Bankruptcies announce themselves in the balance sheet long before the share price shows it: the Z-score condenses five ratios into a default probability.
Save the strategy and StockScorer checks its rules for you every day. You'll get a heads-up whenever a stock crosses the upper or lower score threshold.
MSCI World over the same period: +240.1 %
Past performance is not a reliable indicator of future results.
| Score | Number of stocks |
|---|---|
| -3 | 4,175 |
| -2 | 0 |
| -1 | 0 |
| 0 | 0 |
| +1 | 5,939 |
| +2 | 0 |
| +3 | 6,151 |
Every stock runs through the same disclosed rules. The points add up to a score, traceable down to the individual rule.
Each rule checks a metric against a threshold, for example ROE above 15 %.
You decide how much each rule counts: from +1 to +3 or −1 to −3.
The sum is the score. That gives three classes: high, medium, low match.
These terms describe only the match with the criteria, not a recommendation to buy or sell.
The Z-score translates five balance-sheet ratios into a 24-month default probability.
Manufacturers: above 2.99 safe, below 1.81 acute danger. Other non-financials: above 2.6 safe, below 1.1 acute danger (Z'' model).
Retained earnings weigh heavily: pure hope stocks without history fall through.
Financials run on balance-sheet leverage by design; they are classified neutrally instead of wrongly punished.
These exact rules run over every stock daily.
After a free sign-up: the full profile in the rule editor to copy, adjust thresholds and save as your own starting profile.
See & copy all rulesThe backtest shows a total return of +5.7%. The MSCI World reaches +240.1% over the same period.
Survivorship-free since June 30, 2015: the index composition is applied point-in-time.
Historical period, quarterly rebalancing, no taxes or fees. Past performance is not a reliable indicator of future results.
Purely mechanical rule application, no curated-list effect: stocks with a high match can fall just like any other stock.
For risk-aware investors as a pre-filter before any other strategy: before valuation or growth matter, survival should be secured. Particularly valuable for small caps and optically cheap stocks whose low price may reflect default risk.
The Z-score was developed by Edward Altman in 1968 and remains the best-known bankruptcy prediction model: a discriminant analysis condenses five balance-sheet ratios (working capital, retained earnings, operating earnings power, market value versus liabilities and asset turnover) into a single figure for the default risk of the next 24 months. Altman himself explicitly limits this 1968 model to manufacturing companies; for all other non-financials (trade, services, technology) he followed up in 1995 with the Z'' model, which drops the sales term. StockScorer therefore applies whichever model fits the sector: the classic Z-score for manufacturers, Z'' for the rest. Above the respective upper zone the safe zone begins, below the lower zone the distress zone with acute bankruptcy risk. Both models punish young companies without an earnings history hard: exactly what makes them excellent filters against hype stocks without substance. For banks and insurers, neither metric is structurally applicable; they are classified neutrally.
For manufacturing-adjacent sectors (industrials, basic materials, cyclical and defensive consumer, energy), StockScorer uses the classic Altman Z-score, weighted from five ratios: working capital to total assets (liquidity), retained earnings to total assets (accumulated substance), EBIT to total assets (earnings power), market capitalisation to total liabilities (market buffer) and sales to total assets (asset turnover). A value above 2.99 signals the safe zone (+3), 1.81 to 2.99 the grey zone (+1), below 1.81 the distress zone (−3).
For all other non-financial, non-real-estate sectors (trade, services, technology, health care), Altman's own 1995 Z'' model applies instead since the fidelity round, dropping the sales term, with its own zones: above 2.6 safe (+3), 1.1 to 2.6 grey zone (+1), below 1.1 distress (−3). The thresholds are deliberately identical for all size classes: a high bankruptcy risk is as unacceptable for a large cap as for a small cap (risk metric, universal). Financials, real-estate companies and stocks without a computable Z-score or Z'' park at +1 in the middle zone: not assessed is not the same as risky.
Manufacturers (classic Z-score): > 2.99 safe zone (+3), 1.81–2.99 grey zone (+1), < 1.81 distress zone (−3).
Other non-financials (Z'' model, 1995): > 2.6 safe zone (+3), 1.1–2.6 grey zone (+1), < 1.1 distress zone (−3).
Financial/real-estate company or Z-score/Z'' not computable: neutral middle classification (+1).
The strength: the Z-score is purely balance-sheet based and thus immune to story and sentiment; in studies it flagged a large share of bankruptcies one to two years in advance. As a gatekeeper before value strategies it prevents the classic mistake of investing in a "cheap" company whose price simply reflects its default risk.
Fidelity update as of 23 August 2026: Altman himself calibrated the 1968 model only for manufacturing companies; StockScorer now applies it only there and uses Altman's own Z'' model (1995, no sales factor, different zone boundaries of 2.6/1.1) for all other non-financials. Previously the classic Z-score ran sector-independently for all non-financials, which was systematically too conservative for services and technology names with lean balance sheets. One documented refinement remains: the grey zone earns +1 instead of 0 points here so it becomes visible as a middle classification; in the original point system it would be indistinguishable from the lower zone.
Limits: neither model is defined for financials (hence their neutral treatment). The sector assignment between the two models follows a fixed list of manufacturing-adjacent industries; edge cases (e.g. conglomerates) may be assessed more accurately under the other model.
For anyone wanting to systematically sort out default risks, as a standalone screen or as a risk gate flanking other strategies (the StockScorer Score uses exactly this gate as a knockout).