The score for stocks
Eight weighted categories become a transparent score from 0 to 100, with reasoning.
- Fundamentals83
- Growth73
- Profitability98
- Valuation70
- Technical analysis81
- Momentum65
- Stability81
- Capital allocation87
Every metric earns 0 to 100 points through disclosed thresholds, and every category gets a grade with its reasoning. Weighted together, the eight categories make up the score.
Example: Apple
Apple Inc. achieves a MEQIRI Score of 78/100 (Strong). Its standout strengths are fundamentals and profitability, while valuation holds the score back. The price sits about 64% above fair value, so there is no margin of safety.
Valuation signal
Significantly overvaluedIn our model the price counts as cheap below $180.49 and expensive above $270.74 (uncertainty low, based on the price swings of the past year).
The two parts of the fair value come to $197.75 (cash-flow model) and $235.43 (industry comparison), and the gap between them shows how uncertain the estimate is. The price is above both.
Calculated with a cost of equity of 9.7%. At 10.7% the fair value would be $201.85, at 8.7% it would be $237.38.
Fundamentals are strong, with stable profits and strong cash generation.
All thresholds
Every metric is mapped onto 0 to 100 points through anchor points. Between two anchor points the score moves in a straight line, below the first and above the last their points apply. If a metric is not measured, it drops out and the others share its weight. The exceptions are the growth rates of revenue, earnings per share, free cash flow and dividend: if no rate can be measured there, it counts at 50 points. If everything in a category is missing, the category counts at 50 points. Financials are banks, insurers, brokers, asset managers and lenders. Some metrics say nothing there, so others count instead. The tables are built from the same numbers the score calculates with.
Fundamentals16% of the score
Profitable years and years with positive free cash flow over the fiscal years available, from three years on. Cash conversion counts only with a profit.
| Metric | Weight | Financials | Value and points | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Years with a profit in % | 30% | 55% |
| ||||||||||
| Cash conversion (free cash flow to earnings) in %, only from a 1% net margin | 25% | not used |
| ||||||||||
| Years with positive free cash flow in % | 20% | not used |
| ||||||||||
| Return on invested capital (ROIC) in % | 25% | not used |
| ||||||||||
| Return on equity (ROE) in % | not used | 45% |
|
Growth16% of the score
Growth per year over up to five years, up to and including the last twelve months. It is scored only from three year-on-year comparisons, before that growth counts at 50 points. If a series slipped into a loss, the first anchor point applies. Out of a loss no rate can be measured, it then counts at 50 points. For European banks revenue does not count, because it is reported differently from year to year.
| Metric | Weight | Financials | Value and points | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Revenue in % per year | 40% | 53% |
| ||||||||||||
| Earnings per share in % per year | 35% | 47% |
| ||||||||||||
| Free cash flow in % per year | 25% | not used |
|
Profitability16% of the score
Measured over the last twelve months. If a value is above 1.5 times the median of prior years, the part above counts only at 50% (from three prior years). Return on equity counts only with positive equity.
| Metric | Weight | Financials | Value and points | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Net margin in % | 32% | not used |
| ||||||||||||
| Return on invested capital (ROIC) in % | 28% | not used |
| ||||||||||||
| Return on equity (ROE) in % | 22% | 100% |
| ||||||||||||
| Gross margin in % | 18% | not used |
|
Valuation8% of the score
For the points, fair value is compared with the price, relative to the price. A plus means the price is below fair value. (The valuation signal shows the gap to fair value the other way round: there a plus means the price is above it.) For the points, the distance is multiplied by 20 and divided by the stock's threshold. The threshold depends on last year's volatility and is 20, 30, 40 or 50%. For a calm stock the distance counts as it is, a volatile one needs a larger distance for the same points. If the two parts of the fair value, the cash-flow part and the industry part, are more than a factor of two apart, at least 40% applies even when the price was calm, for mortgage REITs always, because their fair value is book value alone. For equity REITs the fair value uses FFO instead of earnings, for US REITs following the industry rule from the filings with the SEC and at most 1.15 times operating cash flow. For equity REITs the cash-flow model uses AFFO, so new construction and acquisitions do not count as running costs and preferred dividends are deducted; if the first revenue estimate differs from revenue by more than a factor of 1.25, it runs without estimates. Negative multiples, a P/E below 3, a P/E distorted by one-off items and a P/E after a one-time earnings collapse (trailing earnings below half of the prior year and below a third of what analysts expect) do not count, PEG only with a profit.
| Metric | Weight | Financials | Value and points | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Fair value versus price in % | 40% | 62% |
| ||||||||||||
| P/E | 25% | 38% |
| ||||||||||||
| PEG | 20% | not used |
| ||||||||||||
| EV/EBITDA | 15% | not used |
|
Technical analysis11% of the score
Uptrend means the price is above the 50-day line and that line is above the 200-day line. Downtrend is the opposite, everything else counts as sideways. With fewer than 200 trading days, trend and distance do not count.
| Metric | Weight | Value and points | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Trend | 35% |
| ||||||||||||
| Distance to the 200-day line in % | 30% |
| ||||||||||||
| RSI (14 days) | 20% |
| ||||||||||||
| MACD | 15% |
|
Momentum11% of the score
Price change without dividends. A window the price history is too short for does not count.
| Metric | Weight | Value and points | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Price 1 month in % | 25% |
| ||||||||||
| Price 6 months in % | 35% |
| ||||||||||
| Price 12 months in % | 40% |
|
Stability11% of the score
Debt to equity does not count with negative equity. Volatility is annualised from daily closing prices.
| Metric | Weight | Financials | Value and points | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Debt to equity | 30% | not used |
| ||||||||||||
| Leverage (assets per unit of equity) | not used | 67% |
| ||||||||||||
| Interest coverage (earnings per unit of interest) | 25% | not used |
| ||||||||||||
| Current ratio | 20% | not used |
| ||||||||||||
| Annual volatility in % | 25% | 33% |
|
Capital allocation11% of the score
The change in share count runs from the first comparable fiscal year. Dividend growth comes from the individual payments without special dividends: today's annual dividend against the annual dividend on the same day five years ago. If a company only resumed paying after a pause in the twelve months before that day, the run rate counts then and now, that is the latest payment times the number of payments per year. For annual and semi-annual payers these are the last one or two payments, so one annual dividend, even when two annual dividends came shortly after each other after the pause. With no dividend, or a dividend cut within the last three years, fixed points apply. A dividend also counts as cut when no payment has come for 450 days, or for 230 days at companies that last paid at least four times a year. A company that never paid regularly, only special dividends, counts as paying no dividend. If dividend growth cannot be measured, for example because no dividend was paid five years ago, it counts at 50 points.
| Metric | Weight | Financials | Value and points | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Return on invested capital (ROIC) in % | 40% | not used |
| ||||||||||||||
| Return on equity (ROE) in % | not used | 40% |
| ||||||||||||||
| Change in share count in % | 40% | 40% |
| ||||||||||||||
| Dividend growth in % per year | 20% | 20% |
|
Score for ETFs
An ETF does not get a score of its own but the score of its holdings. That shows you how strong the companies are that you buy with the fund.
- The ETF's score is the average score of its holdings, weighted by their share of the fund.
- Holdings without a score, such as stocks from the UK and Australia or smaller companies from Japan and Canada, do not count. If less than half of the fund is rated, the ETF has no score.
- Synthetic ETFs hold a basket of other securities as collateral. They get the score of their index, computed from the largest physical ETF on the same index. If there is none, they have no score.
- Only equity and real estate ETFs have a score. Bond, money market and commodity ETFs do not.
- Costs and returns are not part of the score; they are shown separately.
- Crash risk counts, over ten years of history with reinvested distributions, how often the ETF fell by 30% or more within 12 months.
Custom weighting
With Pro you decide how much each of the eight categories counts, for example without the price trend if you invest long term based on fundamentals, or with growth counting double.
The grades stay the same
Every category is measured the same way for everyone, with the same metrics and thresholds. You do not change what is measured, only how much it counts in your score.
Eight sliders add up to 100%
Move one slider up and the others shrink in the same proportion. Momentum at zero takes the price movement out entirely, and more weight on valuation makes expensive stocks stand out clearly.
Your score applies wherever it matters to you
Once saved, stock pages, your watchlist, the screener and your morning email use your score, with the official one alongside. Public pages and the sector comparison keep showing the official score, which stays the same for everyone and is therefore comparable.
Try it
Move the sliders or pick an example. The table recalculates instantly, using the real category grades from the nightly run. The table shows 161 large US stocks from the S&P 500, not every stock in the screener. The ranks apply within this selection only.
Examples
Your weights
Always 100% in total. Move one slider up and the others shrink in the same proportion.
With MEQIRI Pro you save your weighting, and stock pages, watchlist, screener and morning email then use your score.
See plansChanges to the method
Method as of version s29 from 7 Nov 2026, produced by SSM Ventures UG (haftungsbeschränkt). After every change we recalculate the score history with the new method so that comparisons over time hold.
| Date | Version | Change |
|---|---|---|
| 7 Nov 2026 | s29 | The formulas stay the same, the data are checked more strictly. For depositary receipts (ADRs), the number of ordinary shares reported to the SEC applies, divided by the ratio per receipt. If the data provider has mixed up receipt and share, market cap, share count and earnings per share are converted. The market cap of the Hong Kong listing no longer replaces that of the receipt, because it often counts only one share class there. If a company now pays once a year, only the latest payment counts for the dividend yield, even if the ex-date moves earlier. A higher payment reported as a regular dividend counts as an increase until the next payment shows whether it stays. Examples: Autohome had a P/E of 1.3 instead of about 25, IRSA a dividend yield of 24.1% instead of 12.9%. |
Show earlier changes
| Date | Version | Change |
|---|---|---|
| 6 Nov 2026 | s28 | The formulas stay the same, the data are checked more strictly. If a quarter contains the whole previous fiscal year, revenue and profit from the SEC apply. A profit that includes non-controlling interests becomes the profit attributable to shareholders, even if the data provider has no earnings per share. A cash flow statement a thousand times too high is converted, and a fresh quarter whose profit does not match the reported earnings per share is calculated with the earnings per share. A quarter listed as a fiscal year does not count as a year. Revenue growth is not measured if it spans a year without revenue or differs clearly from the SEC figure. For mortgage REITs, EV/EBITDA does not count. If no valuation metric is measured, the valuation light shows ‘not measured’ and valuation does not count in the score. Examples: Kyndryl had a P/E of 10.7 instead of about 46, Virtu of 3.7 instead of about 8, Barings BDC revenue growth of 70.7% instead of 31.5% a year. |
| 5 Nov 2026 | s27 | The formulas stay the same, the data are checked more strictly. A fresh balance sheet that the data provider takes over incompletely on the reporting day, for example with equity of 0 or without total assets, only counts with the complete report; until then the last complete balance sheet applies. The same goes for a fresh quarter without share count and profit. Special dividends do not count as dividends, even when they come several times a year or are paid together with the quarterly rate; only the rate in them counts. For companies that report to the SEC, we also replace a quarterly value that is really the whole fiscal year, a profit that includes non-controlling interests, a wrong share count and debt of 0, if the SEC report is consistent in itself. Examples: PTC had a debt to equity ratio of 0 instead of about 0.46, W. R. Berkley a dividend yield of 2.45% instead of 0.48%, Open Text a P/E of 5.9 instead of 10.3. |
| 4 Nov 2026 | s26 | The formulas stay the same. For companies that report to the US Securities and Exchange Commission, we check the data provider's statements against the SEC's figures. If a value in one of the last five quarters or two fiscal years is clearly wrong at the data provider, such as 0, a thousand times too high or a jump the SEC does not show, and it differs by more than a factor of 1.5, the SEC figure applies. Examples: Republic Services had $48M instead of $14.1B in debt, BioMarin a thousand times its revenue. |
| 3 Nov 2026 | s25 | The formulas stay the same, the data are checked more strictly. If debt is missing from the latest annual balance sheet, the debt of the last complete balance sheet applies. Special dividends do not count as dividends, even when the data provider lists them in one payment with the regular rate. Dividend growth above 200% a year counts as a data error. A second quarterly report with the same date, filed late, belongs to the quarter before the filing. Examples: Apple had $2B instead of about $84B in debt, Weyco a dividend yield of 5.7% instead of 2.0%. |
| 30 Oct 2026 | s24 | The formulas stay the same, but the data provider's statements are checked more strictly before the calculation. If a quarter's share count jumps without a split while the reported earnings per share fit the old count, the old count applies. Rows converted into a currency twice are converted back, a fresh quarter without a usable income statement only counts with the complete report, a half year listed as a quarter becomes a quarter, and empty cash flow statements as well as balance sheets with missing debt count as not measured. A large quarter counts if the quarters before it rise steadily. Earnings per share use the profit that matches the reported diluted earnings per share. For insurers and for banks outside the US with an erratic revenue series, revenue growth does not count. Examples: Royal Bank of Canada had a P/E of 12.9 instead of about 18, BNY was calculated with two instead of four quarters, and Nanya Technology was rated on the prior year instead of the last twelve months. |
| 30 Oct 2026 | s23 | The cash flow part of the fair value now has a limit for extreme cases: if free cash flow after the years with analyst estimates would be extrapolated at more than 50% a year, it grows at most as fast as the company's revenue, measured by the revenue estimates or, without them, by the revenue of the last five fiscal years. Over time free cash flow cannot grow faster than revenue, otherwise the margin would rise without limit. For all other stocks the calculation stays as it was. Toyota was previously extrapolated from two analyst years at 85.7% growth to free cash flow of about half of today's revenue, and the fair value was 70% above the price. |
| 26 Oct 2026 | s22 | If more than a fifth of the profit in a period comes from discontinued operations, for example the sale of a division, return on equity, return on assets, net margin, cash conversion and earnings per share growth with that period are now not measured either, for the last twelve months as well as for the fiscal year. Earnings growth starts after such a year, and in the excess return model for banks and insurers such years do not count towards the average return on equity. If the revenue of a year with discontinued operations falls below a third of the prior year, it only contains continuing operations: earlier years no longer count for revenue growth, and the margins of that year are not measured. Tiptree previously showed a return on equity of 68.3% and revenue shrinking 77% a year after selling its subsidiary Fortegra. |
| 26 Oct 2026 | s21 | If the last four quarters are missing, the P/E falls back to the last fiscal year or the data provider's trailing twelve month figure. If more than a fifth of that profit comes from discontinued operations, for example the sale of a division, P/E and PEG are now not measured. Tiptree previously showed a P/E of 1.5 after selling its subsidiary Fortegra. |
| 23 Oct 2026 | s20 | For banks and insurers the net margin no longer counts; profitability there comes from return on equity alone. Their revenue is stated after interest expense, so a net margin of 25 to 35% is normal there and almost always earned full points. In the backtest the score warns just as well afterwards. |
| 23 Oct 2026 | s19 | Utilities with power, gas or water networks calculate the cash flow part of the fair value with the dividend discount model, without the minimum spread of the cost of equity over growth. They permanently invest more than they depreciate, so the cash flow model showed them far too cheap. |
| 22 Oct 2026 | s18 | After a merger with many new shares, the fair value uses estimated earnings per share times today's share count when the earnings estimate is still based on the old share count. |
| 22 Oct 2026 | s17 | Fair value after a build-out: if a company invests more than it depreciates, capital spending in the cash flow model falls step by step to depreciation over the years with analyst estimates, that is to maintaining the existing base. Before, the rate of the build-out years applied forever. |
| 22 Oct 2026 | s16 | Per-share figures are converted to another share class only if the P/E from market value and annual earnings confirms it. Units on the Brazilian exchange, bundles of several shares, are treated as not measured for the P/E and fair value. |
| 22 Oct 2026 | s15 | If the share count in the financial statements does not match the market value (off by more than four times), the P/E, PEG, EV/EBITDA and fair value are treated as not measured until the data match. |
| 22 Oct 2026 | s14 | Fair value far from the price: if the model gives less than a quarter or more than four times the share price, the fair value is treated as not meaningful. The valuation then relies on the P/E and the other metrics. |
| 19 Oct 2026 | s13 | One-off earnings slump: if trailing earnings are below half of the prior year and below a third of expectations (at least three analysts), the P/E does not count, and the forward P/E is shown next to it. |
| 19 Oct 2026 | s12 | One-off items in earnings are detected from the correct quarterly report, the earliest after quarter end. |
| 19 Oct 2026 | s11 | Cyclical companies whose average margin over ten years is zero or negative get no fair value; their valuation then comes from P/E and other ratios. |
| 18 Oct 2026 | s10 | Fair value of real estate trusts (REITs) from FFO and AFFO instead of earnings, mortgage REITs from book value. Dividend growth from individual payments; a fresh cut counts immediately. |
| 18 Oct 2026 | s9 | Cash conversion only from a 1% net margin, cyclicals at peak margin with their ten-year average margin, no dividend model from a 12% yield, return on capital as calculated by S&P. |
| 17 Oct 2026 | s8 | Cash flow part of the fair value with analyst estimates, interest rates per country and industry beta, banks via excess returns. |
| 17 Oct 2026 | s7 | Fair value half from the cash flow model, half from earnings times the industry P/E; the traffic light threshold depends on the uncertainty. |
| 17 Oct 2026 | s6 | If reported and adjusted earnings differ by more than half, the P/E does not count, and earnings do not carry the fair value. |
| 17 Oct 2026 | s5 | After checking the S&P 500 against the SEC: banks with revenue after interest expense, no P/E on a trailing twelve-month loss, short fiscal years excluded. |
| 17 Oct 2026 | s4 | Valuation without a substitute when there is no fair value; PEG only with positive earnings; P/E below 3 not measured. |
| 16 Oct 2026 | s3 | Growth only from three year-on-year comparisons, that is from four fiscal years; younger companies count at 50 there. |
| 15 Oct 2026 | s2 | Valuation weight 8% instead of 15% after the backtest; banks and insurers without EV/EBITDA, PEG and free cash flow. |