Fundamental scores — Piotroski and Graham

Ready-made scores exist in every tool. What changes here is the house rule: every component comes out in the open, with the account it came from, and the score is not published when the data does not add up. None of them is investment advice — they are syntheses of fundamentals, for you to check and judge.

Piotroski F-Score (0 to 9)

Created by Joseph Piotroski (2000) to separate, among "cheap" stocks, the ones that are improving from the ones that are merely falling. It is the sum of 9 YES/NO tests (1 point each), comparing the fiscal year with the previous one. It measures fundamental momentum improving, not absolute quality — an excellent company in a slowdown year may score 5; the number goes up when the operation improves relative to itself.

It only applies to companies on the common chart of accounts. Banks and intermediation insurers have no F-Score (their chart of accounts is a different one).

Profitability (4 points)

1. roa_positivo — net income / total assets > 0.
2. cfo_positivo — Cash from Operating Activities (DFC, account 6.01) > 0.
3. roa_crescente — ROA of the year > ROA of the previous year.
4. qualidade_lucro — CFO / assets > ROA (earnings are backed by cash, not just by accrual).

Leverage and liquidity (3 points)

1. alavancagem_caiu — long-term debt / assets fell versus the previous year.
2. liquidez_subiu — current ratio (current assets / current liabilities) rose.
3. sem_diluicao — the share count did not increase (there was no dilutive issuance).

Efficiency (2 points)

1. margem_bruta_subiu — gross margin (gross profit / revenue) rose.
2. giro_subiu — asset turnover (revenue / assets) rose.

Partial score declared. A criterion without data — typical of the 1st year of the series (no previous year to compare against) or with no DFC in the file — is left out and does not score; the response carries avaliados (how many of the 9 could be measured), with the flag parcial_N_de_9. We never guess the missing point.

Convention: readings above 7 usually indicate fundamentals in clear improvement; below 3, deterioration. Like any yardstick, it is a signal, not a verdict.

Graham criterion (cheap / expensive)

Benjamin Graham proposed, for the defensive stock, P/E ≤ 15 and P/B ≤ 1.5. The product of the two ceilings is 22.5 — hence the rule of thumb: the stock passes the Graham criterion when P/E × P/B ≤ 22.5.

We use the point-in-time P/E and P/B (price of the 1st trading session after the financial statements were published — see multiplos.md), from the most recent fiscal year with both available. aprovado: true means only "it passes Graham's cheapness yardstick" — not that the company is good, nor that today's price is still there. A premium growth company (WEG, for example) fails by definition, and that is expected.