What is Altman Z-Score: Corporate Insolvency Prediction, Financial Ratios, and Credit Risk?
Professor Edward Altman developed the Z-Score in 1968 at New York University using discriminant analysis on 66 manufacturing companies — half of which had filed for bankruptcy. He identified five financial ratios that, when combined with empirically derived weights, predicted bankruptcy two years in advance with 80–90% accuracy. The original model was specifically calibrated for publicly traded manufacturing companies. Over 50 years later, it remains the most widely used quantitative bankruptcy prediction tool in corporate finance, credit analysis, and distressed debt investing.
Mathematical Foundation
Altman Z-Score (Original Manufacturing Model)
= Liquidity ratio. Working Capital / Total Assets. Working Capital = Current Assets − Current Liabilities. Measures a company's short-term liquidity — can it meet near-term obligations? A negative WC (more current liabilities than assets) produces a negative X₁ contribution, sharply reducing the Z-Score. Weight: 1.2. A WC/TA of 0.30 contributes 0.36 to Z.
= Cumulative Profitability ratio. Retained Earnings / Total Assets. Retained Earnings are the cumulative undistributed profits reinvested in the business since founding. Young companies have low RE (haven't had time to accumulate profits). Mature profitable companies have high RE. A negative RE (cumulative losses exceed cumulative profits) is a major distress signal. Weight: 1.4. RE/TA of 0.20 contributes 0.28 to Z.
= Operating Efficiency / Return on Assets. Earnings Before Interest & Taxes / Total Assets. Measures how productively the company deploys its asset base to generate operating profit, before the effects of capital structure (interest) and taxes. This is the MOST HEAVILY WEIGHTED ratio (3.3×) because it directly measures operational health independent of financing decisions. A company with high debt but strong operations can still score well on X₃.
= Solvency ratio. Market Value of Equity / Total Liabilities. For publicly traded firms: MVE = shares outstanding × current stock price. For private firms, Altman's revised model (Z') substitutes book value of equity for MVE. Measures how much the equity market cushion protects creditors. A high MVE/TL means creditors could liquidate equity to pay debts. Weight: 0.6. MVE/TL of 2.0 contributes 1.20 to Z.
= Asset Turnover. Total Sales / Total Assets. Measures how efficiently the company generates revenue from its asset base. A capital-intensive manufacturer might have S/TA = 0.8 (1 dollar of assets generates 80 cents of revenue). A service business might have S/TA = 3.0+. This ratio captures the competitive and operational efficiency of the business. Weight: 0.999 (nearly 1.0). Note: This ratio varies significantly by industry — compare against industry peers for meaningful analysis.
Laws & Principles
- Zone Thresholds and Their Predictive Accuracy: Safe Zone (Z > 2.99): Original 1968 study correctly classified 95% of safe firms two years before outcomes were known. Companies in this zone misclassified as bankrupt were called 'Type II errors' — statistically rare. Grey Zone (1.81 ≤ Z ≤ 2.99): The 'zone of ignorance' — Altman found that firms in this range had mixed outcomes. The Z-Score is indeterminate here; additional qualitative analysis is required. Distress Zone (Z < 1.81): In the original study, 94% of companies with Z < 1.81 that subsequently went bankrupt were correctly predicted. The lower the Z-Score, the higher the probability and typically the shorter the time to default. Z < 1.23 often indicates near-term liquidity crisis rather than long-term structural problems.
- Limitations of the Original Z-Score Model: Asset-Heavy Bias: The original model was calibrated on manufacturing firms. Service companies, tech firms, and financial institutions have very different asset/revenue ratios (X₅) and equity structures (X₄), making the original thresholds unreliable. Altman addressed this with Z' (adjusted for private firms) and Z'' (for non-manufacturing firms). Market-to-Book Sensitivity: X₄ uses market value of equity. In a bull market, even financially marginal companies can have high MVE, inflating their Z-Score. In a bear market, Z-Scores compress across the board even for healthy firms. As-Reported Data: The model takes accounting data at face value. Aggressive revenue recognition, off-balance-sheet debt, or management of reported WC can mask true financial health. Audited financial statements reduce (but don't eliminate) this risk.
- Z' for Private Companies and Z'' for Non-Manufacturing: Z' model (1983) replaces MVE in X₄ with Book Value of Equity (total equity on the balance sheet): X₄' = Book Value Equity / Total Liabilities. Adjusted weights: Z' = 0.717×X₁ + 0.847×X₂ + 3.107×X₃ + 0.420×X₄' + 0.998×X₅. Thresholds shift: Z' > 2.90 = Safe, 1.23–2.90 = Grey, Z' < 1.23 = Distress. Z'' model (1995) eliminates the sales/assets ratio (X₅) which varies too much across industries: Z'' = 6.56×X₁ + 3.26×X₂ + 6.72×X₃ + 1.05×X₄'. Used for non-manufacturing and service companies. Safe Zone: Z'' > 2.60, Distress: Z'' < 1.10.
- How Creditors and Analysts Use the Z-Score in Practice: Credit Rating Crossover: A Z-Score below 2.0 often correlates with a sub-investment grade (below BBB) credit rating. Most Z < 1.81 companies have high-yield ('junk') bond ratings or no public rating. Distressed Debt Investing: Hedge funds specializing in distressed investing use the Z-Score as a first-pass screening tool. Firms with Z between 1.0 and 1.81 may have securities trading at significant discounts — these are candidates for restructuring plays. Lender Covenant Monitoring: Commercial banks sometimes include Z-Score floors in loan covenants (e.g., 'borrower shall maintain Z-Score ≥ 1.80'). Falling below the covenant triggers acceleration clauses or requires additional collateral. M&A Due Diligence: Acquirers calculate the target's Z-Score as part of financial due diligence to assess latent bankruptcy risk that may not be visible in headline revenue growth numbers.
- Historical Predictive Accuracy and Academic Validation: Altman's 1968 original study: 72% accuracy two years before bankruptcy. 94% accuracy correctly classifying bankrupt firms. Over 50 years of subsequent academic testing across multiple economic cycles has confirmed the model's utility, though accuracy varies by period and industry. 1990s out-of-sample testing: 85–92% accuracy. 2008 Financial Crisis: The model performed well for industrial firms but missed bank failures (which require the Z'' model or bank-specific models like the Merton structural model). The Z-Score has been tested and validated in 31 different countries, with culture-specific threshold adjustments recommended by Altman for non-US markets.
Step-by-Step Example Walkthrough
" A credit analyst evaluates a mid-size manufacturing company with: Working Capital = $1.2M, Retained Earnings = $0.8M, EBIT = $0.4M, Market Cap = $5.0M, Sales = $6.0M, Total Assets = $4.0M, Total Liabilities = $2.5M. "
- 1. X₁ = 1.2M / 4.0M = 0.300. Contribution: 1.2 × 0.300 = 0.360.
- 2. X₂ = 0.8M / 4.0M = 0.200. Contribution: 1.4 × 0.200 = 0.280.
- 3. X₃ = 0.4M / 4.0M = 0.100. Contribution: 3.3 × 0.100 = 0.330.
- 4. X₄ = 5.0M / 2.5M = 2.000. Contribution: 0.6 × 2.000 = 1.200.
- 5. X₅ = 6.0M / 4.0M = 1.500. Contribution: 0.999 × 1.500 = 1.499.
- 6. Z = 0.360 + 0.280 + 0.330 + 1.200 + 1.499 = 3.669.
Final Result: Z = 3.669 — Safe Zone (> 2.99). The company has strong liquidity (X₁ positive), reasonable retained earnings (X₂), healthy asset utilization (X₅ = 1.5), and most importantly a large market cap cushion (X₄ = 2.0). The analyst should note that if the stock price fell 60% (MVE dropped to $2M), X₄ would be 0.800, reducing Z to 2.149 — entering the Grey Zone. Monitoring market cap trends is the most sensitive leading indicator for this particular firm.