This is an original data study. We took every US-listed stock in the invest-like.com universe (12,543 tickers as of 25 May 2026) and ran each one through seven separately implemented value-investing frameworks: Warren Buffett's quality-and-moat lens, Benjamin Graham's defensive-investor screen, Philip Fisher's growth-quality test, Peter Lynch's growth-at-a-reasonable-price thesis, Joel Greenblatt's Magic Formula, Charlie Munger's mental-models filter, and Terry Smith's Fundsmith framework.
Then we asked the question almost no public source has data to answer: when you run the same universe through seven independent value frameworks, how much do they actually agree?
The answer surprised us. So we wrote it up.
TL;DR
- 47 US-listed stocks pass all 7 named-investor frameworks at a B-plus or better (score >= 60 on the invest-like 0-to-100 scale). That is 0.4 percent of the 12,543-stock universe and 0.71 percent of the 6,621-stock apples-to-apples cohort (stocks that have been scored on all 7 frameworks).
- The 47-stock consensus cohort is tech and industrials heavy, not consumer staples heavy. Technology is 40.4 percent of the list. Industrials 17.0 percent. Healthcare 12.8 percent. Consumer Defensive (the classic "Buffett sector") is just 4.3 percent.
- Median market cap of the consensus cohort is $69.2 billion. Mean is $610 billion (skewed by NVIDIA, Alphabet, and Apple).
- Greenblatt's Magic Formula is by far the strictest framework on the current universe (4.2 percent pass rate). Graham's defensive screen is the loosest (19.1 percent pass rate). That is a 4.5x spread in selectivity across "value" investors who all claimed to be doing roughly the same job.
- The distribution is bimodal. Of the 6,621 stocks scored on all 7 frameworks, 4,493 (67.9 percent) pass zero frameworks and 47 (0.71 percent) pass all 7. Almost nothing lives in the middle.
The rest of this post shows the methodology, the full distribution, the 47-ticker consensus list, the framework hardness ranking, and how you can reproduce any of it through the invest-like.com public API.
Dated snapshot: the underlying scores were last refreshed on 25 May 2026 at 03:30 UTC. Citations of specific numbers below are stable against that snapshot. The 47-stock cohort, sector mix, and framework pass rates will drift over time as new fundamentals get ingested.
Methodology
The data study covers the 12,543 stocks in the invest-like.com primary universe. 10,812 (86 percent) are US-domiciled. The remainder are foreign-domiciled issuers with US listings (ADRs and direct listings from Taiwan, Netherlands, Canada, Ireland, etc.). When we say "US-listed" below, that is the universe we mean.
Each of the 7 frameworks is implemented as a deterministic scoring function. The function takes a stock's normalized fundamentals (most recent 5 years from FMP) and returns a single 0-to-100 score, a letter grade (A through F), and a verdict ("yes", "maybe", "no") on whether the stock fits that framework. The frameworks are independent. There is no blending step in this study, no AI rewriting on top of the numbers, and no manual overrides.
The 7 frameworks, in alphabetical order, and what each measures:
- Buffett, wide-moat quality compounding. Sustained ROIC, durable margins, owner-friendly capital allocation, fortress balance sheet, fair valuation.
- Fisher, growth quality scuttlebutt. Revenue durability, R&D intensity, management quality signals, qualitative competitive advantage.
- Graham, defensive value. Low P/E, low P/B, fortress balance sheet, dividend record, earnings stability over 10 years.
- Greenblatt, Magic Formula. Ranks the universe jointly on earnings yield and return on capital, then keeps the top decile.
- Lynch, growth at a reasonable price. PEG ratio under 1, mid-cap sweet spot, comprehensible business model, insider activity.
- Munger, mental-models filter. Concentration on quality, capital intensity penalty, simplicity premium, balance-sheet conservatism.
- Smith, Fundsmith framework. Quality of earnings (cash conversion, ROCE persistence), pricing power, low-capital reinvestment.
The full methodology hub is at /methodology/, with per-framework explainers at /methodology/buffett-fit/ and /methodology/deal-breakers/. The same scoring functions power every verdict on the production site and every payload returned by the public API at /api/public/verdict/[ticker].
Pass threshold. We define "passes a framework" as a score of 60 or above on the 0-to-100 scale. That maps to a B-minus or better letter grade. We chose 60 (not 70 or 80) because it preserves enough sample size to see the distribution shape and because a B-minus is the rough cutoff where the invest-like verdict flips from "weak" to "partial" fit.
Apples-to-apples cohort. Some frameworks (Munger, Smith, Fisher) have been scored on fewer stocks than others (Buffett, Graham). When we report cross-framework counts, we restrict to the 6,621 stocks that have been scored on all 7 frameworks. That is the apples-to-apples cohort. We say so explicitly each time it matters.
Headline finding 1: only 47 stocks pass all 7 frameworks
Across the 6,621 stocks that have been scored on every framework, here is the histogram of how many frameworks each stock passes:
| Frameworks passed (at score >= 60) | Stock count | Share of cohort |
|---|
| 0 | 4,493 | 67.86% |
| 1 | 1,274 | 19.24% |
| 2 | 314 | 4.74% |
| 3 | 175 | 2.64% |
| 4 | 148 | 2.24% |
| 5 | 99 | 1.50% |
| 6 | 71 | 1.07% |
| 7 | 47 | 0.71% |
Two things stand out.
First, the distribution is bimodal. Two-thirds of US-listed stocks pass zero of the seven frameworks. Less than 1 percent pass all seven. Almost nothing lives in the middle. That is not a coincidence. Value-investing frameworks are designed to be selective, and the criteria overlap heavily on fundamentals like balance-sheet strength and earnings quality, so a stock that fails one of them tends to fail several.
Second, the number to remember is 47. That is the count of US-listed stocks where every one of Warren Buffett, Ben Graham, Phil Fisher, Peter Lynch, Joel Greenblatt, Charlie Munger, and Terry Smith would, on the surface of the documented criteria, say "yes" right now. It is 0.4 percent of the 12,543-stock universe and 0.71 percent of the apples-to-apples cohort.
For comparison, the S&P 500 has 503 constituents. The Russell 1000 has roughly 1,000. The Russell 3000 has 3,000. The all-seven consensus cohort is smaller than any of those benchmark indexes by an order of magnitude or more.
Headline finding 2: framework hardness varies 4.5x
The seven frameworks share the "value investing" label, but their selectivity on the current universe is wildly different. Restricting to the 6,621-stock apples-to-apples cohort:
| Framework | Stocks passing at score >= 60 | Pass rate | Average score |
|---|
| Greenblatt (Magic Formula) | 280 | 4.23% | 21.0 |
| Smith (Fundsmith) | 400 | 6.04% | 33.9 |
| Buffett (moat + quality) | 433 | 6.54% | 35.0 |
| Munger (mental models) | 465 | 7.02% | 35.5 |
| Fisher (growth quality) | 641 | 9.68% | 39.8 |
| Lynch (GARP) | 783 | 11.83% | 30.8 |
| Graham (defensive) | 1,267 | 19.14% | 40.2 |
Greenblatt's Magic Formula is 4.5 times stricter than Graham's defensive screen. That is a surprisingly large gap inside the value-investing tent.
The why is intuitive in hindsight. Greenblatt's Magic Formula is explicitly a relative-rank screen (top decile of universe on combined earnings yield + ROIC), so the math caps its pass rate by construction. Graham's defensive criteria, by contrast, are absolute thresholds (P/E under 15, P/B under 1.5, current ratio above 2x). When the market trades at elevated multiples, fewer stocks clear Graham's absolute hurdles, but at a normal market valuation more stocks pass. The current universe sits at a moment where Graham's absolute thresholds are surprisingly forgiving on the long tail of small-caps, banks, and forgotten cyclicals.
Buffett, Munger, and Smith land in a tight cluster at 6 to 7 percent pass rate. That fits the lore: all three are looking for the same kind of business (high ROIC, durable margins, low reinvestment need, owner-friendly management) just with different vocabulary.
Lynch is the only framework where the average score (30.8) is meaningfully lower than its pass rate would suggest. That is because Lynch's PEG-under-1 criterion is binary and unforgiving for slow growers, while everything else (mid-cap sweet spot, comprehensible business, insider buying) is more permissive. The result: a fat left tail and a healthy pass count, but a low cohort average.
Headline finding 3: the consensus cohort skews tech, not staples
There is a strong intuition in the value-investing community that "Buffett-style stocks" means consumer staples, packaged food, beverages, household products, tobacco. Coca-Cola, Procter and Gamble, Kraft Heinz, Philip Morris.
The data from the 47-stock all-seven-frameworks cohort says that intuition is dated.
| Sector | Stocks in 7-of-7 cohort | Share |
|---|
| Technology | 19 | 40.4% |
| Industrials | 8 | 17.0% |
| Healthcare | 6 | 12.8% |
| Financial Services | 5 | 10.6% |
| Consumer Cyclical | 3 | 6.4% |
| Communication Services | 2 | 4.3% |
| Consumer Defensive | 2 | 4.3% |
| Energy | 1 | 2.1% |
| Real Estate | 1 | 2.1% |
Technology alone is 40 percent of the cohort. Add Industrials and Healthcare and you have 70 percent of the consensus list. Consumer Defensive (the classic Buffett sector) is two names: Philip Morris International and Monster Beverage. Consumer Staples in the old textbook sense is not where the cross-framework consensus lives in 2026.
Why has the consensus moved? Three reasons that fall out of the data without us having to argue them.
First, the modern Tech sector contains an unusually large number of high-ROIC capital-light franchises (semiconductors with dominant equipment positions, vertical software, payment networks classified under Financial Services). These businesses score well on every framework that rewards quality compounding, which is six of the seven we test.
Second, the Industrials cohort is dominated by specialty services and electrical-equipment franchises (Cintas, Vertiv, Trane, Comfort Systems, Powell Industries) that share the same capital-light, sticky-revenue profile that consumer staples used to monopolize. These names have moats built on contract switching costs, regulatory friction, or installed base economics.
Third, classic consumer staples have lost pricing power in the last decade. Many of them now fail Graham's balance-sheet thresholds (leveraged buyback programmes) and Smith's cash-conversion test (working-capital intensity). The names that used to define the value-investing canon increasingly do not pass the canon's own tests.
The 47 names that pass all 7 frameworks
This is the full consensus list as of 25 May 2026, ordered by market cap. Cohort average score across all 7 frameworks shown in the rightmost column.
| Ticker | Company | Sector | Market cap | Avg score |
|---|
| NVDA | NVIDIA Corporation | Technology | $5.22T | 83.4 |
| GOOGL | Alphabet Inc. | Communication Services | $4.63T | 68.3 |
| GOOG | Alphabet Inc. | Communication Services | $4.59T | 68.3 |
| AAPL | Apple Inc. | Technology | $4.54T | 73.6 |
| TSM | Taiwan Semiconductor Manufacturing | Technology | $2.10T | 74.1 |
| AVGO | Broadcom Inc. | Technology | $1.96T | 83.1 |
| LLY | Eli Lilly and Company | Healthcare | $1.00T | 79.7 |
| V | Visa Inc. | Financial Services | $630B | 77.7 |
| ASML | ASML Holding N.V. | Technology | $629B | 79.4 |
| MA | Mastercard Incorporated | Financial Services | $441B | 83.0 |
| LRCX | Lam Research Corporation | Technology | $382B | 81.3 |
| AMAT | Applied Materials, Inc. | Technology | $343B | 72.1 |
| PM | Philip Morris International Inc. | Consumer Defensive | $295B | 71.0 |
| KLAC | KLA Corporation | Technology | $247B | 86.7 |
| ANET | Arista Networks, Inc. | Technology | $194B | 82.4 |
| APP | AppLovin Corporation | Technology | $162B | 86.7 |
| ISRG | Intuitive Surgical, Inc. | Healthcare | $155B | 72.0 |
| VRT |
The next 27 names, smaller by market cap, are Monster Beverage (MNST), Airbnb (ABNB), Hilton Worldwide (HLT), Cintas (CTAS), Comfort Systems USA (FIX), Autodesk (ADSK), Fastenal (FAST), IDEXX Laboratories (IDXX), MSCI (MSCI), Ubiquiti (UI), Fair Isaac (FICO), VeriSign (VRSN), Texas Pacific Land (TPL), Rollins (ROL), Mettler-Toledo (MTD), UL Solutions (ULS), Nextpower (NXT), Rambus (RMBS), Medpace Holdings (MEDP), Powell Industries (POWL), Manhattan Associates (MANH), Sezzle (SEZL), Sprott (SII), Neptune Insurance Holdings (NP), Napco Security Technologies (NSSC), IRadimed (IRMD), and Postal Realty Trust (PSTL).
The smallest member of the cohort is Postal Realty Trust at $832 million market cap. The largest is NVIDIA at $5.22 trillion. The cohort spans 6,277x in market cap.
The highest cohort-average scorer is Neptune Insurance Holdings (NP) at 88.9 across all seven frameworks. The lowest is Sprott Inc. (SII) at 64.0. That spread (24.9 points) inside the 7-of-7 cohort is large, and it is a useful reminder that "passes all 7" is a bar, not a ranking. Two stocks both pass every framework, but one is rated significantly higher on average than the other.
Per-stock verdicts and full per-framework breakdowns for any of the 47 are available through the public verdict endpoint, for example /api/public/verdict/NVDA or /api/public/verdict/MNST. The same data is rendered on the human-readable pages at /buffett/NVDA/, /buffett/AAPL/, and the full set.
What this means for investors
Three observations, framed as observations and not advice.
Observation 1: cross-framework consensus is a strong signal, but it is not a buy list. A stock that passes all 7 frameworks is one that satisfies the published criteria of seven separate value-investing approaches. That is genuinely informative, because the frameworks were not co-designed and they emphasize different things (Graham's balance sheet, Buffett's moat, Lynch's growth-at-a-reasonable-price). When all 7 agree, it is unlikely to be a coincidence. But valuation moves daily, and a stock that passes all 7 today can fail Greenblatt in a month if its earnings yield compresses, or fail Lynch if its growth slows. The 47-stock cohort is a snapshot of cross-framework agreement, not a recommendation.
Observation 2: the consensus cohort is concentrated in expensive tech. Many of the 47 names trade at premium multiples (NVDA, AVGO, KLAC, APP, FICO). They score well because their underlying business quality is exceptional, not because they are cheap by classical Graham standards. An investor who only owns the 7-of-7 cohort is implicitly making a bet that quality compounds even at premium entry multiples. That has worked well for the last decade, in the same way Buffett's evolution from cigar butts to compounders worked well after 1972. It is worth remembering that classical Graham investors would not own most of this cohort.
Observation 3: the histogram tells you to be picky. Two-thirds of US-listed stocks pass zero value-investing frameworks. That is not a glitch in the data. It is a structural feature of public markets: most listed companies are not high-quality compounders, do not trade at defensive-investor multiples, and do not have the earnings stability or balance-sheet strength to clear any of the seven frameworks. If you are a retail investor with limited time, the implication is that the bar for adding a name to your portfolio should be high, and that the universe of names worth owning is probably smaller than your intuition suggests.
This is not investment advice. The frameworks measure documented criteria from published books and letters. Whether the criteria still produce excess returns going forward is an empirical question that no historical analysis can settle.
How to reproduce this study
Everything in this post is reproducible from the public surface of invest-like.com.
- Per-stock verdicts:
/api/public/verdict/[ticker] returns the full 7-framework breakdown for any ticker (score, grade, verdict, per-criterion pass/fail) in JSON. No authentication required.
- Cross-framework consensus endpoint:
/api/public/consensus returns the top stocks by consensus score across all 7 frameworks.
- Latest verdicts feed:
/api/public/latest-verdicts returns the most recent scoring runs across the universe.
- Methodology hub: /methodology/ documents each framework's criteria and scoring function.
- OpenAPI spec: /api/public/openapi.json covers the full public API surface.
- Bulk machine-readable index: /llms-full.txt contains a flat dump of the most important fields for LLM consumption.
If you replicate the methodology and find different numbers, we want to hear about it. The frontier on this kind of cross-framework analysis is empirical and there is no canonical answer. Email zaid@invest-like.com with your methodology and we will publish a correction or comparison post.
FAQ
Why 7 frameworks and not 4 or 10?
We score every stock against the 7 named-investor frameworks where the criteria are sufficiently documented to implement deterministically. Buffett, Graham, Fisher, Lynch, Greenblatt, Munger, and Smith all have published books or letters that define their selection criteria in enough detail to implement. We have not added Lou Simpson, Walter Schloss, Seth Klarman, Howard Marks, or Mohnish Pabrai because their criteria are less precisely documented or are more concentrated portfolios than diversified screens. The list may grow over time. The criteria for each of the 7 are public at /methodology/.
Why is the apples-to-apples cohort only 6,621 stocks instead of all 12,543?
The 7 frameworks have different data requirements. Munger's filter requires 5 years of capital-allocation history, Smith's framework requires consistent cash-conversion data, Fisher requires R-and-D-to-revenue history. Stocks that lack the underlying fundamentals (recent IPOs, ADRs with incomplete US reporting, micro-caps with sparse FMP coverage) get scored on fewer frameworks. Restricting cross-framework counts to the 6,621 stocks scored on all 7 gives an apples-to-apples comparison. The full universe of 12,543 is used only for context.
Is the 47-stock count going to change?
Yes. As fundamentals refresh quarterly and scoring functions get re-run, individual stocks will cross the threshold in both directions. The exact 47 is a snapshot as of 25 May 2026. The shape of the distribution (the bimodal "pass-zero or pass-all" pattern) is stable across runs we have observed since the production system launched in late 2025.
Why is Graham loosest if Graham is famously strict?
Graham's published criteria are absolute thresholds (P/E under 15, P/B under 1.5, current ratio above 2x, dividend record). When applied to a universe of 6,621 stocks that includes thousands of small-cap banks, REITs, and forgotten cyclicals, a surprisingly large fraction of the long tail clears the absolute thresholds. Greenblatt is stricter because his Magic Formula is a relative rank (top decile of the universe). The relative-rank construction caps the pass rate by design.
Did you include foreign stocks?
The primary invest-like universe is US-listed, but 13.8 percent of the 12,543-stock list are foreign-domiciled issuers with US listings (ADRs and direct US-listed shares from Taiwan, the Netherlands, Canada, Ireland, etc.). They are included in the analysis. Examples from the 47-stock cohort: TSM (Taiwan), ASML (Netherlands), TT (Ireland), SII (Canada).
Can I cite these numbers?
Yes. Cite them as invest-like.com 2026, "We scored 12,500 US stocks against 7 named-investor frameworks", snapshot 25 May 2026. Link target is this page. The dataset is publicly accessible through the API endpoints listed above, and the licence terms are public at /press/. If you want a copy of the underlying snapshot CSV for academic use, email zaid@invest-like.com.
Educational disclaimer
This post is an empirical study of how seven documented value-investing frameworks score the current US-listed equity universe. Nothing in this post is a recommendation to buy or sell any security. Past framework performance does not predict future returns, and a stock passing all seven frameworks at a given snapshot does not guarantee positive future performance. The frameworks measure characteristics that the original investors documented as part of their process, but markets are competitive and the persistence of any factor premium is an open empirical question. Readers should consult a licensed financial advisor for personalised investment advice.
Data snapshot: 25 May 2026 03:30 UTC. Methodology hub: /methodology/. Source data: /llms-full.txt. Public API: /api/public/openapi.json.