Piotroski F-Score: 9 Financial Tests
Table of Contents
The Piotroski F-Score gives investors a structured way to examine a cheap-looking company’s financial condition. A low price relative to book value can accompany a recovering business or a deteriorating one. The score asks nine specific questions about profitability, cash generation, financing and operating efficiency before the investor goes further.
Calculating it takes more care than checking nine familiar ratios. The original method uses different asset denominators for different tests, compares two annual periods, and examines common-equity issuance rather than simply whether shares outstanding increased. This guide follows the original paper’s detailed definitions, identifies two methodological ambiguities, and calculates a complete hypothetical score from the underlying inputs.
What Is the Piotroski F-Score?
Joseph Piotroski introduced the F-score in a paper published in the Journal of Accounting Research, Supplement 2000. Each favorable accounting signal earns one point; the nine points are added without different weights. The resulting integer ranges from 0 to 9. Four tests concern profitability and cash flow, three concern leverage, liquidity and equity financing, and two concern operating efficiency. [1]
A higher F-score means more of the specified signals are favorable. It does not measure intrinsic value, the percentage chance of avoiding bankruptcy, or the probability that a stock will outperform.
The original research concerned high book-to-market stocks: companies with substantial accounting book equity relative to the market value of their equity. With matching definitions and positive book equity, high book-to-market corresponds to low price-to-book. The F-score helped distinguish financial strength within that value-stock setting; it did not replace the initial valuation screen.
This article uses a U.S. reporting perspective and a nonfinancial operating-company example. Broader financial ratio analysis supplies the individual concepts, while the F-score imposes a particular nine-test checklist.
Which Financial Statements Do You Need?
Let t be the most recent complete fiscal year and t−1 the preceding year. Annual income and sales describe flows over a year; balance-sheet assets and liabilities describe balances on a date. The SEC’s financial-statement guide explains that distinction and why the statements and notes must be read together. [2]
Collect two annual flow periods and total assets at three consecutive year-ends. For example, scoring 20X2 requires 20X2 and 20X1 income and sales, plus assets at the ends of 20X0, 20X1 and 20X2. The earliest asset balance is needed because 20X1’s beginning assets are the assets reported at the end of 20X0.
| Input | Periods or dates required | Where to look |
|---|---|---|
| Income, net sales and cost of goods sold | Years t and t−1 | Annual income statements and related notes |
| Operating cash flow | Year t | Operating section of the cash-flow statement |
| Total assets | Year-ends t, t−1 and t−2 | Comparable balance sheets, including the earlier annual report if necessary |
| Current assets, current liabilities and long-term debt | Year-ends t and t−1 | Balance sheets and debt notes; include current maturities of long-term debt in the debt input |
| Common-equity issuance | During year t | Cash-flow financing disclosures, equity statement and notes describing issuance and repurchases |
Use consistent fiscal periods, currency units, accounting policies and consolidation boundaries. Operating cash flow is neither free cash flow nor the total change in cash. An unavailable input also differs from a reported zero: an incomplete checklist should remain incomplete rather than becoming an apparently complete score.
Define Income and the Asset Denominators
The original paper specifies net income before extraordinary items. That is a historical accounting label: FASB’s 2015 update eliminated extraordinary-item classification from U.S. GAAP. Document how the actual statement line maps to the historical input, particularly when discontinued operations or unusual transactions matter. Do not substitute EBITDA, adjusted earnings or earnings per share without identifying a different implementation. [1] [3]
Let At mean total assets at year-end t. Beginning assets for year t are therefore At−1. Average assets for that year are (At−1 + At) / 2.
CFO ratiot = Operating cash flowt / At−1
Our general return on assets guide uses average assets for its ordinary ratio discussion. For this named score, retain the beginning-asset definition above. A familiar ratio name does not make the denominator interchangeable.
The Nine Piotroski F-Score Criteria
The checklist below follows Sections 2.3.1–2.3.3 of the original paper. Each row contributes either one point or zero. For the ordinary calculations shown here, assets, sales and current-liability denominators must be positive and the required data must be available. [1]
| Test | One-point condition | What it examines |
|---|---|---|
| 1. Positive profitability | ROAt > 0 | Positive income relative to beginning assets |
| 2. Positive operating cash flow | CFO ratiot > 0 | Cash generated by operations |
| 3. Improving profitability | ROAt > ROAt−1 | A better accounting return than the prior year |
| 4. Cash flow exceeds income | CFO ratiot > ROAt | Operating cash flow exceeds the specified income measure |
| 5. Declining leverage | Long-term debt / average assets is lower than in t−1 | Less long-term debt relative to the asset base; debt includes its current portion |
| 6. Improving liquidity | Current assets / current liabilities is higher than at the prior year-end | A rising current ratio |
| 7. No common-equity issuance | No common equity was issued during year t | Reliance on new common-equity financing |
| 8. Improving gross margin | (Sales − COGS) / sales is higher than in t−1 | More gross profit per sales dollar |
| 9. Improving asset turnover | Sales / beginning assets is higher than in t−1 | More sales per dollar of beginning assets |
The cash-versus-income test uses the same positive asset denominator on both sides, so it is equivalent to comparing the dollar amounts of operating cash flow and specified income. Do not compare a cash-flow dollar amount directly with a percentage ROA. The original accrual construction keeps depreciation within the earnings-to-cash comparison rather than removing it as a separate adjustment.
Strict Inequalities, Ties and Two Source Ambiguities
Zero is not positive, and unchanged is not improved. A zero ROA or CFO ratio earns no point; unchanged ROA, current ratio, gross margin or turnover earns no improvement point. If operating cash flow equals income, test 4 earns zero. Compare unrounded values so display rounding does not create a false tie.
For leverage, the paper describes a falling ratio as favorable and a rising ratio as unfavorable without explicitly assigning exact equality. This guide uses a strict-decline convention: unchanged leverage earns zero. That includes a company with zero long-term debt in both years; its missing improvement point does not mean being debt-free is economically bad. State the convention when comparing scores.
There is also an inconsistency in the original PDF: Section 2.3.3 defines asset turnover using beginning assets, while Table 1’s turnover footnote says average assets. This guide follows the detailed method section and uses beginning assets throughout. A provider using the other denominator may obtain a different signal. This disclosed choice makes the example reproducible; it does not establish the contents of the original author’s implementation code. [1]
A Worked Piotroski F-Score Example
Consider Northfield Tools, a fictional U.S. manufacturer. The two fiscal years below each cover twelve months on the same reporting basis. Amounts are hypothetical USD millions, except shares, which are in millions of shares. Total assets at the end of 20X0 were 800, the beginning balance for 20X1.
There are no extraordinary or discontinued items, so the supplied income is the specified income input. There are no preferred shares, noncontrolling interests, stock splits, conversions, dividends or other equity movements. The table is a selected statement extract, not a complete set of financial statements.
| Input | 20X1: prior year | 20X2: current year |
|---|---|---|
| Net sales | 960 | 1,300 |
| Cost of goods sold | 672 | 884 |
| Specified income / net income | 40 | 60 |
| Operating cash flow | 50 | 90 |
| Current assets | 300 | 330 |
| Other assets | 700 | 770 |
| Total assets | 1,000 | 1,100 |
| Current liabilities | 200 | 220 |
| Current portion of long-term debt, already within current liabilities | 30 | 20 |
| Noncurrent long-term debt | 240 | 232 |
| Other noncurrent liabilities | 110 | 168 |
| Book equity | 450 | 480 |
| Common shares outstanding, millions | 10 | 9 |
The balance sheets reconcile. For 20X1, assets are 300 + 700 = 1,000, liabilities are 200 + 240 + 110 = 550, and 550 + 450 = 1,000. For 20X2, assets are 330 + 770 = 1,100, liabilities are 220 + 232 + 168 = 620, and 620 + 480 = 1,100. Do not add the current debt portion to those liability totals again.
Calculate Each Point
Begin with the asset bases: 20X1 uses 800 beginning assets and (800 + 1,000) / 2 = 900 average assets. For 20X2, the corresponding amounts are 1,000 and (1,000 + 1,100) / 2 = 1,050.
During 20X2, Northfield issued 1 million common shares for 20 and repurchased 2 million shares for 50. Shares therefore fell from 10 + 1 − 2 = 9 million. Equity also reconciles: 450 + 60 income + 20 issuance − 50 repurchases = 480. These gross issuance facts determine test 7.
| Test | Calculation and comparison | Points |
|---|---|---|
| 1. Positive ROA | 60 / 1,000 = 6% > 0 | 1 |
| 2. Positive CFO | 90 / 1,000 = 9% > 0 | 1 |
| 3. Improving ROA | 6% > 40 / 800 = 5% | 1 |
| 4. CFO exceeds income | 9% > 6%, equivalently 90 > 60 | 1 |
| 5. Declining leverage | (232 + 20) / 1,050 = 24% < (240 + 30) / 900 = 30% | 1 |
| 6. Improving current ratio | 330 / 220 = 1.50, equal to 300 / 200 = 1.50 | 0 |
| 7. No common-equity issuance | Common shares were issued, despite the net share-count decline | 0 |
| 8. Improving gross margin | (1,300 − 884) / 1,300 = 32% > (960 − 672) / 960 = 30% | 1 |
| 9. Improving turnover | 1,300 / 1,000 = 1.30 > 960 / 800 = 1.20 | 1 |
| F-score | 1 + 1 + 1 + 1 + 1 + 0 + 0 + 1 + 1 | 7 / 9 |
The company earns all four profitability points, one of the three financing/liquidity points, and both efficiency points: 4 + 1 + 2 = 7. ROA improves by 1 percentage point, leverage declines by 6 percentage points, and gross margin improves by 2 percentage points. The unchanged current ratio and actual equity issuance explain the two missing points.
The useful follow-up is to investigate those events. Was the equity raised for an attractive investment or to cover a funding shortfall? Why did issuance and repurchases occur in the same year? A net share-count field alone would conceal the first question. The original paper identifies offerings from financing statements; it does not define this point as merely “shares did not increase.” [1]
What Is a Good F-Score?
The original paper’s main comparison called 8 or 9 high scores and 0 or 1 low scores, within its high-book-to-market sample. Scores between those groups contained more mixed signals. Northfield’s 7 therefore falls below the original main high-score group, even though seven of its nine conditions are favorable. These group definitions are research classifications, not universal buy and sell rules. [1]
The study used 14,043 firm-year observations from 1976 through 1996, drawn from the highest book-to-market quintile using prior-year distribution cutoffs. In Table 3, the high-score group’s mean one-year market-adjusted return was 13.4%, compared with 5.9% for the complete high-book-to-market sample: a historical difference of 13.4% − 5.9% = 7.5 percentage points. That is a comparison of historical portfolio means, not a promised annual return or a current expected premium. [1]
Results were particularly associated with smaller, less actively traded and less followed firms. The paper also recognizes potential data-snooping effects on out-of-sample performance. Those characteristics, changing markets and implementation costs matter when someone attempts to turn an academic result into an investment strategy. A favorable group average never establishes that an individual high-score stock will succeed.
F-Score vs P/B, Altman Z-Score and Beneish M-Score
These tools can complement one another because they organize different questions. Keep their inputs and interpretations separate rather than treating several favorable labels as independent proof of investment quality.
| Tool | Main task | What it does not establish |
|---|---|---|
| Piotroski F-score | Count specified favorable historical financial signals | Intrinsic value, a fair share price or a calibrated outcome probability |
| Price-to-book | Compare market equity with accounting book equity | Whether a low multiple accompanies recovery or deterioration |
| Altman Z-Score | Screen for financial-distress characteristics using an identified model | A raw score’s current bankruptcy probability or an attractive purchase price |
| Beneish M-Score | Screen statement characteristics associated with potential earnings manipulation | Proof of fraud or assurance that unflagged accounts are reliable |
Ordinary fundamental analysis then connects the statements to business economics, competitive position, obligations and valuation. The F-score’s cash-versus-income point adds one accounting check; it does not perform the full reporting-quality investigation taught by the Beneish comparison.
How to Use the F-Score in a Value Screen
- Define the stock universe. Record the market, reporting basis, sector treatment and valuation screen. Reproducing the original high-book-to-market approach differs from ranking every stock by F-score alone.
- Fix the information date. Use filings and disclosures actually available then. The original paper began measuring returns at the start of the fifth month after fiscal year-end to allow annual information to become available; that historical convention is not a guarantee that every current filing is available on that schedule.
- Reproduce the inputs. Retain filing dates, statement lines, the three asset dates, debt maturities, gross equity events and the denominator/tie conventions. Reconcile any data-provider figure to the filings before relying on a score.
- Inspect the nine outcomes. Explain which signals changed and by how much. A liquidity point gained through inventory accumulation deserves different scrutiny from one supported by cash collection.
- Research the investment separately. Examine cash needs, refinancing, industry conditions, management decisions and price. The broader value investing framework explains why a screen is the start of an investigation.
For a historical backtest, later restatements must not silently replace the information available at the original selection date. Keep the original public data and later corrections distinct. For current analysis, a restatement can be important new evidence, but it cannot retroactively become something an investor knew earlier.
Common Mistakes When Calculating the F-Score
- Using one asset denominator everywhere. This guide’s ROA, CFO ratio and turnover use beginning assets; leverage uses average assets. State the turnover source discrepancy rather than silently mixing conventions.
- Collecting only two year-end asset balances. The prior year’s beginning-assets calculations require the third, earlier balance.
- Scoring a tie as an improvement. Preserve strict tests and the disclosed leverage tie convention; do not classify rounded display values.
- Using lower shares outstanding as proof of no issuance. Repurchases can offset issuance. Diluted weighted-average shares also differ from actual gross equity events.
- Confusing long-term debt with all liabilities. Include the current portion of long-term debt in leverage, but do not substitute current liabilities plus all noncurrent liabilities.
- Replacing missing data with zero. Do not normalize an incomplete checklist to nine or present an undefined ratio as an observed unfavorable signal.
- Treating a high score as a valuation conclusion. The nine points contain no share-price input; an expensive stock can receive a high score.
Limitations of the Piotroski F-Score
Any strictly positive qualifying improvement earns the same one point, whether tiny or substantial. A current ratio that rises from 1.50 to 1.51 earns the liquidity point; the count alone does not describe the size or durability of the change.
Accounting inputs require judgment. Acquisitions, asset impairments, divestitures, changing cost classifications and reporting policies can alter several ratios together. A smaller asset base can raise turnover without stronger customer demand. Operating cash flow can reflect working-capital timing, and its excess over income is not proof that every reported number is reliable.
Sector and life-cycle differences matter. Banks, insurers and other businesses with different statement structures need sector-specific analysis; a conventional gross-margin/current-ratio checklist may not describe their economics well. Growth companies can issue equity to finance useful projects. Cyclical businesses can earn favorable points during a temporary rebound. These are limits on interpretation, not a claim that the original paper explicitly excluded every such sector.
The signals are context dependent. Piotroski selected them for financially strained value firms. Lower leverage, no new equity and improving liquidity can be sensible favorable signs there, while the same financing decision can have a different meaning elsewhere. The paper itself discusses that conditional interpretation. [1]
Historical statements lag events. A company can face a new cash shortfall, lawsuit, loss of customers or financing disruption after its latest year-end. Undefined ratios, nonpositive required denominators or insufficient history also prevent the ordinary complete calculation. Retain those limits rather than manufacturing a precise score.
Frequently Asked Questions
Sources and Methodology
- Joseph D. Piotroski, Value Investing: The Use of Historical Financial Statement Information to Separate Winners from Losers, Journal of Accounting Research 38, Supplement 2000. Original paper hosted by Ivey Business School; Sections 2.3 and 3, Table 1 definitions, Table 3 results, financing footnote 16 and conclusion.
- U.S. Securities and Exchange Commission, Beginners’ Guide to Financial Statement, February 4, 2007. Statement timing, operating and financing cash flows, notes and MD&A.
- Financial Accounting Standards Board, Accounting Standards Update 2015-01, January 2015. Summary and relevant amendments eliminating extraordinary-item classification.
Sources accessed November 15, 2026. Northfield Tools and all its figures are hypothetical. The calculation uses the original detailed method section’s beginning-assets turnover definition and a disclosed strict-decline leverage convention. It is an educational calculation, not a replication of every empirical data-processing step or a current issuer rating.
Disclaimer
This article is for educational and informational purposes only and does not constitute investment or accounting advice. Historical screens can miss material risks and do not guarantee investment performance. Review original filings and consult a qualified professional about your circumstances.
