How Kaplio judges a company
Two layers, both free. The Summary says what the business is and at what price it trades: four lights (valuation, business, balance sheet, results), a price range from its own decade and its sector, real peers, dividend and the return you would get if consensus is right. The methods are two verdicts and two lists: Buffett and Lynch say what that investor would have done today with this company, rule by rule and with the source of each rule; Greenblatt and Druckenmiller give a ranking position and a momentum level, which are not a verdict. On top of them, the Graham defensive-investor basket, published as a list in the Radar. Everything is deterministic and traceable; where a rule cannot be computed from public accounts (the quality of management, for instance) we say so instead of inventing a number. None of it is a recommendation.
The Kaplio Summary
The main verdict on every company page. Four lights, each judged against the thresholds of its own sector category (a bank is not judged like a software company): valuation, business, balance sheet and results. A price range built from the company's own P/E of the last decade (weekly, on known trailing earnings) and the median of its sector; banks and insurers by a justified price-to-book and their real peers. Then the peers with their multiples, the dividend (yield, payout, streak) and the expected return over three years if consensus earnings are met and the multiple returns to its median.
What it looks at:
- Valuation: cheap / fair / expensive against the decade and the sector
- Business: margins, returns on capital, growth against its category
- Balance sheet: net debt, coverage, Altman Z
- Results: last quarters against consensus and against the prior year
- Range: own median P/E of the decade and sector median; P/B for banks and insurers
- Expected return: consensus EPS × median P/E + dividend
Buffett method
A computable distillation of the shareholder letters, Buffettology and Hagstrom's work, written down first as a dossier and then as rules. First the circle of competence and the routing by business type: banks, insurers and commodity businesses are declared outside the circle rather than scored. Then the purchase rules (sustained return on equity measured on a decade median, margins, retained earnings, contained debt) and the veto rules (accounting, business and structural). The result is a business quality from 0 to 100 that does not include the price. The price is judged separately: the ten-year equity-bond projection or, failing that, the farm rule, against a 10 % hurdle. Approve, wait, veto or too hard.
What it looks at:
- Circle of competence and routing by business type
- Purchase rules: ROE median over a decade, margins, retention, debt
- Veto rules: accounting, business, structure
- Quality 0-100 without the price
- Price: equity-bond over ten years or farm rule, hurdle 10 %
- Verdict: approve / wait / veto / too hard
Lynch method
From One Up on Wall Street and Beating the Street. The method starts by classifying the company into one of Lynch's six categories (slow grower, stalwart, fast grower, cyclical, turnaround, asset play), because each is bought and sold by different rules. Then the purchase and veto rules of the category and the sector checklists (banks, insurers, retailers, utilities). Quality is a 0-100 without the price. The price is judged with the adjusted PEG for growers, with net cash per share for asset plays and turnarounds, and with the inverted P/E signal for cyclicals.
What it looks at:
- Classification into six categories, with a confidence level
- Purchase and veto rules by category
- Sector checklists
- Quality 0-100 without the price
- Price: adjusted PEG (growth + yield) / P/E, net cash per share, cyclical P/E
- Verdict: approve / wait / veto
Greenblatt method (a ranking)
A ranking, not a grade. Every night Kaplio sorts the 2,043 companies it covers (market cap of $2 billion or more) by the two figures of The Little Book That Beats the Market (2006): return on capital (operating profit of the last twelve months over net working capital plus net fixed assets) and earnings yield (that same operating profit over enterprise value at today’s price). The two ranks are added and the resulting list gives each company a position and a percentile; the first 30 are “the formula’s 30”, published in the Radar. As in the book, banks, insurers, utilities and real estate are left out, along with companies with negative operating profit or accounts older than fifteen months. Greenblatt’s own summary: buy good companies (high return on capital) at bargain prices (high earnings yield), hold for a year and rotate.
What it looks at:
- Return on capital = EBIT / (net working capital + net fixed assets)
- Earnings yield = EBIT / enterprise value at today’s price
- Universe: Kaplio coverage (2,043 companies of $2bn or more) without financials, utilities or REITs
- Combined position = rank by return on capital + rank by earnings yield
- Output: rank, percentile (100 = best) and whether it is among the formula’s 30
Druckenmiller method (momentum)
Not his method: a computable reconstruction of what Druckenmiller has said about momentum, relative strength and revisions (The New Market Wizards, 1992; the Lost Tree Club talk, 2015). Six rules, each with its source: D1 absolute momentum over 6 and 12 months, both positive and in the top third of the coverage; D2 relative strength against the S&P 500 over 6 and 12 months; D3 consensus revisions (next fiscal year EPS estimate at least 3 % above 60 days ago); D4 leading sector (its sector in the top third of 3-6 month returns); D5 price within 10 % of the 52-week high; D6 traded liquidity of $20 million a day or more. The level is strong, medium or weak depending on how many it passes; it is momentum, not a value verdict. The seventh rule, the macro liquidity regime (D7), is declared and not computed: we do not yet have the Fed and yield-curve data. Nor do we measure his risk management, his sizing by conviction or his currency and bond bets, which are half of his record.
What it looks at:
- D1 momentum over 6 and 12 months: both positive and, over 12 months, better than 70 % of the coverage
- D2 relative strength: ahead of the S&P 500 over 6 and 12 months
- D3 revisions: next fiscal year’s estimated earnings per share ≥ +3 % versus 60 days ago
- D4 leading sector: top third of 3-6 month returns
- D5 52-week high: price ≥ 90 % of the high
- D6 liquidity: average traded volume ≥ $20 million a day
- D7 macro regime (Fed, curve, credit): declared, not yet computable
- Level: strong / medium / weak by rules passed; not a value verdict
Graham defensive investor
A basket, not a verdict on one company. The seven checks Benjamin Graham sets for the defensive investor in The Intelligent Investor, chapter 14, applied every night to the Kaplio coverage: size (market cap of $2 billion or more, Jason Zweig’s update of the 1972 sales threshold); balance sheet (current assets at least twice current liabilities, and long-term debt no larger than net working capital); positive earnings in each of the last ten years; dividend paid without interruption over the twelve years of history we hold (Graham asked for twenty); earnings per share at least a third higher than ten years ago, comparing three-year averages; a P/E of 15 or less on the average earnings of the last three years; and a price of 1.5 times book value or less, or else P/E times price-to-book of 22.5 or less. For banks and insurers the balance-sheet check does not apply, as in the book, and they pass with six of six. Graham designed the filter for a diversified portfolio of ten to thirty stocks, not to judge a single company, and in 2026 almost no large company with buybacks or a strong brand passes the price checks: that is why it lives as a list in the Radar and not as a row on the company page. A list, not a recommendation.
What it looks at:
- G1 size: market cap ≥ $2bn (Zweig, 2003)
- G2 balance sheet: current assets ≥ 2 × current liabilities and long-term debt ≤ net working capital; not applied to banks and insurers
- G3 stability: positive earnings in each of the last ten years
- G4 dividend: paid without interruption over the twelve years of available history
- G5 growth: three-year average EPS ≥ +33 % versus ten years earlier
- G6 price: P/E ≤ 15 on three-year average earnings
- G7 assets: P/B ≤ 1.5, or P/E × P/B ≤ 22.5
- Output: passes N of 7 (6 for financials); those passing all of them, in the Radar. Source: The Intelligent Investor, ch. 14