Try it in the terminal
- AAPL AI
- See how Apple scores against all 12 investor styles.
- MSFT AI
- Check Microsoft, then press Ask AI on a card for a written take.
- KO AI
- Try a steady dividend payer and compare the value-style cards with the growth-style ones.
What it shows
A grid with one card for each of 12 investors. Each card shows a score ring from 0 to 100, a verdict (Bullish, Neutral, Bearish or Not enough data), and a checklist of the rules for that style. Each rule shows a tick, a cross or a dash, and the stock's actual figure.
At the top a summary bar counts how many cards are bullish, neutral and bearish, and shows the average score across the cards that could be scored.
The rules are inspired by each investor's published approach. They are not that person's actual views, their current holdings or a prediction of what they would do.
How to open it
Type the ticker, a space, then AI, for example AAPL AI. You can also open the stock and click the AI tab. Do not confuse it with ASK, which is the chat with the AI analyst, or EAI, which reads earnings filings.
The 12 checklists
Each check compares one figure with a fixed threshold. These are the thresholds the terminal uses. Percentages are percent, and the others are ratios shown with an x.
- Warren Buffett (quality compounders at a fair price): return on equity at least 15%, operating margin at least 20%, debt-to-equity at most 0.5, five-year earnings growth at least 10%, P/E at most 25.
- Charlie Munger (great businesses, few decisions): return on invested capital at least 15%, gross margin at least 40%, debt-to-equity at most 0.8, five-year revenue growth at least 5%.
- Benjamin Graham (deep value with a margin of safety): P/E at most 15, price-to-book at most 1.5, current ratio at least 2, pays a dividend (yield at least 0.01%), five-year earnings growth at least 0%.
- Peter Lynch (growth at a reasonable price): PEG at most 1, five-year earnings growth at least 10%, debt-to-equity at most 0.6, P/E at most 30.
- Philip Fisher (outstanding growth companies): five-year revenue growth at least 10%, gross margin at least 40%, operating margin at least 15%, net margin at least 10%.
- John Templeton (buy at maximum pessimism): P/E at most 12, price-to-book at most 1.5, share price at most 70% of its 52-week high.
- Joel Greenblatt (good and cheap): return on invested capital at least 20%, EV/EBITDA at most 10.
- Seth Klarman (margin of safety, patient value): price-to-book at most 1.5, free cash flow yield at least 6%, debt-to-equity at most 0.5.
- Bill Ackman (concentrated quality, cash-generative): free cash flow yield at least 4%, operating margin at least 20%, debt-to-equity at most 1.5, five-year revenue growth at least 5%.
- Michael Burry (contrarian deep value on cash flow): free cash flow yield at least 10%, EV/EBITDA at most 8, price-to-book at most 1.
- Stanley Druckenmiller (growth and momentum): revenue growth over the last twelve months at least 15%, earnings growth over the last twelve months at least 20%, share price at least 85% of its 52-week high.
- Cathie Wood (disruptive innovation): five-year revenue growth at least 20%, revenue growth over the last twelve months at least 25%, gross margin at least 50%.
What the figures mean
All figures come from Finnhub's company fundamentals and are mostly for the trailing twelve months (the last four reported quarters). A few terms in plain English:
- ROE and ROIC: how much profit a company makes on the money shareholders (ROE) or all investors (ROIC) have put in. Higher is better. ROIC is taken from the latest annual figure.
- Gross, operating and net margin: the share of each dollar of sales that is left after the cost of goods, after running costs, and after everything including tax.
- Debt / equity: how much the company owes compared with what its shareholders own. Lower means less borrowing.
- Current ratio: short-term assets divided by short-term bills. Above 2 means a comfortable cushion.
- P/E: price divided by yearly earnings per share. P/B: price divided by book value, roughly what the company's assets are worth on paper.
- PEG: the P/E divided by the five-year earnings growth rate. The terminal calculates it only when growth is above zero.
- FCF yield: free cash flow (cash left after running costs and investment) as a percentage of the price. The terminal works it out as 100 divided by the price-to-free-cash-flow ratio, and only when that ratio is positive.
- EV/EBITDA: the company's total value (including debt) divided by its operating profit before some costs. Lower is cheaper.
- Price vs 52-week high: today's price as a percentage of the highest price of the last year. It uses the live price.
How scoring works
- Each check is a pass, a fail, or no data.
- The score is the percentage of the checks with data that passed, rounded to a whole number. Checks with no data are left out of the calculation.
- At least two checks need data. If fewer than two do, the card shows a dash and the verdict Not enough data.
- Verdict: 70 or more is Bullish, 40 to 69 is Neutral, and below 40 is Bearish.
- The average score at the top is the plain average of all cards that have a score.
- A missing dividend yield counts as 0%, not as no data. So a company that pays no dividend fails Graham's dividend check instead of skipping it.
Scoring is done in your browser, using the company data and the live price. No AI is used for the scores, the ticks or the verdicts.
Tips
- Look at which checks failed, not only at the score. A Bullish card that passes only two of two checks is much thinner than one that passes five of five.
- Compare cards that disagree. A stock that is Bullish for Fisher and Wood but Bearish for Graham and Burry is a growth stock that is not cheap. That is useful information about what kind of stock it is.
- Use FA (financials) and DCF (valuation) to look more closely at any figure, and EAI to read what management says about it.
- A figure the data provider does not have (for example a P/E for a company with no reported earnings) removes that check from the score instead of failing it. Look for dashes and the words no data on the card.
Limits and good to know
- The thresholds are simple rules of thumb applied to every company alike. Banks, utilities, biotech and other sectors have very different normal ranges.
- Figures can be missing or stale for some stocks, and the figures are only as good as the data provider's.
- A card's AI take is cached for each stock, investor and day, so asking again the same day returns the same text. Each person can request 10 a minute and 40 a day, and there is a daily cap for everyone together. These limits are counted per server.
- The AI can be wrong, even when it quotes your figures. Its text is written in a style, and it is not a statement by the investor.
- This screen is information, not investment advice.
Questions about Investor Legends
Is this what Warren Buffett (or anyone else) would actually do?
No. The cards are simple rule checks inspired by each investor's published ideas. They do not reflect the investor's real opinions, current holdings or full process. The AI text is written in a style and is not a quote or an endorsement from them.
Why does a card show Not enough data?
Fewer than two of that card's checks had a figure to test. This happens a lot with newer companies, funds, or non-US stocks. The figures come from Finnhub, which does not have everything for every ticker.
Why did Ask AI say there is not enough data?
The AI take is only generated when the stock has a company name and the card has a score, so there is enough to talk about. Cards showing a dash are refused before any AI call.
Can I trust the AI commentary?
It is a short summary of the numbers on the card, not research. The AI is told to use only the figures it is given and not to invent facts or quotes, but it can still misjudge. The score and ticks are calculated by plain code, so check the checklist against the commentary. The terminal is information, not investment advice.
Why does Ask AI say too many requests?
Each new commentary is a paid AI call. Each person can request 10 a minute and 40 a day. Repeat requests for the same stock, investor and day are served from a saved copy and do not count against these limits. Wait a little and try again.
Why do the Bullish, Neutral and Bearish labels not say whether to buy?
They only describe how many of that style's rules the stock passes: 70 or more, 40 to 69, or below 40. They are not a price prediction or a recommendation.
Do I need an account?
No. The legend checklists and Ask AI work without signing in.
