What does the Altman Z-Score actually predict?

The probability that a publicly-traded manufacturing firm will file for bankruptcy within roughly two years. Altman's 1968 study correctly classified 95% of bankrupt firms in the original sample one year out and 72% two years out — a remarkable level of predictive accuracy from just five ratios.

What's the difference between Z, Z', and Z''?

Z (original 1968) uses Market Value of Equity / Total Liabilities for X4 — applies to public manufacturers. Z' (1983) replaces market value with book value of equity for private companies. Z'' (1995) drops the asset-turnover ratio (X5) to apply to non-manufacturers and emerging-market firms. Choose the variant that matches the company you're analyzing — using the wrong coefficients on a service business produces misleading scores.

What are the cutoff zones?

For original Z: above 2.99 = Safe Zone (low default risk), 1.81-2.99 = Grey Zone (warrants investigation), below 1.81 = Distress Zone (high probability of default). Z' uses 2.90 and 1.23. Z'' uses 2.60 and 1.10. The grey zone is where most useful analysis happens — clearly safe and clearly distressed firms don't need a model.

Why does the model use these five ratios specifically?

Altman used Multiple Discriminant Analysis (MDA) on 22 candidate ratios from 66 firms (33 bankrupt, 33 healthy). The five surviving ratios — working capital/assets, retained earnings/assets, EBIT/assets, market equity/total liabilities, sales/assets — each captured a different dimension (liquidity, cumulative profitability, current profitability, market signal, asset productivity). The coefficients were fitted by MDA to maximize between-group variance.

Does the Z-Score work in 2026?

It still has predictive value, but the original cutoffs are arguably stale. Altman himself published a 2017 update suggesting Distress Zone be moved from 1.81 to 1.23 for modern firms, reflecting changes in capital structure and accounting. For service-economy companies, Z'' is the preferred variant; for tech firms with heavy intangibles, the model under-detects strength because retained earnings are dampened by buybacks and stock-based comp.

How does Z-Score compare to credit ratings?

Z-Score and S&P / Moody's credit ratings agree directionally but not perfectly. A Z above 3.0 typically corresponds to investment grade (BBB-/Baa3 or better); Z between 1.81 and 3.0 maps to BB/Ba; Z below 1.81 maps to B/Caa or worse. Credit ratings incorporate qualitative factors (management, industry, sovereign risk) that pure-quant Z does not — but Z is auditable, transparent, and free.

Finance

Altman Z-Score Calculator

Predict corporate bankruptcy risk using Edward Altman's original five-ratio model.

Working Capital ($)

Total Assets ($)

Retained Earnings ($)

EBIT ($)

Market Cap (Equity MV) ($)

Total Liabilities ($)

Sales / Revenue ($)

Altman Z-Score Formula (Public Manufacturing):
Z = 1.2×X1 + 1.4×X2 + 3.3×X3 + 0.6×X4 + 1.0×X5
X1=WC/TA, X2=RE/TA, X3=EBIT/TA, X4=MV Equity/TL, X5=Sales/TA

Edward Altman's Z-Score is the most widely cited bankruptcy-prediction model in finance — taught in every CFA curriculum, used by credit officers, and quoted in courtroom valuation disputes. The model is almost 60 years old and still works directionally because it captures something fundamental: distressed firms tend to share five financial signatures, and combining them in the right linear weights extracts a single risk number that beats most one-ratio screens.

Altman's 1968 study and its method

Working with 66 publicly-traded U.S. manufacturers — 33 that had filed bankruptcy between 1946-1965 and 33 healthy matched controls — Altman ran Multiple Discriminant Analysis on 22 candidate ratios. Five survived the statistical pruning. The discriminant function fit on the surviving five correctly classified 95% of bankrupt firms one year ahead of filing and 72% two years ahead — extraordinary out-of-sample accuracy for a model that costs five line items of financial statements to compute.

The five ratios and what they tell you

Ratio	Captures	Weight
X1 = Working Capital / Total Assets	Liquidity buffer	1.2
X2 = Retained Earnings / Total Assets	Cumulative profitability & firm age	1.4
X3 = EBIT / Total Assets	Current operating productivity	3.3
X4 = Market Equity / Total Liabilities	Market-implied solvency cushion	0.6
X5 = Sales / Total Assets	Asset turnover efficiency	1.0

X3 (EBIT/Assets) carries the largest coefficient — current earning power matters more than any single other factor for short-horizon default prediction.

Reading the zones

Z-Score	Zone	Interpretation
Z > 2.99	Safe	Default risk < 1% within 1 year
1.81 < Z < 2.99	Grey	Watch — could deteriorate
Z < 1.81	Distress	Material bankruptcy probability within 2 years

When to use Z, Z′, or Z″

Z (original): Public manufacturers. Uses market value of equity in X4.
Z′ (1983): Private manufacturers. Replaces market value with book value of equity. Coefficients shift slightly; cutoffs become 2.90 (safe) and 1.23 (distress).
Z″ (1995): Non-manufacturers and emerging-market firms. Drops X5 entirely. Cutoffs 2.60 and 1.10. This is the variant most appropriate for retailers, banks, technology firms, and services.

Worked example: $2M asset manufacturer

X1 = 500K/2M = 0.25 → contribution 0.30
X2 = 300K/2M = 0.15 → contribution 0.21
X3 = 400K/2M = 0.20 → contribution 0.66
X4 = 1.5M/800K = 1.875 → contribution 1.125
X5 = 3M/2M = 1.50 → contribution 1.50
Z = 0.30 + 0.21 + 0.66 + 1.125 + 1.50 = 3.795 → comfortably in Safe Zone

Modern critiques and adjustments

Three caveats every practitioner should keep in mind in 2026:

Capital structure shifts. Modern public companies hold more cash and more financial debt than 1968 firms. Z's coefficients reflect a different era. Altman's own 2017 update suggests easing the distress cutoff to ~1.23.
Intangible-heavy firms. Software, biotech, and platform companies have book values that radically understate true asset value. Z reads them as risky when they're not.
Accounting changes. Lease capitalization (ASC 842) raised total assets and total liabilities, mechanically shifting X1 and X4 without changing economic reality. Adjust pre/post for cross-year comparison.

FAQ

Is the Z-Score used by rating agencies?

Indirectly — Moody's KMV and S&P use proprietary structural and ratio models, but Z-style ratios are inside them. Z is a transparent, free benchmark for what the proprietary models capture.

Can Z-Score be used for banks?

No — bank balance sheets violate Z's underlying assumptions (working capital, sales, asset turnover don't translate). Use bank-specific models like CAMELS or stress tests.

Should I use trailing or forward EBIT?

Altman's original calibration uses last-12-month financials. For forward-looking analysis, run Z on consensus-forecast financial statements in addition to trailing.

How often should Z be updated?

Quarterly for credit monitoring of public firms. Annually for private companies based on audited financials. Real-time updates aren't meaningful — Z is a structural model, not a market signal.

What about emerging markets?

Altman's emerging market Z (Z'') is the appropriate variant; it adds a country-risk modifier in some implementations. For very low-disclosure jurisdictions, treat Z as one of several inputs, not a stand-alone screen.

Are there better modern models?

The Merton structural model (KMV) outperforms Z marginally on public firms because it uses option-pricing logic on equity volatility. Machine learning models can outperform Z on large samples but lose its interpretability. For most practitioners, Z remains the right baseline.

Related calculators

Sources

Educational only; not investment advice. Reviewed by Priya Venkatesan, CFA, on March 1, 2026.