Where the F-Score comes from
Joseph Piotroski, an accounting professor, introduced the score in a 2000 paper in the Journal of Accounting Research. His starting point was the cheapest stocks by book-to-market ratio, the group value investors screen for when they look for stocks below book value. Many of those companies are cheap because they are in trouble. Piotroski asked whether simple accounting signals could separate the ones getting stronger from the ones getting weaker.
He chose tests that need nothing beyond the financial statements, each with a clear pass or fail. That simplicity is why the score still appears in screeners decades later.
The nine tests
| # | Test | Passes (scores 1) if | Group |
|---|---|---|---|
| 1 | Return on assets | Net income before extraordinary items, divided by beginning total assets, is positive | Profitability |
| 2 | Operating cash flow | Cash flow from operations is positive | Profitability |
| 3 | Change in return on assets | Return on assets is higher than last year | Profitability |
| 4 | Accruals | Operating cash flow divided by assets is higher than return on assets | Profitability |
| 5 | Change in leverage | Long-term debt to average total assets fell from last year | Leverage and liquidity |
| 6 | Change in current ratio | Current ratio rose from last year | Leverage and liquidity |
| 7 | Share issuance | No common equity issued in the year | Leverage and liquidity |
| 8 | Change in gross margin | Gross margin is higher than last year | Operating efficiency |
| 9 | Change in asset turnover | Sales divided by beginning total assets is higher than last year | Operating efficiency |
Profitability: tests 1 to 4
The first two tests ask whether the company makes money at all, on an accounting basis and in cash. The third asks whether profitability is improving. The fourth is the most interesting: it passes when cash flow exceeds reported earnings. When earnings run well ahead of cash, the gap is accruals, such as rising receivables, and earnings built on accruals tend to be lower quality. Our guide on how to read a cash flow statement explains where that gap shows up.
Leverage and liquidity: tests 5 to 7
These ask whether the balance sheet is getting safer. Less long-term debt relative to assets and a higher current ratio both pass. The share issuance test fails if the company sold new common stock during the year. For a distressed company, issuing equity often means it could not fund itself from operations, and it dilutes existing holders; see what dilution is.
Operating efficiency: tests 8 and 9
A rising gross margin suggests better pricing or lower costs. Rising asset turnover means the company is generating more sales from the same asset base. Together they show whether the business is working better, not just whether the accounts look better.
A worked example
That score says Company A's finances improved on almost every measure. It says nothing about whether 0.6 times book is the right price, whether book value is real, or what could happen next.
How investors use it
The F-Score works as a second filter on a value screen. You first find stocks that look cheap on book value, earnings or cash, then use the score to separate improving companies from deteriorating ones. Many of the stock screeners built for value investors include it as a filter, and it appears in broader research platforms too; see the 12 best stock research tools for individual investors for where.
It also pairs well with distress measures. The Altman Z-Score estimates bankruptcy risk from the current level of key ratios. The F-Score measures direction. A company with a low Z-Score and a high F-Score is still risky but moving the right way. One with both low is the case to approach with most caution.
| Situation | What the score suggests | What to check next |
|---|---|---|
| Cheap stock, score 8 to 9 | Fundamentals improving across the board | Whether the improvement is one-off |
| Cheap stock, score 0 to 2 | Deteriorating on most measures | Debt maturities, going-concern language |
| Cheap stock, score 4 to 6 | Mixed picture | Which tests failed and why |
| Expensive stock, high score | Strong and improving, but priced for it | Valuation, not the score |
Where this goes wrong
The score is mechanical, and mechanical rules miss context.
- It rewards change, not level. A company recovering from a bad year scores well even if it is still weak. A stable, highly profitable company with flat ratios can score lower.
- One-off items swing it. An asset sale can lift return on assets, cut debt and raise the current ratio at once, adding three points without any improvement in the business.
- Share issuance is blunt. A company issuing stock to fund an acquisition fails the same test as one issuing to stay alive. Stock issued for employee compensation can also trip it, depending on how the vendor defines the test.
- It does not fit financial companies. Banks and insurers have no gross margin or current ratio in the usual sense.
- It uses annual data. By the time the 10-K is out, the year is already months old. Check the latest quarter.
- It ignores price and events. A high score on an overvalued stock is still an overvalued stock, and a high score on a stock that is a value trap still leaves you in the trap.
How it fits with event-driven work
For special situations, the F-Score is a quick read on whether a cheap balance sheet is getting stronger or weaker going into an event, such as a refinancing, an asset sale or a strategic review. It can also flag where the next filing might surprise. It cannot tell you whether the event will happen or what the stock is worth afterwards. Dombey's published research uses balance sheet screens like this as a starting point before each catalyst; see the research archive. For screens that start even closer to the balance sheet, our guide on how to screen for net-net stocks covers one.
The bottom line
The Piotroski F-Score adds up nine pass-or-fail tests on profitability, balance sheet strength and efficiency to give a score from 0 to 9. It was designed to sort cheap stocks into improving and deteriorating companies, and it is most useful as a second filter after a value screen. It measures direction rather than quality, can be moved by one-off items and ignores price. Use it to decide where to read the filings more closely, not as a conclusion.