The original 1968 Z-Score targets public manufacturing firms and uses market value of equity in ratio D. Z' (1983) replaces market value with book value of equity so it works for private companies, with re-estimated coefficients. Z'' (1995) drops the Sales/Assets ratio and re-weights the others to apply to non-manufacturers and emerging-market firms, where asset turnover varies too much across industries.

Use it as a fast bankruptcy-risk screen for non-financial corporates with standard balance sheets: credit analysis, supplier due diligence, audit risk assessment, and equity screening. Pick the variant that matches the firm: original Z for public industrials, Z' for private industrials, Z'' for service, retail, or emerging-market companies. It is not designed for banks, insurers, or other financial institutions.

Altman's original sample showed roughly 95% accuracy one year before bankruptcy and about 72% accuracy two years before. Accuracy fades beyond two years, so the Z-Score is best treated as a one to two year early-warning indicator rather than a long-run forecast. A persistent slide through the grey zone is usually more informative than a single reading.

Replication studies since the 1990s generally find 70 to 80% accuracy on broad samples, which is strong for such a simple model but lower than the original 1968 result. Critics note that accounting standards, intangible-heavy business models, and changes in capital structure have reduced fit, especially for technology and asset-light firms. Most practitioners now use Z alongside cash-flow ratios, credit spreads, and qualitative review.

Avoid using Z on banks, insurers, broker-dealers, and other financial institutions because their balance sheets are dominated by financial assets and deposits that make the ratios uninformative. The original Z is also unreliable for service firms, retailers, and real-estate companies; use Z'' for those. Treat early-stage, pre-revenue, or restructuring firms with extra caution since the model assumes a stable going concern.

Altman's original specification uses market capitalization divided by total liabilities to capture how much equity cushion the market sees relative to debt. Market value is forward-looking and reflects investor expectations of future cash flows, which book value can miss. Z' substitutes book equity for private firms where no market price exists, accepting some loss of predictive power in exchange for applicability.

No. Z-Score is a statistical screen, not a guarantee. Even firms scoring above 2.99 can fail because of fraud, sudden liquidity events, regulatory shocks, or strategic mis-steps that the five ratios cannot capture. Treat a safe zone reading as a green light to keep monitoring, not as a clean bill of health, and revisit it whenever new financials are filed.