Methodology

The 5-pillar rating

A single Buy/Hold/Sell verdict from 0–100, computed as a weighted average of five pillar scores, each itself 0–100.

Formula

Details

• •Quality (30%) — return on equity, return on assets and net margin.
• •Financial Health (25%) — current ratio, debt-to-equity and interest coverage.
• •Growth (20%) — year-over-year change in return on assets and in asset efficiency.
• •Value (15%) — 60% from the company's valuation ratios against its sector median, 40% from the spread between price and our intrinsic-value estimate.
• •Momentum (10%) — 52-week and 3-month price return.
• •Dividing by the weights actually used means a company we can only score on three pillars is not silently penalised for the two we could not compute. It is instead marked down in confidence.

Bands

• •75 and above — Strong Buy
• •60 to 74.9 — Buy
• •45 to 59.9 — Hold
• •30 to 44.9 — Sell
• •Below 30 — Strong Sell

Confidence

Confidence is reported as HIGH when all five pillars scored, MEDIUM when three or four did, and LOW below that. Treat a LOW-confidence rating as a prompt to read the filing yourself, not as a weak signal.

Limitations

• ◆The weights are a judgement, not a discovered truth. A different reasonable investor would weight value higher and momentum lower, and would get a different answer from the same data.
• ◆Momentum is price-derived, so it imports market sentiment into what is otherwise a fundamentals score. It is weighted lowest for that reason.
• ◆The Value pillar's sector-relative half requires a sector median built from at least five peers. Where a company has no meaningful peer set, that half is unavailable and the intrinsic-value signal carries the pillar alone.
• ◆The rating says nothing about what you paid, your tax position, your time horizon or your concentration. It cannot: it has never met you.

Altman Z-Score — distress risk

A bankruptcy-risk screen published by Edward Altman in 1968, combining five balance-sheet and earnings ratios into one number.

Formula

Bands

• •Above 2.99 — 'safe' zone
• •1.81 to 2.99 — 'grey' zone
• •Below 1.81 — 'distress' zone

Limitations

• ◆It was derived from manufacturing companies. Applied to a bank or an insurer it is meaningless — their balance sheets are structurally leveraged and the working-capital term does not describe anything real. We show an explained N/A rather than a misleading number for those filers.
• ◆It is also weak for asset-light businesses, where 'total assets' understates the productive base, and for pre-revenue companies, where the sales term collapses.
• ◆The coefficients are from 1968 and have never been re-fitted. Treat the zones as a rough triage, not a probability.

Piotroski F-Score — fundamental momentum

A 0–9 score from Joseph Piotroski (2000). Each of nine yes/no tests of profitability, leverage and operating efficiency scores one point if the company passes.

Formula

Bands

• •8–9 — strong on all three dimensions
• •4–7 — mixed
• •0–3 — weak

Limitations

• ◆It needs a prior fiscal year on file. A recent IPO or a newly covered filer will show N/A rather than a low score — absence of data is not a failing grade.
• ◆Piotroski designed it to separate winners from losers *within* a basket of already-cheap stocks. A high F-Score on an expensive company is a much weaker signal than the number suggests.
• ◆Every test is year-over-year, so it rewards improvement from a low base as much as sustained excellence.

Beneish M-Score — earnings-manipulation indicators

An eight-variable model from Messod Beneish (1999) that flags accounting patterns statistically associated with earnings manipulation.

Formula

Bands

• •Above -1.78 — exhibits characteristics common among manipulators
• •Below -1.78 — does not

Limitations

• ◆This is the one to be most careful with. A score above the threshold does NOT mean a company has committed fraud, and we do not say that it has. It means the company's accounting ratios moved in a pattern that, across Beneish's sample, occurred more often among manipulators. Fast-growing and acquisitive companies trip it routinely and innocently.
• ◆It requires two consecutive years of comparable data and is not meaningful for financial companies, where the accrual and receivables terms do not describe the business.
• ◆Use it as a reason to read the filing more carefully, never as a conclusion.