient. Distress that develops fast — a currency shock doubling the rupee cost of dollar debt, a commodity spike destroying working capital — can outrun the published numbers. The score remains valuable precisely because true balance-sheet rot is usually slow, and slow rot is what it was built to catch. One last boundary deserves stating plainly: the model's original calibration comes from mid-twentieth-century American manufacturers, and while Altman and others have re-estimated versions for other eras and markets, no re-estimation has been run specifically on Pakistani failure data at public scale. Treat the zone boundaries as well-tested guideposts imported from a related climate, not as survey markers measured on local soil — directionally trustworthy, precise only in appearance.
This article is educational content, not financial advice: use the Z-score as one screen among several, and never as a substitute for reading the accounts themselves.Key takeaway: the Altman Z-score fuses liquidity, accumulated profits, earning power, market value against debt, and asset turnover into one gauge — above 2.99 historically safe, below 1.81 distressed. It fits manufacturers, not banks, and its trend warns earlier than its level.
