Methodology
Every stock on this site gets an overall score from 0 to 100 and one fair value — our estimate of what a share is worth. The score comes from five angles on the business; the fair value comes from three independent estimates combined into one number. This page explains, in plain language, what each number means and how to read it. Prices are the 2026-11-16 close unless marked LIVE during market hours.
(a) What the score is
Every stock gets an overall score from 0 to 100, built from five pillars, each scored 0–100: Conviction, Valuation, Quality, Technical (Timing), and Trust. Each stock also gets a verdict — Growth, Value, Watch, or Avoid — based on its scores and how its price compares with our fair value.
227 stocks scored today — see them all on All Stocks.
(b) Each pillar, in plain words
Conviction — the Story-Tracker
How good the business story and its financials look: whether sales and profits are growing, analysts are raising their forecasts, and the stock is leading the market.
Valuation
How the share price compares with our fair value. Our fair value combines three independent estimates: one based on the company’s cash flow, one based on its growth and dividend compared with its earnings multiple, and one based on next year’s expected earnings at its sector’s going multiple.
One-off gains are taken out and one-off charges are added back before earnings are valued, so a single unusual quarter does not distort the picture — in either direction. A gain that will not repeat makes a business look more profitable than it is; a charge that will not repeat makes it look less profitable than it is.
The growth behind the cash-flow and growth-based estimates is also kept honest: when the implied growth rate is far out of line with what the business itself is growing at — or when no analyst publishes a long-term growth estimate for the name at all — the growth used is held to a level the business’s own revenue growth can support. The third estimate, based on next year’s earnings at the sector’s multiple, is unaffected.
Quality / Forensics
Whether the reported profits are real: backed by cash actually coming in, customers paying what they owe, and borrowing kept in check.
Technical / Timing
What the price chart says about timing. It is about when, never about whether a business is worth owning, and it is the smallest part of the score.
Trust
How much we can rely on the numbers behind the other four pillars: whether the data is complete, analysts broadly agree, and insiders are not all selling. There is no insider feed in this build, so that part stays neutral — it is never read as “no selling”.
(c) The ladder — valuation context, not buy/sell instructions
The ladder marks where the price sits against our fair value, in four steps from well below it to well above it. It is valuation context — a way to see how far the price is from our estimate — not an instruction to buy or sell at any level. No level on the ladder is a recommendation.
That restraint is evidence-based, not modesty. Three backtests measured whether the ladder separates forward returns:
Backtest v2 — 12 quarter-ends, Dec 2023 – Sep 2026, 5 tickers (Variant A)
- 38 of 50 observations above Exit.
- Six months after being above Exit: +19.9% average return, 72% of observations positive, +10.0% vs SPY.
- Correlations, upside vs forward return: upside vs forward return: +0.03 (1M) / −0.11 (3M) / +0.00 (6M).
Full 749-name backtest — 2,498 fair-value observations, 748 names, 12 quarter-ends
- Six-month return below Buy: +6.5% (n=579, −4.1% vs SPY).
- Six-month return above Exit: +9.5% (n=1,048, −0.6% vs SPY).
- Correlations: −0.013 (1M) / −0.011 (3M) / −0.054 (6M).
Backtest v3 — CAPM-discount variant, full 749-name universe
- Six-month return below Buy: +6.0% (n=655).
- Six-month return above Exit: +9.8% (n=951).
- Correlations: −0.01 (1M) / −0.01 (3M) / −0.06 (6M).
Read plainly: stocks above Exit did not underperform stocks below Buy — in the wide backtest they did better. The ladder does not separate forward returns, so it is published as context and never as a signal.
(d) CAPM cross-check — display only
A second, market-based check on our cash-flow estimate. It uses live market interest rates instead of our fixed assumptions, and it is display-only: it does not change any fair value, score or verdict.
A full backtest found it does not separate future returns any better than the estimate it checks — which is why it stays display-only.
(e) Data sources & cadence
| Data | Source | Cadence |
|---|---|---|
| Fundamentals | SEC filings | Daily |
| Prices / technicals | Yahoo Finance | Daily close |
| Free cash flow / operating cash flow | Yahoo quarterly cash-flow statements | Quarterly |
| Forward EPS / P/E / price targets / beta | Yahoo Finance | Daily |
| Insider activity | Form 4 / 13G-D filings | As filed (no live feed in this build — see Trust) |
| Macro (Market Pulse) | FRED + Yahoo | Every 6 hours (feed not wired in this build) |
| Conviction (Story-Tracker) | Story-Tracker | Weekly, Monday pre-market |
| Top Picks | Engine output | Frozen at publication; changes only by decision |
One field is deliberately discarded: Yahoo’s freeCashflow field was found wrong, so free cash flow is taken from the quarterly cash-flow statement instead.
(f) How Top Picks performance is tracked
Top Picks are the four launch names plus the top five Growth and top five Value names as of the 16 Nov 2026 close.
The list is frozen at publication. It changes only when a new decision is made — names are never rotated in automatically.
Performance is tracked in two groups, one for each entry date, and each group is measured on its own — the two groups are never blended into one return.
The S&P 500 is the benchmark, shown with returns since each entry date. It carries no costs or dividends either, same as the picks.
Reading cards — every metric, explained
Every number on a stock card, explained in plain language: what it is and how to read it. Scores are on a 0–100 scale unless stated otherwise. A fair value is our estimate of what a share is worth — it is research, not a buy or sell recommendation.
- DCF leg
- PEGY leg
- Forward-sector leg
- Fair value
- Conviction
- Valuation
- Quality
- Technical (Timing)
- Trust
- Valuation ladder
- DCF sensitivity grid
- The Stock Check vs Wall Street
- CAPM cross-check
Prices at the 13 November 2026 close; engine run 16 November 2026.
DCF leg
This estimate values a company by the cash it is expected to generate for its owners in the years ahead, worked back to what that cash is worth today. A higher value means the company’s expected cash flow supports a higher price — but it leans heavily on assumptions about future growth, so read it together with the other two estimates.
PEGY leg
This estimate asks whether the price you pay for each dollar of earnings is reasonable once you count how fast earnings are growing and the dividend you collect while you wait. If growth plus dividend income is high compared with the price-to-earnings ratio, the stock looks cheap on this view; if it is low, it looks expensive.
Forward-sector leg
This estimate values a stock the way the market usually does: next year’s expected earnings per share, priced at the multiple its whole sector currently trades at. It is the most market-priced of the three estimates, so when it stands far above the other two, the fair value is resting on forecasts rather than on today’s cash flow.
Fair value
Our fair value is one number built from three independent estimates, so no single estimate can dominate it. When the price is below our fair value, our estimates on balance say the stock is worth more than it costs — and when it is above, they say the opposite.
Conviction
Conviction measures whether the business itself is worth following — whether its story and its financials are both getting better. A high score here means a good business, not a cheap price: a strong company can still be too expensive to score well overall.
Valuation
Valuation turns our fair value into a 0–100 score: the further the price sits below our fair value, the higher the score. Always read it with how far apart the three estimates are in mind — a high score means less when the estimates disagree wildly.
Quality
Quality checks whether reported profits are real — backed by cash actually coming in, customers paying what they owe, and borrowing kept in check. When profits run well ahead of cash, treat the earnings with caution, however big they look.
Technical (Timing)
Technical, labelled Timing on the card, asks only whether now looks like a good moment to act, based on the price chart alone. It says nothing about whether the business is good, and a low score here never makes a stock Avoid on its own.
Trust
Trust measures how much we can rely on the numbers behind the other pillars — whether the data is complete, analysts broadly agree, and insiders are not all selling. A low score means read the rest of the card with more caution, not that the business itself is bad.
Valuation ladder
The ladder shows where the price sits against our fair value in four steps, from well below it to well above it. These are reference points for context only — never instructions to buy or sell.
DCF sensitivity grid
This grid shows the cash-flow estimate not as one number but as a table, so you can see how much it moves when growth or interest-rate assumptions change. If the whole grid sits below the share price, the cash-flow view cannot support the price on any plausible assumption.
The Stock Check vs Wall Street
Wall Street analysts publish their own price targets, and this card puts our fair value next to theirs so you can see where we agree and where we disagree. A small gap means rough agreement; a large gap means one of us is wrong — and when none of our three estimates sits near their average target, we say so plainly.
CAPM cross-check
This is a second opinion on our cash-flow estimate, built with live market interest rates instead of our fixed assumptions. It changes nothing — no fair value, score, or verdict — and is shown only so you can see how the estimate holds up under a market-based view.