Piotroski F-score screener: audit all nine signals

The nine tests and their denominators

Let t be the latest complete fiscal year, t-1 the preceding year, and A the total assets at each year end. Beginning assets for t are A(t-1); beginning assets for t-1 are A(t-2). Average assets for t are [A(t) + A(t-1)] / 2. Use the same reporting scope, currency and units across both years. NI means net income before extraordinary items; OCF means cash flow from operations. Every test below is worth one point when true and zero when false, provided its inputs are known.

• Definition choice: the Chicago Booth reprint describes beginning assets for turnover in section 2.3.3, while its Table 1 note says average assets. This worksheet follows section 2.3.3. Record that choice when reconciling a provider score.
• Boundary choice: the paper describes falling leverage as favorable but does not explicitly assign equality in that sentence. This worksheet requires a strict decline. Providers that accept unchanged leverage can differ by one point.
• Issuance evidence: an unchanged or declining share count can hide issuance offset by buybacks. Weighted-average diluted shares also change for reasons other than a common equity offering. Review the equity note instead of substituting either field.
• A reported net-income field is not automatically the historical NI definition. Reconcile extraordinary, discontinued and other adjustments in the filing before entering it. Keep the definition consistent across the two years.

Reproduce the three-company screen

The initial worksheet uses FY 2025 versus FY 2024, with a March 10, 2026 screen date. All monetary inputs are hypothetical USD millions. Year-end assets for 2023, 2024 and 2025 are 90, 100 and 110 for each company. Prior-year income, sales, cost of sales, debt, current assets and current liabilities are 5, 100, 65, 20, 45 and 30. The latest current input set was publicly available on March 2, 2026; the prior set on March 3, 2025. These are teaching assumptions, not actual company disclosures.

• Alder: ROA rises from 5/90 = 0.055556 to 12/100 = 0.12. CFO/assets = 0.15 exceeds 0.12. Leverage falls from 20/95 = 0.210526 to 18/105 = 0.171429.
• Alder: current ratio rises from 1.5 to 2.0, gross margin from 35% to 40%, and turnover from 100/90 = 1.111111 to 130/100 = 1.30. No issuance supplies the remaining point.
• Birch: positive earnings, positive cash flow and rising ROA supply three points. CFO is below income; leverage rises to 30/105 = 0.285714; liquidity, margin and turnover fall; issuance fails the seventh test.
• Cedar: missing OCF affects signals 2 and 4. Seven confirmed points and two unknowns imply a range of 7 to 9. Entering zero OCF would falsely convert uncertainty into two failures.

Availability dates prevent future-data leakage

A fiscal year end identifies the economic period; it is not the date investors could see the accounts. Record the original filing or release timestamp for each input and use the latest one as the input-set availability date. This worksheet uses dates only, so it requires both availability dates to be strictly earlier than the screen date. It cannot prove a user-supplied date or inspect a linked filing.

Missing, invalid and weak are different states

A complete score ranges from 0 to 9. A partial range runs from the number of confirmed passes to that number plus unknown signals. The filter admits only complete scores of at least 8. A partial upper bound below 8 can already rule a company out, but a lower bound of 8 still does not create a complete score. This conservative completeness rule is a worksheet policy.

• A negative valid income or cash-flow amount remains negative and can fail a signal. It is not missing. Zero at a strictly positive test also fails.
• Zero or negative assets, sales denominators or current liabilities make the affected ratios unavailable. Negative debt, current assets or cost of sales require reconciliation. No infinity or invented zero is scored.
• If current liabilities are zero, the current-ratio test remains unknown even if the business has no current debt. The denominator is all current liabilities, not only interest-bearing debt.
• A high score counts favorable historical changes; it does not measure valuation, fraud probability or return certainty. Banks, insurers and businesses without comparable gross profit or current balance-sheet classifications need a different review.

Use the candidate list in Pineify and TradingView

Pineify Market Screener supplies technical fields and saved symbol collections. Its current field registry does not include the nine accounting signals or a native F-score threshold. AI Finance Agent can request a provider financial score and annual financial statements by ticker. The provider score response does not expose the nine flags, their fiscal periods or their availability dates, so it cannot replace this audit.

• Preserve a candidate sheet with ticker, fiscal years, input values, source references, availability dates, nine flags, score range and inclusion reason. The local worksheet resets on reload; save your reviewed figures separately.
• The original research concerns high book-to-market stocks. Substituting a low-P/E universe or applying the score to every growth stock changes the experiment. This worksheet does not reproduce the historical portfolio or its returns.