SageRanks is not a black box. This page documents exactly how every ranking on this site is computed (each metric, each weight, each hard limit) so you can judge the tool instead of trusting it blindly.
On top of the ranking, each mode applies hard limits: absolute thresholds the investor himself insisted on. Breaching one doesn't change the rank, but flags a warning on the company so the breach is impossible to miss.
Warnings are graded by how far past the limit the company is. The bands are our own editorial choice, applied identically to every mode and metric:
For lower-bound limits (such as Buffett's gross margin > 15%) the distance is measured the inverted way: how far below the bound the value sits, and a non-positive value always counts as the worst band. Each warning's level is written out in text alongside the colour, so the grading works without colour vision.
Every index page also offers a strict screen toggle (off by default): it hides the companies that break any of the mode's hard limits and re-ranks the qualifiers among themselves. Sometimes nothing qualifies; that is not an error, it is the screen doing its job. Fisher mode has no strict toggle, because Fisher set no numeric limits: he trusted judgment, not rules. The same limits also power Passing the screen: live cross-index lists of every company that clears each legend's criteria right now.
| Metric | Better when | What it tells you |
|---|---|---|
| P/B | Lower | Price relative to book value: the classic margin-of-safety measure. |
| PEG | Lower | P/E divided by earnings growth: what you pay for each unit of growth. |
| ROE | Higher | Return on equity: how efficiently the company compounds shareholder capital. |
| Gross margin | Higher | Pricing power. Durable high margins are the fingerprint of a moat. |
| Revenue growth | Higher | Top-line momentum: the raw material of every growth story. |
| Analyst upside | Higher | Distance from current price to the mean analyst target. |
| Buy % | Higher | Share of analysts rating the stock a buy. |
| Score deviation | Lower | Disagreement between analysts: lower means a firmer consensus. |
| Beta | Lower | Volatility versus the market: a rough stability filter. |
| RSI | Closest to 50 | 14-day momentum. Near 50 = neither overbought nor oversold. |
New to any of these? The glossary defines every metric and concept in plain English.
Buy with a margin of safety, wait for fair value. Graham tested price two ways in chapter 14: no more than 1.5 times book value, and no more than 15 times earnings. This mode ranks on exactly those two, weighted equally, because his own 22.5 rule lets one offset the other inside a single budget.
Reweighted 30 August 2026: beta and return on equity are gone. Beta because Graham rejected volatility as a definition of risk in print, in the endnote to chapter 5 — “we find this use of the word ‘risk’ more harmful than useful for sound investment decisions” — and because we had already removed it from Buffett’s mode on weaker evidence than that. Return on equity because it is not one of his seven criteria, and because P/B = P/E × ROE is an identity: measured across our own universe that weight had an effective variance share of minus eight percent, so it pulled against the thing it was there to support.
P/B ×4P/E ×4High return on equity and durable pricing power reveal a moat. Pay a fair price for quality rather than a bargain price for mediocrity.
ROE ×4Gross margin ×3PEG is the master metric: growth is only worth buying when the price hasn't caught up with it. Revenue growth and quality confirm the story.
PEG ×4Revenue growth ×3ROE ×2Upside ×1Score deviation ×1Buy% ×1Find companies growing faster than the market believes, with the margins to fund their own expansion. Fisher held winners for decades: no hard limits, conviction over rules.
Revenue growth ×4Upside ×3Gross margin ×2Buy% ×2Score deviation ×1RSI ×1Buy good businesses (high return on capital) at cheap prices (high earnings yield), and nothing else. Greenblatt's literal formula ranks on EBIT/EV and return on capital. Until 30 August 2026 we had neither figure and used earnings ÷ price with return on equity instead — a version that priced a company as if it carried no debt, and then rewarded the borrowing that created the debt. We now carry enterprise value and return on assets, so the cheap pillar is EV/EBITDA: value measured against the whole business, debt included, which is the entire reason Greenblatt reached for enterprise value. The quality pillar is return on assets, which leverage cannot inflate the way it inflates return on equity. It is still a proxy and not the literal formula — EBITDA is not EBIT, because it ignores what it costs to keep the assets running, and return on assets is not return on capital — but both pillars and the hard limit now measure what Greenblatt measured.
EV/EBITDA ×4ROA ×4Market data comes via Yahoo Finance and is refreshed by our pipeline once an hour on trading days, while the market in question is open. Every market is then swept again overnight, open or not, so nothing on the site is older than the last night. Open pages re-fetch automatically every 5 minutes while visible, so what you see is never older than the last pipeline run. Fundamentals are trailing (TTM) figures and update on the providers' own cadence.
| Region | Price refresh | Fundamentals & consensus |
|---|---|---|
| Mag7, GRANOLAS, S&P 100, TSX 60 | Every hour 09:00–17:00 ET | Daily |
| Nordic | Every hour 09:00–17:30 CET | Daily |
| DACH, Benelux, Med | Every hour 09:00–17:30 CET | Daily |
| Asia Pacific | Every hour overnight CET (SGX/ASX/KRX) | Daily |
| Baltic | Every hour 09:00–17:30 CET | Daily |
Every other market (FTSE 100, CAC 40, Nikkei, Hang Seng, Taiwan, Thailand, New Zealand, Baltic) runs on the same hourly cadence inside its own regional window. On top of that, one nightly sweep covers all seventeen files regardless of opening hours, so a market that was closed all day still has fresh figures in the morning. A row that could not be refreshed keeps its previous value and carries the timestamp of the last time it actually came from Yahoo.
Rankings are relative. #1 means best within that index, by that philosophy, not a buy signal. A deep-value #1 in an expensive universe is merely the least expensive.
Sector bias is real. Graham's P/B rule structurally punishes asset-light businesses (software, pharma) and flatters banks and industrials. Compare within sectors mentally, or use multiple modes.
Analyst data lags. Targets and ratings are updated by analysts on their schedule, not ours. Fresh news moves prices before it moves consensus.
Approximation, not replication. The modes translate each investor's published principles into today's data. Greenblatt's actual formula ranks on EBIT/EV and return on capital; we rank on EV/EBITDA and return on assets. Those are the closest measures Yahoo publishes, and they are not the same measures: EBITDA flatters companies that must keep replacing their assets, and return on assets says nothing about how the assets were financed. Closer than the earnings-yield version it replaced, but still an approximation.
We have no price history. Every ranking on this site is a snapshot of today. We cannot tell you what a company looked like six months ago, whether a rank is improving or a P/B has halved since spring. That is a real limit, not a feature we forgot.
Fisher has no hard limits at all, and that is deliberate. His fifteen points are entirely qualitative — management quality, research, the sales organisation — and he set no numeric screens. So an empty Fisher view means there are no rules to break, not that nothing is wrong. It is the one mode where the absence of a warning tells you nothing.
Beta in Graham's mode is our reading, not his. Graham measured risk as price against value, not as volatility. We weight beta there as a proxy for "defensive", which is an interpretation we are making on his behalf. We removed it from Buffett's mode entirely, because he rejected beta as a measure of risk explicitly.
Not investment advice. SageRanks is an analysis and education tool. Nothing on this site is a recommendation to buy or sell any security. Do your own research. That's rather the point.