Altman Z-Score Calculator (Bankruptcy Prediction)
Altman Z-Score calculator. The 1968 5-factor bankruptcy prediction model. Safe, grey and distress zones with each weighted ratio shown, and what Altman's own accuracy tests do and do not say.
Altman Z-Score Calculator
Z-Score decomposition
| Factor | Definition | Ratio | Weight | Contribution |
|---|---|---|---|---|
| X1 — Liquidity | WC / Total Assets | — | 1.2 | — |
| X2 — Cumulative Profitability | Retained Earnings / Total Assets | — | 1.4 | — |
| X3 — Operating Efficiency | EBIT / Total Assets | — | 3.3 | — |
| X4 — Market Leverage Buffer | Market Equity / Total Liab | — | 0.6 | — |
| X5 — Asset Turnover | Sales / Total Assets | — | 1.0 | — |
| Total Z-Score | — | |||
How to use the Altman Z-Score calculator
Gather 7 balance-sheet + P&L line items
You need: Working Capital (current assets − current liabilities), Retained Earnings, Total Assets, EBIT (operating income), Sales, Market Capitalization, Total Liabilities. All from the most recent 10-K or annual report; market cap from your broker. All in the same currency, ideally $ millions for typical company-scale precision.
Enter the values
Negative working capital is normal for some industries (retail with high payables, software with deferred revenue) — enter the actual value, negative or positive. Retained earnings can be negative for young firms or those with accumulated losses — also enter the actual value. The Z-Score formula handles negative values correctly.
Read the headline Z-Score
Three zones: Z > 2.99: SAFE (the upper bound in common use). 1.81 ≤ Z ≤ 2.99: GREY (warning zone — heightened scrutiny). Z < 1.81: DISTRESS (in Altman's sample ~72% of the firms that DID fail had scored here two years out. That is the test's SENSITIVITY, and it is not the same as the share of low scorers who fail — reading it that way requires far more firms to fail than ever do). The zone map shows where your firm sits on the 0-5 scale.
Study the decomposition table
The Z-Score is the sum of 5 weighted ratios. Look at the contribution column to identify which factors are dragging the score: low X3 (EBIT/Assets) = poor operating profitability. Low X4 (Equity/Liab) = high leverage. Low X1 (WC/Assets) = liquidity stress. Low X2 (RetEarn/Assets) = limited cumulative profitability (often young firms). Low X5 (Sales/Assets) = asset productivity problem.
Use Z-Score as a screen, not a verdict
Altman Z is a 1968 model for US public manufacturers. It works less well for: (1) Banks + insurance — different balance-sheet structure. (2) Service / software firms — asset-light. (3) Private firms — use Z' variant with book equity. (4) Non-US firms — different accounting standards. Treat the Z-Score as one input into credit analysis; combine with cash-flow analysis (FCF calculator) and qualitative factors (management, industry trajectory).
Altman Z-Score — the 1968 model still used in credit analysis worldwide
Edward Altman published the Z-Score in 1968 in the Journal of Finance ("Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy", 23(4):589-609). He used multiple discriminant analysis, a statistical technique that finds the weighted combination of measurements that best separates two groups, on 66 US manufacturers: 33 that had filed for bankruptcy between 1946 and 1965 and 33 that had not, matched by industry and size. The five ratios he chose and their weights are still the model this calculator uses.
The formula
As Altman restates it in his 2000 revisit (Altman, Predicting Financial Distress of Companies: Revisiting the Z-Score and ZETA Models (2000)): Z = 1.2X1 + 1.4X2 + 3.3X3 + 0.6X4 + 1.0X5, where X1 is working capital over total assets, X2 retained earnings over total assets, X3 EBIT over total assets, X4 the market value of equity over the book value of total liabilities, and X5 sales over total assets. The original paper wrote the first four as percentages with coefficients one hundred times smaller; Altman points out that the two forms give identical scores. EBIT over assets carries the largest weight.
What the accuracy figures mean
On his original sample, the model classified 95% of the 66 firms correctly using data from one financial statement before bankruptcy. Two statements before, it correctly identified 72% of the bankrupt firms. Those figures describe how many failing firms the model caught. They do not say how many low-scoring firms go on to fail, and the second figure is often misquoted that way. Altman's own later data show why: in 1999 more than 20% of US industrial firms on Compustat scored below 1.81, far more than ever went bankrupt. In tests on later samples of bankrupt firms, from 1969 to 1999, the model caught between 82% and 94% of them a year ahead using a cutoff of 2.67.
"I advocate using the lower bond of the zone-of-ignorance (1.81) as a more realistic cutoff Z-Score than the score 2.675." Altman, Predicting Financial Distress of Companies: Revisiting the Z-Score and ZETA Models (2000)
The zones
This calculator flags scores below 1.81 as distress, 1.81 to 2.99 as grey and above 2.99 as safe. The 1.81 boundary is the one Altman recommends as a cutoff, and he calls the range where his model misclassifies firms the "zone of ignorance". The single cutoff with the fewest errors in his original tests was 2.675. The 2.99 upper boundary is the figure in common use for the original model.
When the model does not fit
The model was built on US public manufacturers. Banks and insurers have balance sheets made of financial assets, so its ratios do not mean what they mean for a manufacturer. Asset-light and young firms can score low for reasons that are not distress, such as negative retained earnings while they grow. For firms without a share price, Altman re-estimated the model with the book value of equity (Z′ = 0.717X1 + 0.847X2 + 3.107X3 + 0.420X4 + 0.998X5, grey area 1.23 to 2.90), and for non-manufacturers he dropped sales over assets (Z″). A 2017 international study by Altman, Iwanicz-Drozdowska, Laitinen and Suvas (2017) found the general model "works reasonably well for most countries", with prediction accuracy of about 0.75, and above 0.90 when re-estimated for each country.
Altman's five weighted ratios, his own accuracy tests and the cutoffs
Altman published the Z-Score in the Journal of Finance in 1968, using 66 US manufacturers.
Half the sample had filed for bankruptcy between 1946 and 1965; the other half were matched survivors.
The formula is 1.2X1 + 1.4X2 + 3.3X3 + 0.6X4 + 1.0X5, with EBIT over assets weighted most.
One statement before failure the model classified 95% of the original sample correctly.
Two statements before failure it caught 72% of the bankrupt firms. That is not the share of low scorers that fail.
In 1999 over 20% of US industrial firms on Compustat scored below 1.81, Altman reports.
Altman recommends 1.81 as the more realistic cutoff; the lowest-error cutoff in his original tests was 2.675.
On bankruptcies from 1969 to 1999 the model caught 82% to 94% one year ahead at a 2.67 cutoff.
For private firms Altman uses Z′ with book equity, with a grey area of 1.23 to 2.90.
A 2017 study of 34 countries found accuracy of about 0.75, above 0.90 with country-specific models.
Frequently asked questions
-
On Altman's original 66-firm sample, 95% were classified correctly one financial statement before bankruptcy, and 72% of the bankrupt firms were identified two statements before. In his later tests on bankrupt firms from 1969 to 1999 the model caught 82% to 94% one year ahead, using a cutoff of 2.67. Accuracy falls further out: 48% at three years and 29% at four on the original sample. The figures measure how many failures the model catches, not how many low scorers fail; Altman notes that the share of healthy firms wrongly flagged has risen to around 15 to 20%.
-
Reasonably well, with adjustments. A 2017 study by Altman and co-authors tested it on firms from 31 European and three non-European countries and found prediction accuracy of about 0.75 for the general model, rising above 0.90 when the model was re-estimated for each country with extra variables. For non-manufacturers and emerging markets Altman uses the Z″ variant, which leaves out sales over assets.
-
Look at the decomposition to see which ratios pull the score down, and at the trend over several years rather than one snapshot. Altman calls this range the "zone of ignorance" because his model misclassifies firms there; in one of his tests, 10 of 14 misclassified firms scored between 1.81 and 2.67. Combine the score with cash flow, debt maturities and interest cover before drawing conclusions.
-
A bank's assets are mostly loans and securities, its liabilities mostly deposits, and its leverage high by design. Working capital and sales over assets do not mean for a bank what they mean for a manufacturer, so the ratios and the weights fitted to manufacturers do not carry over. Banks are assessed with capital, asset-quality and liquidity measures instead.
-
For firms without a market value of equity, Altman re-estimated the whole model with book equity in X4: Z′ = 0.717X1 + 0.847X2 + 3.107X3 + 0.420X4 + 0.998X5. The cutoffs change too: below 1.23 is distress and above 2.90 safe, with a wider grey area between. This calculator uses the original model; putting book equity into it gives only a rough reading.
-
Market-based models such as Merton's and the KMV approach estimate default risk from the market value of a firm's equity and its volatility. Altman notes that the Z-Score's fourth ratio also uses market value, as a proxy for how far assets can fall before liabilities exceed them. Market models need share-price data and more computation; the Z-Score needs five ratios from the accounts.
-
Negative working capital means current liabilities exceed current assets. For a firm in difficulty it can signal a liquidity squeeze. For some business models it is normal, for example when customers pay before the firm pays its suppliers, or when a subscription business collects revenue in advance. The Z-Score cannot tell these apart, so read a low X1 against how the business is run.
-
No. It classifies firms into groups; it does not estimate when a firm might fail. Its accuracy is highest one statement before failure and falls with time, which is why a score is most useful when tracked over several periods.
-
No. All 7 input values (WC, RE, EBIT, sales, market equity, total assets, total liab) stay in your browser. The Z-Score computation + zone classification + decomposition table run as client-side JavaScript. Open DevTools → Network when you click Calculate and you'll see zero outbound requests. Safe for confidential credit analysis.
-
Altman EI. Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy. Journal of Finance 1968;23(4):589-609, doi:10.1111/j.1540-6261.1968.tb00843.x. His 2000 paper, "Predicting Financial Distress of Companies: Revisiting the Z-Score and ZETA Models", restates the model, the Z′ and Z″ variants and later test results.
Method & sources
How it computes
Z = 1.2X1 + 1.4X2 + 3.3X3 + 0.6X4 + 1.0X5 (working capital, retained earnings and EBIT over total assets; market equity over total liabilities; sales over total assets), as restated by Altman (2000), with zones below 1.81, 1.81-2.99 and above 2.99 and each weighted ratio shown.
What this tool implements
- The page and the result hint said '~70% of distress-zone firms filed for bankruptcy within 2 years'. Altman's 72% is the share of bankrupt firms correctly classified two statements before failure, and he reports over 20% of US industrial firms scoring below 1.81 in 1999. Corrected everywhere.
- From Altman (2000): 95% of the 66-firm sample correct one statement before; 82-94% of later bankruptcies caught at a 2.67 cutoff; 1.81 recommended as the more realistic cutoff, 2.675 the lowest-error one; Z' coefficients and 1.23/2.90 cutoffs. The 2.99 upper bound is stated as the figure in common use, not attributed.
- Removed as unsourced: 'used by every major rating agency', Altman's age, the Texas Ratio detail, a book citation, and Walmart/Costco examples. The 2017 international study (accuracy about 0.75) is cited.
Sources
- Altman EI. Predicting Financial Distress of Companies: Revisiting the Z-Score and ZETA Models (2000), NYU Stern; read via Internet Archive. https://web.archive.org/web/2024/https://pages.stern.nyu.edu/~eal…
- Altman EI, Iwanicz-Drozdowska M, Laitinen EK, Suvas A. Financial Distress Prediction in an International Context: A Review and Empirical Analysis of Altman's Z-Score Model. J Int Financ Manag Account 2017;28(2):131-171. https://doi.org/10.1111/jifm.12053
Pick up where you left off
Stored only in this browser — never sent to our servers.