Stocktoria
Do the scores work? A 15-year backtest
Companies with a high Piotroski F-score and a safe Altman Z historically outperformed weak-scoring ones. Growth of $1 by score bucket, 2011–2025, with the method and caveats.
Over 15 years, $1 spread equally across companies scoring a Piotroski F of 7–9 grew to 7.301× — versus 1.792× for the weakest (F 0–3) and 3.512× for the whole universe. The ranking is monotonic: stronger scores, higher returns.
Growth of $1 by score bucket
Average forward 1-year return by fiscal year · equal-weight
| Year | All (equal-weight) | Altman distress | Altman safe | Piotroski F 0-3 | Piotroski F 4-6 | Piotroski F 7-9 |
|---|---|---|---|---|---|---|
| 2011 | +13.4% | +12.4% | +13.2% | +7.1% | +17.5% | +13.5% |
| 2012 | +29% | +53.8% | +32.5% | +28.5% | +28.4% | +35.8% |
| 2013 | +9.1% | +3.2% | +12.1% | +7.7% | +9.1% | +16% |
| 2014 | -8.8% | -20.2% | -6% | -14.7% | -5.5% | -3% |
| 2015 | +22.9% | +20.2% | +20.9% | +21.3% | +23.8% | +23% |
| 2016 | +18.9% | +29.1% | +25% | +16% | +19.9% | +24.7% |
| 2017 | -1.1% | +0.4% | +1.2% | -5.4% | +1.2% | +1.3% |
| 2018 | +5.3% | +3.2% | +9% | +1.6% | +6.6% | +9.3% |
| 2019 | +22% | +50.1% | +23.8% | +31.8% | +16.8% | +11.3% |
| 2020 | +14.4% | +1.3% | +10.1% | +6% | +19.8% | +26.9% |
| 2021 | -12% | -18% | -11.8% | -20.8% | -8% | -1.9% |
| 2022 | -5.8% | -14.4% | -0.9% | -12.4% | -2.5% | +5.8% |
| 2023 | +12% | +8.8% | +13.9% | +4% | +15.1% | +19.5% |
| 2024 | +7.7% | +7.6% | +11.2% | +5.2% | +7.3% | +14.4% |
| 2025 | +13.6% | -5.7% | +18.1% | -1.2% | +17.1% | +23.7% |
| Avg / yr | 9.4% | 8.8% | 11.5% | 5% | 11.1% | 14.7% |
Method
For every company and fiscal year we recompute that year's Piotroski F and Altman Z″ zone, then measure the return from that fiscal year-end to one year later using split-adjusted monthly closes. Returns are equal-weighted, averaged within each bucket per year, and compounded into the curves above.
Caveats — read these
Illustrative, not a live strategy. Our fundamentals cover companies that are still listed, so this carries survivorship bias — delisted losers are missing, which overstates every bucket (especially the weak ones). Returns are gross and equal-weight, with no transaction costs, taxes or slippage, and rebalanced annually on an idealized date. Past results do not predict the future. This is evidence the scores have sorted winners from losers in our data — not a promise that they will.