How this instrument works
Altman's Z-Score compresses five accounting ratios into one number that historically separated companies heading toward bankruptcy from companies that survived. Edward Altman built it in 1968 at New York University by running discriminant analysis on 66 publicly traded manufacturers, half of which had filed for bankruptcy within the prior year. Five weights sit inside that formula — 1.2, 1.4, 3.3, 0.6, and 1.0 — not round numbers picked for convenience, but coefficients that best separated those two groups in his original sample.
Each ratio captures something others miss. Working capital over total assets measures short-term liquidity, cash available against near-term obligations. Retained earnings over total assets is a proxy for both age and cumulative profitability, which is why young, fast-growing firms tend to score lower even when healthy. EBIT over total assets measures operating efficiency before distortions from tax rates and leverage, and it carries outsized weight because it discriminated best in Altman's original sample. Market value of equity over total liabilities substitutes live market pricing of risk for a book figure that can lag reality by a full accounting period. Sales over total assets closes out that formula with a turnover measure: how much revenue each dollar of assets generates.
This model has real limits. It was calibrated on public manufacturers using 1960s data, so its market-value-of-equity term makes it awkward for private companies, and its weights were never re-estimated for service or technology firms with light balance sheets — Altman later published separate Z' and Z'' variants for exactly that reason. A single score also cannot see fraud, pending litigation, off-balance-sheet debt, or a covenant breach maturing next quarter; it only reads what four ratios and one market price already contain.
- Enter Working capital ⁄ total assets, Retained earnings ⁄ total assets, and EBIT ⁄ total assets straight from the balance sheet and income statement.
- Enter Market value of equity ⁄ total liabilities — use share price times shares outstanding for market value, not book equity.
- Enter Sales ⁄ total assets using trailing twelve-month revenue over your most recent total assets figure.
- Read your Altman Z-Score and check it against three zones: below 1.81, from 1.81 to 2.99, and above 2.99.
- Change one ratio at a time to see how much a swing in leverage or profitability moves its score across a zone boundary.
Worked example — default-sheet Z-score of 2.856
Take this default sheet: Working capital ⁄ total assets of 0.15, Retained earnings ⁄ total assets of 0.20, EBIT ⁄ total assets of 0.12, Market value of equity ⁄ total liabilities of 1.5, and Sales ⁄ total assets of 1.1. Its formula multiplies each ratio by its weight — 1.2(0.15) + 1.4(0.20) + 3.3(0.12) + 0.6(1.5) + 1.0(1.1) — giving 0.18 + 0.28 + 0.396 + 0.9 + 1.1, which sums to a Z-score of 2.856.
That number sits in the grey zone, between the 1.81 distress cutoff and the 2.99 safe-zone cutoff — 0.134 short of being read as clearly healthy on Altman's original scale. A grey-zone reading does not predict failure; it flags a company whose ratios overlap both historically bankrupt and historically solvent groups in Altman's 1968 sample, which is exactly why analysts treat it as a prompt for closer reading of financial statements rather than a verdict on its own.
Questions
What counts as a healthy Altman Z-Score?
Altman's original scale for public manufacturers reads above 2.99 as the safe zone, 1.81 to 2.99 as the grey zone, and below 1.81 as the distress zone historically associated with bankruptcy within two years in his 1968 study. It is zones, not a single pass mark, because discriminant analysis behind that formula produced overlapping probability bands rather than one hard cutoff.
Why does market value of equity carry more weight than book value here?
Its D term uses share price times shares outstanding instead of book equity because a live market price already reflects investors pricing in risk that the balance sheet has not caught up to yet — a falling stock ahead of falling earnings is common. Substituting book equity understates leverage risk, which is one reason Altman published a separate private-firm variant, Z', for companies without a traded share price.
Does this Z-score work for private companies or non-manufacturers?
Not this original formula — it was fit to publicly traded manufacturers, and its market-value term needs a share price a private company doesn't have. Altman later published Z' for private manufacturers, substituting book equity, and Z'' for private and non-manufacturing firms, which also drops its sales-to-assets term. Applying public-manufacturer weights to a service company skews that score.
Can a company with a safe-zone score still go bankrupt?
Yes. This score reads five accounting ratios and one market price; it cannot see fraud, a covenant breach maturing next quarter, off-balance-sheet debt, or a sudden loss of a major customer. Altman's own back-testing found his model correctly classified roughly 80 to 90 percent of firms one year ahead of failure, which also means a meaningful minority were misclassified in both directions.
How is this different from just checking the current ratio or debt-to-equity?
A single ratio like the current ratio reads one dimension, short-term liquidity, and misses profitability, leverage, and turnover entirely. This Z-score's five weights came from discriminant analysis that tested which combination of ratios best separated companies that failed from companies that survived, so it carries information no single ratio holds alone, at the cost of being harder to sanity-check by eye.
Why does EBIT ⁄ total assets get such a large weight, 3.3?
In Altman's original discriminant analysis, EBIT over total assets — a measure of operating earning power independent of tax rate and debt load — separated bankrupt and non-bankrupt groups in his sample more sharply than any other of the four ratios, so that statistical procedure assigned it a large coefficient. It reflects discriminating power in that dataset, not a claim that profitability matters more in principle.
References
- New York University Stern School of Business — Edward Altman's Z-Score research
- U.S. Securities and Exchange Commission — Investor.gov education hub
Read this first: This instrument shows arithmetic, not advice. Real offers add fees, taxes and terms that vary by lender and place — verify the figures against your actual paperwork before deciding anything.