Comparative valuation screening matrix evaluated against legendary investment frameworks.
How does each company stand against each investor's framework? Toggle investors to combine them; the score is the % of selected investors' rules the company passes. A dash (—) means the rule could not be tested for that company.
Fifteen investors are not scored from this data — their frameworks cannot be tested from the table’s fields. A rule that can’t be measured from the table’s data is labelled “Principle” in each investor’s card, never scored by proxy.
| Company (Ticker) | Score |
|---|---|
| Loading market data… | |
Rules scored from this data — only these eight are scored; the rest of each framework is labelled “Principle” in the investor’s card, not scored. Druckenmiller: Momentum > 0 (confirming only); Buffett: ROE ≥ 15%; Tepper: Growth ≥ 10%; Alden: ROE > 0 and Margin > 0; O’Shaughnessy: Momentum in the top 25 of the table — note the table carries 3-month momentum as a proxy for his six-month momentum, and with 19 tickers every company currently passes this leg; Baker: PEG < 1.0 and less than half the median PEG of the expensive AI names (AAOI). Baker compared P/E, not PEG: this is the table’s PEG implementation of his P/E-gap test, not a threshold he stated. Lynch: fast growers only (growth ≥ 20%): PEG < 1.0 = pass, PEG ≥ 2.0 = fail; below 20% growth, or a PEG between 1.0 and 2.0 (fairly priced by his rule), the leg is not graded. Lynch’s rule uses expected growth; the score uses the table’s growth field as the growth input. O’Neil: ROE ≥ 17% — his CAN SLIM leg; it shows — until the fundamentals data flows, like Buffett’s and Alden’s ROE legs. Graham, Greenblatt, Smith and Minervini are not scored yet: their legs need the new columns in the table below (P/E, P/B, earnings yield, ROIC, ROCE, gross margin, F-score, moving averages), which show — until the updater writes those fields. — means a field is missing, so the rule could not be tested.
Every screen above must beat a low-cost total-market index fund after costs and taxes — if it can’t, own the index. (Bogle)
Price is the latest closing price from Alpha Vantage, refreshed each weekday after the 4:00 p.m. ET close (last close: loading…). Market cap is that price × shares outstanding from the latest SEC filing, recomputed daily. PEG, growth, debt, ROE and margin (net profit margin) come from SEC filings via EDGAR and Alpha Vantage overview data, refreshed weekly (last refresh: loading…; fiscal period varies by ticker). Momentum is the 3-month % price return computed from Stooq daily closes, through loading…. Institutional ownership is the author’s estimate, not live data. The value, quality and momentum columns added in November 2026 (P/E, P/B, earnings yield, ROIC, ROCE, gross margin, F-score, moving averages and the rest) show — until the updater writes those fields from SEC filings and price history. A dash (—) means no verified figure is available.
| Company (Ticker) | PEG | Growth | ROE | Margin | Momentum | Score |
|---|---|---|---|---|---|---|
| Select one or two tickers above to populate and grade the comparison matrix. | ||||||
† CIFR — kept in the table although it trades below the Step 0 price floor of $10 ($7.15 at the last close, 12 Nov 2026). ‡ TSM — kept in the table although it is a foreign ADR (Taiwan), excluded by the strict Step 0 universe rule and by Greenblatt. Both are kept by choice — see Step 0 in “Johnny’s Stock Screener: The Stack” below.
Five steps, in order. Screens find candidates, not buy signals. Thresholds are strict on purpose — if you get zero results, loosen one at a time (e.g. ROIC 12%, P/E 25). Data differs by platform (EBIT, EV, ROIC definitions) — compare 2–3 sources before trusting a single number. Not financial advice. This is my system, built from the committee’s frameworks — not the committee’s consensus.
US-listed common stocks. Market cap $1B+; price $10+; average volume 500K+ shares. Exclude financials and utilities from the ROIC/EBIT steps; screen banks and insurers separately on P/B and ROE.
Cheap (Greenblatt, Graham, Lynch): earnings yield (EBIT / EV) 8%+ — Greenblatt ranks; 8% is my cutoff. P/E 20 or less (strict Graham mode: 15 or less). P/B 3 or less (strict Graham: 1.5 or less; P/E × P/B 22.5 or less). PEG 1 or less (Lynch); under 0.5 is a standout.
Quality (Buffett, Smith, Greenblatt): ROIC 15%+ (Smith ROCE tier: 20%+). ROE 15%+. Debt/equity 0.5 or less (mine); net debt / EBITDA 2 or less (mine). Free cash flow / net income 90%+ (Smith “cash conversion”). Gross margin 40%+ (mine; Smith favors high margins). FCF yield 4%+ (mine). Positive earnings every year for the last 5 years (Graham uses 10).
Health: Piotroski F-score 7+ (mine — Piotroski’s own “strong” line is 8–9).
Growth (O’Neil): current quarter EPS growth vs. year ago 25%+; annual EPS growth, 3-year, 25%+; sales growth, current quarter, 20%+ and accelerating (mine on the number); ROE 17%+.
Trend Template (Minervini): price above the 150-day and 200-day MA; 150-day MA above the 200-day MA; 200-day MA rising for 1+ months; 50-day MA above the 150-day and 200-day; price above the 50-day MA; price 30%+ above the 52-week low; price within 25% of the 52-week high; relative strength top 20% of the universe by 12-month return, excluding the last month (proxy for the IBD RS rating of 80+).
Sum the ranks (1 = best) for: earnings yield, ROIC, FCF yield, 12-month relative strength. Take the top 20–30 (Greenblatt style) or top 10 for a concentrated list. Don’t blend 1A and 1B blindly; cheap stocks often fail the trend rules. Pick one lead style, use the other as confirmation.
Read the last 10-K and last earnings call. Why is it cheap or strong? Is the earnings drop or jump one-time? Check insider activity and share count trend (buybacks or dilution). Check debt maturity and cash flow. Estimate intrinsic value; buy only with a 25–50% margin of safety (Buffett/Klarman).
Entry for momentum names: breakout from a tight base on rising volume (Minervini). Stop loss: 7–8% below entry for momentum trades (Minervini); for value, use thesis-based exits (mine). Risk 0.5–1% of the portfolio per position (mine; Minervini uses more). Average loss should be well below the average gain. Rebalance value picks yearly (Greenblatt, O’Shaughnessy).
Track results vs. a total-market index fund after taxes and fees. If you can’t beat it over 3–5 years, shift core money to the index. John Bogle’s Cost Matters Hypothesis: after costs, the average active investor must underperform a low-cost index fund by exactly the amount of the costs. SPIVA Year-End 2025: 89.5% of US large-cap funds underperformed after costs over 15 years; over 20 years, ~93–94% fell short. Buffett’s 2007–2017 bet proved the point: a Vanguard S&P 500 index fund compounded 7.1%/yr against hedge funds, after fees.
Their frameworks, verified against what each investor actually said — and what they agree on.
Holdings figures are from each fund’s latest SEC 13F filing where one exists — Q2 2026 (as of 30 June 2026, filed August 2026). 13Fs show US-listed long positions only: no shorts, bonds, cash or foreign holdings. Private funds, analysts and venture firms have no 13F; their figures are labelled in each card. Q3 2026 filings are due 16 Nov 2026.
| Investor | Style | Time horizon | Key test | Risk stance | Scored? |
|---|---|---|---|---|---|
| Jim Simons | Quant / statistical arbitrage | 1–2 days to 1–2 weeks | Secret signal — Principle | ~$10B capacity cap; never override models | No |
| Peter Thiel | Venture (startups) | — | Proprietary tech ≥10× substitute | Fails any of seven questions — Principle | No |
| Dan Ives | Sell-side analyst | 12 months | Central AI beneficiary — Principle | Thesis breaks; no stated exit rule — Principle | No |
| Stanley Druckenmiller | Macro / top-down | 18 months ahead | Liquidity expanding, 18-month outlook | Liquidity regime turns / thesis breaks | Partly |
| Warren Buffett | Value (moats) | Long-term (holds) | ≥10% pre-tax, very high probability; price < intrinsic value | Dishonest/incompetent management, undue leverage — Principle | Partly |
| George Soros | Macro (reflexivity) | — | Reflexive dislocation (peg/credit boom unsustainable) | Exit on disproof — Principle | No |
| Ray Dalio | Macro / risk parity | — | All Weather 30/40/15/7.5/7.5, rebalanced annually | Late-cycle debt avoidance; underweight bonds | No |
| David Tepper | Distressed value | — | Distressed claims at 12–20¢ on the dollar | Valuations far ahead of floor — Principle | Partly |
| Seth Klarman | Deep value | — | Margin of safety, up to 50%+ by uncertainty | Sell at intrinsic value; leverage 0 | No |
| Lyn Alden | Macro (fiscal dominance) | — | Deficits 5–7% GDP, debt >100% GDP; three pillars | Over-owned/overvalued — Principle | Partly |
| Jordi Visser | AI macro | — | Hyperscaler capex rising; company is a capex receiver | Crowding; debt-funded capex (capex > OCF) | No |
| Patrick O’Shaughnessy | Quant (factors) + venture | ≥1 year | Cheapest decile + top 25–50 six-month momentum | Quality-screen fail / faded factor rank | Partly |
| Gavin Baker | AI-stack / bottleneck | — | Mid-single-digit P/E vs peers ≥40x (P/E gap) | Debt-funded buildout; TSMC ends wafer shortage | Partly |
| James Lavish | Macro / Bitcoin | — | Reserves <10% GDP + RRP drained | No stated exit test — Principle | No |
| David Sacks | Venture (SaaS) + policy | — | CMGR ≥15%/≥10%; retention ≥100%; gross margin ≥75% | Burn Multiple >2; retention <100%; margin <75% | No |
| Marc Andreessen | Venture (startups) | — | Software company attacking an uneaten industry — Principle | No extreme strength / check-box deal — Principle | No |
| Howard Marks | Credit / distressed debt | — | Price below intrinsic value, margin of safety — Principle | Euphoria / thin risk premiums / no perceived risk — Principle | No |
| Peter Lynch | Growth (fast growers) | Years (10–20-year story) | Fast grower, PEG < 1.0 at 20–25% growth | PEG ≥ 2, growth slows sharply, or story breaks | Partly |
| Benjamin Graham | Value (defensive) | — | P/E ≤ 15, P/B ≤ 1.5; net-net ≤ ⅔ NCAV | Price above Graham Number; fails strength tests | Partly |
| Joel Greenblatt | Quant (value) | 1 year (hold) | Top 20–30 by combined rank | Falls out of top ranks at rebalance | Partly |
| Terry Smith | Quality | Years (holds) | High ROCE, ~100% cash conversion | ROCE fades; cash stops converting | Partly |
| William O’Neil | Growth (momentum) | Weeks–months | CAN SLIM: EPS +25%, RS 80+ | Cut at −7–8% | Partly |
| Mark Minervini | Growth (momentum) | Weeks–months | All 8 Trend Template rules + VCP | Cut at −7–8%; trend break | Partly |
Framework details are from an audit of each investor’s own statements (books, letters, interviews, filings), Nov 2026. “Principle” means the investor stated it but it has no measurable number — it is labelled, not scored.
James Simons (1938–2024), mathematician and former code-breaker, founded Renaissance Technologies; its Medallion Fund averaged about 66% a year gross and about 39% net, 1988–2018. He replaced narrative stock-picking with scientific, computer-driven searches for small, brief price anomalies.
Principle (not measurable): buys when a historical price pattern recurs beyond chance — the signal, threshold and test are secret, so no outside number can be stated.
Secret signalEdge: a tiny statistical edge — Renaissance co-CEO Robert Mercer put the hit rate at 50.75% — repeated over 150,000–300,000 trades a day, with positions held 1–2 days to 1–2 weeks.
50.75% hit rateSells / avoids when size threatens the edge: Medallion is capped at about $10B, profits are returned annually, outside money was expelled in 2005, and traders never override the models.
~$10B capacity cap| Holding | Ticker(s) | $ value | % of disclosed 13F book |
|---|---|---|---|
| Nvidia | NVDA | $1.42B | 1.95% |
| Alphabet (GOOGL $828.2M + GOOG $556.8M combined) | GOOGL + GOOG | $1,385.0M | 1.91% |
| Meta Platforms | META | $1.14B | 1.57% |
| Intel | INTC | ~$1.02B | 1.41% |
| United Therapeutics | UTHR | $925.8M | 1.27% |
| Palantir Technologies | PLTR | $883.0M | 1.22% |
| Exelixis | EXEL | $730.6M | 1.01% |
| VeriSign | VRSN | $616.3M | 0.85% |
| Kinross Gold | KGC | $584.5M | 0.81% |
| CrowdStrike (new in Q2) | CRWD | $571.6M | 0.79% |
13F-reported holdings of Renaissance Technologies’ institutional equities funds (the Medallion Fund’s holdings are not public), quarter ended 30 June 2026 (Q2 2026), filed 13 Aug 2026. Top 10 of 3,140 disclosed positions; disclosed total $72.6B. 13F filings are reported with a lag and cover US-listed long positions only.
Co-founded PayPal (sold to eBay for $1.5B, 2002) and Palantir (2003); first outside investor in Facebook — $500,000 for 10.2% in 2004, sold for over $1B in 2012. He backs only companies that can become monopolies in a small market first, judged by seven questions and the power law — a venture framework for startups, not a public-stock screen.
Buys when a company's proprietary technology is at least 10× better than its nearest substitute in an important dimension.
≥10× vs substituteEdge: a monopoly that strengthens as it grows — network effects, near-zero marginal cost, and a brand rivals cannot copy.
Principle (not measurable)Sells / avoids when a business cannot answer all seven questions — breakthrough tech, timing, big share of a small market, team, distribution, durability, secret — or must compete on price.
Principle (not measurable)| Holding / stake | Disclosed amount or % | Date of disclosure | Source |
|---|---|---|---|
| PayPal — co-founder; company sold to eBay, stock-for-stock | ~$1.5B (company sale value; Thiel's personal proceeds not disclosed) | Closed 3 Oct 2002 (announced 8 Jul 2002) | SEC Form 8-K, Ex. 99.1 (eBay press release), 3 Oct 2002 |
| Facebook (Meta) — personal, first outside investor | $500,000 for 10.2% (company valued at $4.9M) | Aug 2004 | Nasdaq, 8 Feb 2022 ("Peter Thiel To Step Down From Facebook Parent Meta's Board") |
| Facebook (Meta) — exit | $638M (16.8M shares sold at May 2012 IPO) + $395.8M (Aug 2012, after lock-up) = $1.034B total | May / Aug 2012 | Nasdaq, 8 Feb 2022 |
| Palantir — co-founder and chairman; personal stake | 29.8% of Class B shares (Thiel personally); Founders Fund 12.7% of Class B; three founders collectively 30.2% of stock | Aug 2020 (S-1, holdings at Jul 2020) | Palantir S-1, via TechCrunch, 21 Aug 2020 |
| Palantir — personal stake, current | ~2.99%, ~68.87M shares; largest individual shareholder | 6 Apr 2026 | TipRanks / FinanceCharts ownership data (aggregated from SEC filings); consistent with Motley Fool, 27 Aug 2026 (~3%, ~69M shares). Thiel has sold steadily since early 2024, incl. ~2M shares (~$290M) in Mar 2026 (Form 4, 2 Mar 2026) |
| SpaceX — Founders Fund, first check | $20M (first outside institutional investment; Luke Nosek joined the board) | 4 Aug 2008 | SpaceX press release, 4 Aug 2008; Space.com, 5 Aug 2008 |
| SpaceX — Founders Fund, cumulative invested | ~$600M across multiple rounds | Reported Jun 2026 | Bloomberg, Jun 2026 (via TechStartups, 11 Jun 2026) |
| SpaceX — Founders Fund entities + Thiel personal, combined | 427,306,025 shares = 5.5% | Schedule 13G, Aug 2026 (holdings as of 30 Jun 2026) | SEC Schedule 13G, reported by StockTitan / AlphaMaven, 14 Aug 2026; Edge Consultancy, 16 Aug 2026. Aggregates FF II–V, growth and management vehicles, plus Thiel personal entities (Lembas II, IBSY, Barrel Vault LLC, 2025-005 Investments LLC). Note: pre-IPO press (Bloomberg, Jun 2026) put the firm's stake at ~3% — the 13G is higher because it aggregates the funds *plus* Thiel's personal entities |
| Anduril — Founders Fund | $1B in the $2.5B Series G at a $30.5B valuation; largest single check the firm had written at the time | Jun 2025 | TechCrunch, Jun 2025; Bloomberg, 1 May 2026. (Superseded as "largest ever" by the $1.25B Anthropic check in early 2026 — see next row) |
| Anthropic — Founders Fund, first position | $1.25B at a $350B valuation | Early 2026 | Bloomberg, 1 May 2026 |
| OpenAI — Founders Fund | Participant in the $300M+ share sale / tender at a $27–29B valuation; individual check undisclosed | 28 Apr 2023 | TechCrunch, 28 Apr 2023 |
| Stripe, Ramp, Cognition AI — Founders Fund (Growth III) | Backed; amounts undisclosed | 2025–2026 | Bloomberg, 1 May 2026 (Growth III backed 7 companies, avg ~$600M check) |
| Airbnb, Spotify, DeepMind — Founders Fund | Early investor; amounts undisclosed | 2005–2014 era | Founders Fund portfolio (Wikipedia, "Founders Fund") |
| Fund scale (context, not a holding) — Founders Fund AUM ~$17B (2025); Growth III $4.6B (Apr 2025); Growth IV $6B closed 1 May 2026 (largest ever, 4th growth vehicle; $4.5B LPs + $1.5B partners/employees) | — | 1 May 2026 | Bloomberg, 1 May 2026; Wikipedia |
Peter Thiel is a venture investor and files no 13F. Only publicly disclosed stakes are listed, per the sources shown; where a size is not public, the table says so.
Partner and Senior Managing Director at Yorkville Ives (co-founded July 2026 after eight years as Wedbush Global Head of Technology Research); named Nvidia ($300), Microsoft ($615), Palantir ($250), Apple ($400) and CrowdStrike ($335) his Top 5 into 2027 on 7 Oct 2026. He rates stocks on their place in the AI buildout — a sell-side analyst's ratings and targets, not a fund manager's screening rules.
Buys when a company is a central AI beneficiary — chips, cloud, enterprise software or cybersecurity — with accelerating enterprise adoption.
Principle (not measurable)Edge: a platform position in the AI stack — Nvidia compute, Microsoft Azure/OpenAI, Palantir AIP, Apple installed base — judged by channel checks and backlog.
Principle (not measurable)Sells / avoids when the AI adoption thesis breaks — Ives has published no sell or exit rule; his ratings are 12-month Outperform targets, revised when the thesis changes.
Principle (not measurable)Top 5 into 2027
| Rank / name | Rating | Price target | Announced | Source |
|---|---|---|---|---|
| 1. Nvidia (NVDA) | Outperform (shown as "Buy" by Price-Target.com) | $300 | 7 Oct 2026; reiterated 11 Nov 2026 | Yorkville Ives initiation 7 Oct 2026, via Price-Target.com analyst page (accessed 11 Nov 2026); reiteration 11 Nov 2026 per same archive |
| 2. Microsoft (MSFT) | Outperform | $615 | 7 Oct 2026 | Yorkville Ives initiation 7 Oct 2026, via Price-Target.com; MarketWatch 7 Oct 2026 |
| 3. Palantir (PLTR) | Outperform | $250 | 7 Oct 2026 | Yorkville Ives initiation 7 Oct 2026, via Price-Target.com; MarketWatch 7 Oct 2026 |
| 4. Apple (AAPL) | Outperform | $400 | 7 Oct 2026; reiterated 3 Nov 2026 | Yorkville Ives initiation 7 Oct 2026, via Price-Target.com |
| 5. CrowdStrike (CRWD) | Outperform | $335 | Initiated 5 Oct 2026; named Top 5 on 6 Oct 2026 | MarketBeat analyst log: "10/5/2026, Yorkville Ives, Dan Ives, Initiated Coverage, Outperform, $335.00"; Ives X post 6 Oct 2026 (via BeInCrypto 8 Oct 2026); Barchart 13 Oct 2026. Note: Price-Target.com's Yorkville archive does not list a CRWD note — the $335 is corroborated by MarketBeat + MarketWatch + Barchart, not by Price-Target.com. |
Full current call archive
| Company | Rating | Price target | Date of latest note | Source |
|---|---|---|---|---|
| Nvidia (NVDA) | Outperform | $300 | 11 Nov 2026 (reiterated; initiated 7 Oct 2026) | Price-Target.com / Dan Ives page, accessed 11 Nov 2026 |
| Apple (AAPL) | Outperform | $400 | 3 Nov 2026 (reiterated; initiated 7 Oct 2026) | Price-Target.com |
| SpaceX (SPCX) | Outperform | $225 | 26 Oct 2026 (reiterated; initiated 7 Oct 2026) | Price-Target.com |
| Meta (META) | Outperform | $920 | 26 Oct 2026 (reiterated; raised from $900 on 22 Oct 2026; initiated 7 Oct 2026 at $900) | Price-Target.com |
| Tesla (TSLA) | Outperform | $500 | 12 Oct 2026 (reiterated; initiated 7 Oct 2026) | Price-Target.com; Longbridge/Benzinga 9 Oct 2026 ("physical AI" play) |
| Alphabet / Google (GOOGL) | Outperform | $400 | 7 Oct 2026 | Price-Target.com; MarketWatch 7 Oct 2026 |
| C3.ai (AI) | Hold | $12.50 | 7 Oct 2026 (target reached 26 Oct 2026 per archive) | Price-Target.com |
| General Motors (GM) | Outperform | $100 | 7 Oct 2026 | Price-Target.com |
| Microsoft (MSFT) | Outperform | $615 | 7 Oct 2026 | Price-Target.com; MarketWatch 7 Oct 2026 |
| Palantir (PLTR) | Outperform | $250 | 7 Oct 2026 | Price-Target.com; MarketWatch 7 Oct 2026 |
| Oracle (ORCL) | Hold | $155 | 7 Oct 2026 | Price-Target.com |
| Rivian (RIVN) | Outperform | $20 | 7 Oct 2026 | Price-Target.com |
| CrowdStrike (CRWD) | Outperform | $335 | 5 Oct 2026 (see §1a note — not in Price-Target.com archive) | MarketBeat; MarketWatch 7 Oct 2026; Barchart 13 Oct 2026 |
Dan Ives is a sell-side analyst (Yorkville Ives) and runs no portfolio — these are his published ratings and price targets, not holdings.
Stanley Druckenmiller founded Duquesne Capital in 1981 and averaged about 30% a year to 2010, with no losing completed calendar year. He invests top-down, 18 months ahead, on central-bank liquidity — not on current earnings.
Buys when central-bank liquidity is expanding and the outlook 18 months ahead is materially better than today — the direction of liquidity, not a single stock's earnings, is the test.
Liquidity expandingEdge: concentrated sizing only on tremendous conviction with an asymmetric payoff — in 1992 the pound short risked about 0.5–6% against about 15–20% upside, sized to about $10B.
Asymmetric payoffSells / avoids when the liquidity regime turns or the 18-month thesis breaks, cutting the position rather than waiting for current earnings to confirm it.
Thesis broken = exit| Holding | Ticker | Shares (stock) | Stock value | % of disclosed $5.211B | Call notional |
|---|---|---|---|---|---|
| Natera Inc | NTRA | 3,186,306 | $864.9M | 16.60% | — |
| Taiwan Semiconductor Mfg (ADR) | TSM | 589,680 | $281.6M | 5.40% | — |
| STMicroelectronics NV | STM | 3,102,880 | $232.4M | 4.46% | — |
| Insmed Inc | INSM | 1,424,690 | $151.9M | 2.92% | $143.9M |
| Fox Corp (Cl A + Cl B combined) | FOXA / FOX | 2,837,200 | $144.6M | 2.78% | — |
| YPF SA (ADR) | YPF | 3,138,897 | $142.7M | 2.74% | — |
| Amazon.com Inc | AMZN | 541,600 | $129.1M | 2.48% | $109.5M |
| BBB Foods Inc | — * | 2,901,733 | $120.9M | 2.32% | — |
| Alphabet Inc (GOOGL only held) | GOOGL | 336,300 | $120.2M | 2.31% | — |
| iShares MSCI Brazil ETF | EWZ | 3,436,170 | $118.5M | 2.28% | $145.9M |
| Seagate Technology Holdings | STX | 122,000 | $117.7M | 2.26% | — |
| United Airlines Holdings | UAL | 794,795 | $108.1M | 2.07% | — |
| Sea Ltd (ADR) | SE | 1,099,905 | $105.4M | 2.02% | — |
| CDW Corp | CDW | 743,950 | $104.6M | 2.01% | $35.2M |
| NewAmsterdam Pharma Co | — * | 3,070,146 | $104.0M | 2.00% | — |
| SanDisk Corp | SNDK | 34,900 | $79.4M | 1.52% | — |
| Revolution Medicines Inc | RVMD | 399,587 | $74.8M | 1.44% | — |
| Bitdeer Technologies Group | BTDR | 4,074,993 | $64.7M | 1.24% | — |
| CRH plc | CRH | 553,555 | $59.2M | 1.14% | — |
| Delta Air Lines Inc | DAL | 603,000 | $56.5M | 1.08% | — |
13F-reported holdings (Duquesne Family Office), quarter ended 30 June 2026 (Q2 2026), filed 14 Aug 2026 — top 20 stock positions of 95 reported lines; all-rows total $5.211B, stock-only $4.354B. Duquesne’s macro positions are not in the 13F. 13F filings are reported with a lag and cover US-listed long positions only.
Berkshire Hathaway's former CEO, whose per-share market value compounded 19.7% a year from 1965–2025, against 10.5% for the S&P 500 with dividends. He buys understandable businesses with durable advantages, run by honest and competent managers, at a very attractive price — and then holds them.
Buys when he sees a very high probability of at least 10% pre-tax returns, at a price below intrinsic value as measured by owner earnings.
≥10% pre-tax, very high probability (2002 letter)Edge: a moat that shows up as ROE averaging over 20% across 10 years, with no single year below 15%.
ROE >20% avg / ≥15% every year (1987 letter)Principle (not measurable): Sells / avoids when management is dishonest or incompetent, or the business needs undue leverage to earn its returns.
No debt-to-equity threshold stated (1989 letter)| Holding | Value | % of disclosed 13F | Shares |
|---|---|---|---|
| Apple Inc. (AAPL) | $65.95B | 22.04% | 227,917,808 |
| American Express Co. (AXP) | $51.28B | 17.14% | 151,610,700 |
| Alphabet Inc. (GOOGL + GOOG combined) | $37.76B | 12.62% | 105,979,600 |
| Coca-Cola Co. (KO) | $32.51B | 10.86% | 400,000,000 |
| Bank of America Corp. (BAC) | $27.54B | 9.20% | 483,394,015 |
| Chevron Corp. (CVX) | $13.99B | 4.67% | 84,375,856 |
| Occidental Petroleum Corp. (OXY) | $12.87B | 4.30% | 264,941,431 |
| Chubb Ltd. (CB) | $11.67B | 3.90% | 34,249,183 |
| Moody's Corp. (MCO) | $11.17B | 3.73% | 24,669,778 |
| Kraft Heinz Co. (KHC) | $7.69B | 2.57% | 325,634,818 |
| DaVita Inc. (DVA) | $6.43B | 2.15% | 28,880,209 |
| Delta Air Lines Inc. (DAL) | $5.37B | 1.79% | 57,320,000 |
| Sirius XM Holdings Inc. (SIRI) | $3.69B | 1.23% | 124,807,117 |
| VeriSign Inc. (VRSN) | $2.26B | 0.76% | 8,989,880 |
| Kroger Co. (KR) | $2.17B | 0.72% | 39,000,000 |
| Liberty Live Holdings Inc. (LLYVA + LLYVK combined) | $1.62B | 0.54% | 15,573,731 |
| Ally Financial Inc. (ALLY) | $1.24B | 0.41% | 27,000,000 |
| Lennar Corp. (LEN + LEN.B combined) | $1.21B | 0.41% | 13,409,858 |
| New York Times Co. (NYT) | $1.10B | 0.37% | 15,700,000 |
| Capital One Financial Corp. (COF) | $0.60B | 0.20% | 3,000,000 |
| Louisiana-Pacific Corp. (LPX) | $0.45B | 0.15% | 5,664,793 |
| Nucor Corp. (NUE) | $0.41B | 0.14% | 1,857,752 |
| Macy's Inc. (M) | $0.17B | 0.06% | 7,347,426 |
| NVR Inc. (NVR) | $0.08B | 0.03% | 11,112 |
| Jefferies Financial Group Inc. (JEF) | $0.02B | 0.01% | 433,558 |
| D.R. Horton Inc. (DHI) | $0.0006B | 0.00% | 3,564 |
13F-reported holdings (Berkshire Hathaway), quarter ended 30 June 2026 (Q2 2026), filed 14 Aug 2026 — all 26 companies in the filing; disclosed total $299.25B. Cash, bonds and wholly-owned businesses are not in a 13F. 13F filings are reported with a lag and cover US-listed long positions only.
George Soros ran the Quantum Fund to about 30% a year in 1969–2000 ($1,000 to about $4 million), including about $1 billion from the about $10 billion short of sterling on Black Wednesday, 16 September 1992. He bets on reflexive boom–bust cycles — a false belief that moves prices, which then moves fundamentals — sizing hardest when the payoff is asymmetric.
Buys when a trend is still spreading on a false but fertile belief and prices are feeding fundamentals, and the peg or credit boom behind it is demonstrably unsustainable — in 1992, sterling pegged at DM 2.95 while Britain needed rate cuts and Germany needed high rates.
Macro dislocation testEdge: sizes to the payoff, not the probability — Druckenmiller proposed $1.5 billion in 1992 and Soros pushed it to about $10 billion, about 200% of the fund, on about 0.5% downside against 15–20% upside.
Risk/reward about 30:1 or betterPrinciple (not measurable): Sells / avoids when Soros judges the thesis wrong — no fixed stop-loss % exists in his writings; he cuts at once on disproof rather than waiting for vindication.
Exit on disproof| Holding | Ticker | Type | Shares / principal | Value | % of disclosed $8.144B |
|---|---|---|---|---|---|
| Amazon.com Inc | AMZN | Common stock | 1,182,529 sh | $281.8M | 3.46% |
| Alphabet Inc (GOOGL $213.1M + Cl-B depositary shs $60.2M) | GOOGL | Common + dep. shs | 1,796,286 sh | $273.3M | 3.36% |
| Taiwan Semiconductor Mfg (ADR) | TSM | Common stock | 522,158 sh | $249.4M | 3.06% |
| Global Payments Inc | GPN | Convertible note 1.500% | $247.5M principal | $222.7M | 2.73% |
| Nvidia Corp | NVDA | Common stock | 1,064,635 sh | $213.0M | 2.62% |
| Lumentum Holdings Inc | LITE | Convertible notes 0.500% (two issues) | $28.5M principal | $207.5M | 2.55% |
| Electronic Arts Inc | EA | Common stock | 970,990 sh | $199.1M | 2.44% |
| BILL Holdings Inc | BILL | Convertible note + $0.9M stock | $209.7M principal | $182.3M | 2.24% |
| PG&E Corp | PCG | Convertible note 4.250% | $175.0M principal | $179.6M | 2.21% |
| Rivian Automotive Inc | RIVN | Convertible note 4.625% | $131.6M principal | $155.8M | 1.91% |
| Apple Inc | AAPL | Common stock | 529,538 sh | $153.2M | 1.88% |
| Linde plc | LIN | Common stock | 267,349 sh | $138.7M | 1.70% |
| TKO Group Holdings Inc | TKO | Common stock | 578,157 sh | $116.4M | 1.43% |
| Super Micro Computer Inc | SMCI | Depositary shares, Series A (7%) | 2,175,408 sh | $111.5M | 1.37% |
| American Electric Power Co | AEP | Common stock | 768,000 sh | $105.1M | 1.29% |
| Entergy Corp | ETR | Common stock | 913,912 sh | $105.0M | 1.29% |
| Dropbox Inc | DBX | Convertible note | $103.5M principal | $104.6M | 1.28% |
| JD.com Inc | JD | Convertible note 0.250% | $100.0M principal | $98.5M | 1.21% |
| Itron Inc | ITRI | Common stock | 1,137,853 sh | $98.5M | 1.21% |
| Digital Realty Trust Inc | DLR | Common stock | 539,618 sh | $96.9M | 1.19% |
13F-reported holdings (Soros Fund Management), quarter ended 30 June 2026 (Q2 2026), filed 14 Aug 2026 — top 20 of 266 reported lines; all-rows total $8.144B, stock-only $5.855B. The fund’s macro positions are not in the 13F. 13F filings are reported with a lag and cover US-listed long positions only.
Ray Dalio founded Bridgewater in 1975 and built Pure Alpha to about 11.4% a year since December 1991 (Bloomberg, July 2022), including +9.5% in 2008, while All Weather returned about 7.8% a year since 1996 to 2019. He treats the economy as a machine of short-term (5–8 year) and long-term (50–75 year) debt cycles and builds portfolios that survive every regime by balancing risk, not by picking stocks.
Buys when a portfolio can be balanced to equal risk across four regimes — rising/falling growth × rising/falling inflation — at 30% stocks / 40% long-term bonds / 15% intermediate bonds / 7.5% gold / 7.5% commodities, rebalanced at least annually.
Risk parity, not dollarsEdge: holds 15–20 good, uncorrelated return streams, which Dalio says cuts portfolio risk by about 80% without cutting expected return — tested by pairwise correlation near zero, not by position count.
15–20 streams, ~80% risk cutSells / avoids when debt-to-income and debt service show a long-term debt cycle is late and deleveraging is unavoidable — underweight bonds and debt assets, and on 21 August 2026 Dalio put that at 10–15% in gold plus a bit of Bitcoin.
Underweight debt, 10–15% gold| Holding | Ticker | Shares | Value | % of disclosed $24.376B |
|---|---|---|---|---|
| SPDR S&P 500 ETF Trust | SPY | 5,320,308 | $3,973.0M | 16.30% |
| iShares Core S&P 500 ETF | IVV | 3,002,286 | $2,248.4M | 9.22% |
| Nvidia Corp | NVDA | 3,866,195 | $773.6M | 3.17% |
| Alphabet Inc (GOOGL $472.5M + GOOG $26.2M) | GOOGL + GOOG | 1,396,383 | $498.7M | 2.05% |
| Broadcom Inc | AVGO | 1,317,923 | $497.8M | 2.04% |
| Amazon.com Inc | AMZN | 2,025,481 | $482.8M | 1.98% |
| Lam Research Corp | LRCX | 938,224 | $406.6M | 1.67% |
| Vanguard S&P 500 ETF | VOO | 458,381 | $314.8M | 1.29% |
| Advanced Micro Devices Inc | AMD | 538,632 | $312.9M | 1.28% |
| iShares MSCI South Korea ETF | EWY | 1,384,772 | $279.6M | 1.15% |
| Microsoft Corp | MSFT | 711,896 | $265.6M | 1.09% |
| Newmont Corp | NEM | 2,695,170 | $251.7M | 1.03% |
| Applied Materials Inc | AMAT | 291,426 | $210.7M | 0.86% |
| Oracle Corp | ORCL | 1,432,967 | $210.0M | 0.86% |
| Seagate Technology Holdings | STX | 199,822 | $192.8M | 0.79% |
13F-reported holdings (Bridgewater Associates), quarter ended 30 June 2026 (Q2 2026), filed 14 Aug 2026 — top 15 of 997 positions; disclosed total $24.376B. The 13F is only part of Bridgewater’s book, not firm AUM. 13F filings are reported with a lag and cover US-listed long positions only.
Founder of Appaloosa Management (1993), which returned 132% in 2009 on ~$7–7.5B of profit from distressed bank debt and equity, and was up 32% (gross) in H1 2026. He buys distressed claims at a deep margin of safety when a government backstop or seniority bounds the downside, and buys everything when a central bank or government is easing aggressively.
Buys when senior or preferred claims trade at 12–20¢ on the dollar, or equity at $2–3 against a stated government purchase price of $6 — a margin of safety of roughly 50–67% to that floor.
12–20¢ on the dollarEdge: he reads the rescue terms before the market does — the stated government floor is at least 2x the price he pays, so the downside is bounded by seniority and the floor, not by hope.
Floor ≥2x price paidPrinciple (not measurable): Sells / avoids when valuations run far ahead of that floor or a rally becomes extreme — he trimmed Micron 41% in Q2 2026 after a 300%+ H1 run and exited SanDisk after a 591% run — but he states no numeric sell trigger.
No stated trigger| Holding | Ticker | $ value | % of disclosed stock-only portfolio |
|---|---|---|---|
| Amazon.com | AMZN | $1,191.7M | 15.95% |
| Micron Technology | MU | $1,125.4M | 15.06% |
| Taiwan Semiconductor (TSM) | TSM | $788.0M | 10.55% |
| Alphabet — Class C only (no GOOGL line in this filing) | GOOG | $653.7M | 8.75% |
| Uber Technologies | UBER | $555.2M | 7.43% |
| iShares MSCI South Korea ETF | EWY | $489.6M | 6.55% |
| Meta Platforms | META | $380.2M | 5.09% |
| Vistra Corp | VST | $351.4M | 4.70% |
| Nvidia | NVDA | $305.1M | 4.08% |
| NRG Energy | NRG | $257.1M | 3.44% |
| Alibaba Group (ADR) | BABA | $192.0M | 2.57% |
| Boeing | BA | $173.2M | 2.32% |
| Lam Research | LRCX | $165.7M | 2.22% |
| Baidu (ADR) | BIDU | $148.0M | 1.98% |
| American Airlines | AAL | $135.5M | 1.81% |
| Advanced Micro Devices | AMD | $114.7M | 1.54% |
| CoreWeave | CRWV | $107.3M | 1.44% |
| ASML Holding | ASML | $99.5M | 1.33% |
| Broadcom | AVGO | $56.7M | 0.76% |
| Qualcomm | QCOM | $46.2M | 0.62% |
| SpaceX (Space Exploration Technologies) | SPCX | $38.4M | 0.51% |
| Energy Transfer LP | ET | $30.1M | 0.40% |
| MPLX LP | MPLX | $28.3M | 0.38% |
| Whirlpool | WHR | $28.2M | 0.38% |
| Goodyear Tire & Rubber | GT | $10.0M | 0.13% |
13F-reported holdings (Appaloosa Management), quarter ended 30 June 2026 (Q2 2026), filed 14 Aug 2026 — all 25 stock positions; stock-only total $7.47B. Listed options in the filing are not shown. 13F filings are reported with a lag and cover US-listed long positions only.
Founder of The Baupost Group (1982), whose largest partnership returned just over 20% a year net since 1983 with only one losing year at that point — a record that has since cooled to about 4% a year since 2014, with five down years in 44 years. He invests for absolute return by avoiding permanent loss first: buy only at a margin of safety, hold cash when nothing qualifies, and use no leverage.
Buys when price is at least 50% below a conservative estimate of intrinsic value for uncertain or distressed assets, and at a smaller but still significant margin of safety where value is highly certain.
Margin of safety up to 50%+ by uncertainty (1991)Edge: senior claims and distressed debt bought far below recovery value, with no leverage — tested by only five down years in 44 years, the worst −12%.
5 down years / worst −12% (2026)Sells / avoids when price rises to the conservative estimate of intrinsic value, so the margin of safety is gone, or when an investment would need leverage to work.
Sell at intrinsic value; leverage 0| Holding | Value | % of disclosed 13F | Shares |
|---|---|---|---|
| Amazon.com Inc. (AMZN) | $892.3M | 16.48% | 3,743,854 |
| Elevance Health Inc. (ELV) | $493.1M | 9.11% | 1,275,154 |
| Restaurant Brands International Inc. (QSR) | $489.6M | 9.04% | 6,753,112 |
| Alphabet Inc. (GOOG, Class C) | $484.7M | 8.95% | 1,371,931 |
| Ferguson Enterprises Inc. (FERG) | $344.3M | 6.36% | 1,450,594 |
| Genuine Parts Co. (GPC) | $332.3M | 6.13% | 2,816,178 |
| Union Pacific Corp. (UNP) | $322.4M | 5.95% | 1,185,395 |
| Visa Inc. (V) | $240.6M | 4.44% | 701,355 |
| Wesco International Inc. (WCC) | $229.0M | 4.23% | 662,881 |
| Aon plc (AON) | $227.9M | 4.21% | 687,000 |
| Teleflex Inc. (TFX) | $178.6M | 3.30% | 1,409,000 |
| Norwegian Cruise Line Holdings Ltd. (NCLH) | $163.2M | 3.01% | 7,730,000 |
| Molina Healthcare Inc. (MOH) | $144.9M | 2.68% | 633,609 |
| CME Group Inc. (CME) | $136.7M | 2.52% | 619,000 |
| Herbalife Ltd. (HLF) | $121.8M | 2.25% | 9,259,844 |
| Eagle Materials Inc. (EXP) | $119.4M | 2.21% | 530,763 |
| GDS Holdings Ltd. (GDS) | $113.2M | 2.09% | 3,769,818 |
| Americold Realty Trust Inc. (COLD) | $107.4M | 1.98% | 6,830,230 |
| Liberty Global Ltd. (LBTYK) | $94.0M | 1.74% | 8,543,161 |
| Grupo Aeromexico SAB de CV (AERO) | $87.8M | 1.62% | 4,880,188 |
| Axalta Coating Systems Ltd. (AXTA) | $43.6M | 0.81% | 1,274,356 |
| DNOW Inc. (DNOW) | $36.2M | 0.67% | 2,788,000 |
| Pershing Square Inc. (PS) | $12.9M | 0.24% | 392,000 |
13F-reported holdings (Baupost Group), quarter ended 30 June 2026 (Q2 2026), filed 13 Aug 2026 — all 23 positions; disclosed total $5.42B. This is Baupost’s 13F only, not the firm’s whole book. 13F filings are reported with a lag and cover US-listed long positions only.
Founder of Lyn Alden Investment Strategy (2016) and author of Broken Money (2023), with no fund track record — an analyst and author, and a general partner at Ego Death Capital. She invests for the fiscal-dominance regime rather than predicting it, holding a three-pillar portfolio built to survive persistent 5–7% of GDP deficits.
Buys when US fiscal deficits run at 5–7% of GDP with public debt above 100% of GDP, holding profitable equities, commodities/producers and hard monies, and cash equivalents instead of long-duration bonds.
Deficits 5–7% of GDPEdge: the three pillars cover all three regimes — profitable equities for disinflationary growth, commodities/producers and hard monies for inflation or stagflation, and cash equivalents, typically 5–10%, for credit contractions — where a 60/40 has no answer for stagflation.
Cash typically 5–10%Principle (not measurable): Sells / avoids when assets depend on long-duration fiat promises or become over-owned and overvalued — she leans out of what is overbought and into what is cheap, gradually — but she states no numeric sell trigger.
No stated triggerThe three pillars
| Pillar | What it holds | Role |
|---|---|---|
| 1. Profitable equities | High-quality, mostly profitable equities, US + foreign blend | Does best in disinflationary growth |
| 2. Commodities / producers & hard monies | Commodity producers, gold, Bitcoin, energy producers | Does best in inflation / stagflation |
| 3. Cash equivalents | Cash, short-term T-bills; not long-duration bonds | Holds up in disinflationary credit contractions |
Stated positions
| Position / figure | Size | Nature | As-of / source |
|---|---|---|---|
| Cash / liquidity buffer | Typically 5–10% | A tendency, not a rule: "I tend to try to keep that more 5% to 10% level available so that if you have a major sell-off or a major opportunity come your way that you can pull on it" | Bankless podcast, "How to Survive The Gradual Print Era", 16 Feb 2026 (Bankless guest directory; transcript at 01:36:31) |
| Bitcoin — model-portfolio example | 5% | Example, not a rule: "one of my model portfolios this past year had a 5% Bitcoin allocation" | Investors Podcast, BTC008 (Bitcoin Fundamentals), transcript via Podscripts; also, to precious-metals investors: "buy 5% Bitcoin and go to the beach" (Kitco/YouTube transcript, undated in source — treat date as unverified, see §4) |
| Bitcoin — suggested range for liquid net worth | 5% or even 10% of liquid net worth | Her answer to "what % of your wealth in Bitcoin?" in a Peter McCormack interview — a suggestion, not her own allocation | BitcoinWorld summary of McCormack interview (secondary; original date not established — see §4) |
| Gold | Holds; not selling, especially physical gold; "certainly in the hold camp" | Stated position, no % given | Kitco News, 29 Sep 2026 (interview recorded 28 Sep 2026) |
| Gold miners | Sold some mining positions after the 2026 run (gold under $2,000 → peak ~$5,405 Jan 2026, ~$4,194 on 10 Oct 2026); judged the asymmetry gone — upside ≈ downside | A sell, no names or % sold given | Discovery Alert, Oct 2026 ("Lyn Alden Sold Gold Miners, But Her Gold Bull Case Holds"), summarising her recent interview — secondary summary, underlying interview not named/linked; see §4 |
| Bitcoin (personal) | Owns Bitcoin; GP at Ego Death Capital (Bitcoin venture firm) | Stated holding, no size given | Kitco, 29 Sep 2026 |
| Energy producers | "Already positioned for energy"; energy under-owned, 3–5 yr view; plays energy mainly through producer equities; prefers natural-gas producers; holds uranium itself rather than uranium miners | Sector stance, no names/sizes in sources found | Kitco, 29 Sep 2026 ("already positioned for energy"); MacroVoices #386, 27 Jul 2023; 2026 interview transcript (per audit) |
| Latin American banks | Commodity-exposed banks (e.g. Brazil), currency unhedged — "the currency is part of the thesis"; Swiss franc / Norwegian krone as defensive stores of value | Stance, no bank names/sizes in sources found | Discovery Alert, Oct 2026 (secondary summary of her interview) |
| Long-duration bonds | Avoids / underweights — prefers short end of the curve | Stance, no % | June 2026 newsletter; Kitco/YouTube transcript ("I generally like the short end of the curve more than the longer end") |
| Orange Juice (new, Jul 2026) | Co-founded Orange Juice (15 Jul 2026) with Ego Death Capital partners — acquires/improves/permanently holds cash-flowing small/mid-sized businesses, backed by a Bitcoin treasury; raised $40M at launch | A business she co-founded, not a portfolio position | lynalden.com, 15 Jul 2026; Wealthion, 19 Oct 2026 |
Public model portfolio — M1 Finance Newsletter Portfolio (June 2026)
| Slice | Actual % | Target % | Value (as printed) |
|---|---|---|---|
| Dividend Stocks | 26.5% | 27% | $42,404.53 |
| International Equity | 18.2% | 18% | $29,210.13 |
| Commodities (commodity producers) | 13.6% | 12% | $21,808.01 |
| Growth Stocks | 11.4% | 12% | $18,295.90 |
| BIL — SPDR Bloomberg 1–3 Month T-Bill ETF | 5.9% | 6% | $9,474.35 |
| SHY — iShares 1–3 Year Treasury Bond ETF | 4.9% | 5% | $7,883.67 |
| PHYS — Sprott Physical Gold Trust | 4.8% | 5% | $7,802.21 |
| GBTC — Grayscale Bitcoin Trust | 4.4% | 5% | $7,056.70 |
| SLYV — SPDR S&P 600 Small Cap Value ETF | 4.2% | 4% | $6,733.49 |
| SIVR — Aberdeen Standard Physical Silver Shares | 3.8% | 4% | $6,123.74 |
| PPLT — Aberdeen Standard Physical Platinum Shares | 1.9% | 2% | $3,075.11 |
Lyn Alden runs no fund and files no 13F. The pillars and stated positions are as she has stated them (June 2026 newsletter; Bankless, 16 Feb 2026; Kitco, 29 Sep 2026). Model-portfolio values are as printed in her public M1 portfolio (account $159,867.84).
30+ year Wall Street veteran — Morgan Stanley managing director, founding partner of Anchor Point, and Weiss Multi-Strategy president & CIO 2005–2024 — now head of AI Macro Nexus at 22V Research and founder of Visser Labs (2024); his tracked 100-stock AI portfolio was up about 40% in 2026 to October, while his equally weighted 42-token + 4-stock index has outperformed Bitcoin since June 2026. He invests where AI capital expenditure flows — to the physical receivers of that spending — and treats crypto as the payment rails for AI agents.
Buys when hyperscaler AI capex is rising year on year and the company is a receiver of that spending in the bottom layers of the five-layer cake — energy, chips/compute, or networking/storage — rather than a top-layer software business.
CapEx receiversEdge: major benchmarks still weight the last era's software winners, so the physical inputs to AI are underweight until the indexes catch up, a gap he tracks in his 100-stock AI portfolio (up about 40% in 2026 to October) and his thematic basket (up 34% since November 2024).
Benchmark arbitrageSells / avoids when the AI trade is crowded — he began trimming Micron in mid-May 2026 after hedge funds and retail piled in — or when a spender's capex exceeds its operating cash flow and is debt-funded, as with Oracle's $55.7bn capex against $32bn operating cash flow in fiscal 2026.
Crowding / debt-funded capexNamed equity positions
| Position | Action / stance | Size where stated | As-of / source |
|---|---|---|---|
| Micron (MU) | Trimmed / began reducing mid-May 2026 — hedge funds and retail had crowded into the AI trade | No % trimmed stated | Cointelegraph, 27 Oct 2026 |
| Micron (MU) | Sold out around May–June 2026; buying back in from ~Oct 2026 — re-entered around $800; downside "maybe 10%" (HTX summary) / "10% to 20%" (Opening Bell Daily summary of the same appearance); upside 60–80% | No position size stated; entry ~$800 per Phil Rosen appearance | Phil Rosen / Full Signal podcast, 4 Nov 2026 (Opening Bell Daily, 4 Nov 2026; HTX/PANews summary, 8 Nov 2026). Note the two secondary summaries disagree on downside (10% vs 10–20%) — see §4. |
| Micron (MU) | Buying the Nov 2026 selloff — bought Micron most, plus Marvell and Nvidia; first nibbles ~27 Oct 2026, added through following weeks, largest buying week was the recording week (week of 2–6 Nov); anchor = 200-day moving average; Micron went from 200% above its 200-day to 13% below as of Friday's close; concedes Micron could fall toward $500 before bottoming; "not underweight time and price" | No $ size stated | Visser on his own podcast, via BigGo Finance, 8 Nov 2026 |
| Marvell (MRVL) | Bought in Nov 2026 selloff; recycled proceeds from selling other optical names at a tax loss into Marvell after it fell 40% in seven trading days on news of a potential deal, back to roughly its April 2026 price | No $ size stated | Same, BigGo Finance, 8 Nov 2026 |
| Nvidia (NVDA) | Bought in Nov 2026 selloff (alongside Micron/Marvell); "just increased Nvidia" in Aug 2026 | No size stated | BigGo Finance, 8 Nov 2026; Aug 2026 YouTube appearance (see §3d) |
| Eli Lilly (LLY) | Bullish — "purest longevity play"; predicts it "will become the S&P 500's largest company by 2030"; compounded >30%/yr for 10 years; obesity/diabetes drugs non-cyclical, ~10% of 100M obese Americans on treatment; 40% revenue growth, massive beats, raised guidance (GLP-1) | No position size stated; not confirmed whether he holds it vs merely bullish | Phil Rosen / Full Signal, 4 Nov 2026 (Opening Bell Daily; HTX, 8 Nov 2026); Pomp podcast, 31 Oct 2026 (Zortix summary, 1 Nov 2026) |
| Optical names (unnamed) | Sold at a loss for tax purposes in Nov 2026, proceeds recycled into Marvell | Names not given | BigGo Finance, 8 Nov 2026 |
| Oracle (ORCL) | NOT a holding claim — a financing test. Capex $55.7B > operating cash flow $32B in fiscal 2026, $43B debt raised. His test: avoid/flag spenders whose capex exceeds OCF and is debt-funded. Nov 2026 update: Oracle guided to positive free cash flow in 2028 and "massive" positive FCF in 2029, needing only ~$20B additional capital; ~$60B combined equity+debt financing over the past 12 months | — | Cointelegraph, 27 Oct 2026; Phil Rosen appearance via HTX, 8 Nov 2026 |
Crypto
| Position | Action / stance | Size where stated | As-of / source |
|---|---|---|---|
| Bitcoin (BTC) | Largest position (his largest holding); "behaves like a high-beta risk asset"; "to me is not digital gold. It's a digital store of value" (two separate statements — do not merge into one quote) | Largest, no % | Cointelegraph, 27 Oct 2026; Substack "Yes, Virginia, Bitcoin is a Risk Asset"; Coin Bureau interview (per audit) |
| Ethereum (ETH) | Bought — disclosed 9 May 2026 on the Pomp podcast, reported 11 May 2026. Thesis: AI agents "need food… it is tokens"; agents cannot access banking/credit, transact in Ether or stablecoins | No size stated | Cointelegraph, 11 May 2026 (Pomp podcast, Sat 9 May 2026) |
| Solana (SOL) | Held — until August 2026 his crypto holdings were solely BTC, ETH and SOL | No size stated | Cointelegraph, 27 Oct 2026 |
| Broader tokens (from Aug 2026) | Expanded beyond BTC/ETH/SOL from August 2026 into a broader range of tokens; "crypto is his largest allocation" (per Phil Rosen summary); expects 2027 to be a breakout year for blockchain commercial applications/payments | No sizes; no token list published | Cointelegraph, 27 Oct 2026; HTX/PANews summary of Phil Rosen, 8 Nov 2026 |
| 42-token + 4-stock index | Equally weighted index he maintains to track the broader market; has outperformed Bitcoin since June 2026; "plans to hold all its constituents" | Constituent list NOT published; magnitude of outperformance unverified | Cointelegraph, 27 Oct 2026 |
| Privacy theme: Zcash (ZEC), Monero (XMR) | "Invested in" privacy; ZEC and XMR "among projects that fit that investment thesis" — AI makes privacy/identity protection more important | No sizes stated | Cointelegraph, 27 Oct 2026 |
| Dogecoin (DOGE) | Watches as a signal of returning retail enthusiasm, "essential to sustaining a rally" — a signal he watches, not a stated holding | — | Cointelegraph, 27 Oct 2026 |
Jordi Visser runs no fund and files no 13F. These are only the individual positions he has publicly named, with dates; his full tracked 100-stock AI basket is not published.
Patrick O'Shaughnessy is the son of quant pioneer James O'Shaughnessy, ran his father's firm OSAM as CEO from 2018 until its sale to Franklin Templeton (announced 2021, completed January 2022; $6.5 billion AUM, Canvas over $2 billion), and built a second career in venture capital, where his firm Positive Sum put $20 million into Tegus, acquired by AlphaSense for $930 million in 2024. [Quant hat] In public equities he buys only through a systematic value-plus-momentum factor model, and [Venture hat] in private markets he backs early-stage founders pursuing their life's work — two different frameworks, scored separately.
Buys when a stock ranks in the cheapest 10% of its universe on the six-ratio Value Composite Two and among the top 25–50 names on six-month momentum, a measurable margin of safety confirmed by momentum.
Entry · Cheapest decile + top 6-month momentumEdge: a purely systematic six-factor model — value, momentum and shareholder yield to select, financial strength, earnings quality and earnings growth to reject — removes discretionary stock-picking, and its 45-year backtest (1964–2009) returned 21.19% a year against 11.22% for the average stock.
Edge · Six-factor system; backtested 1964–2009Sells / avoids when a holding's financial-strength, earnings-quality or earnings-growth factors flag it as a poor-quality company, or when its factor ranking fades at rebalance — typically holding each name at least one year — and the model, not discretion, decides.
Exit · Quality-screen fail or faded factor rank| Company | What is disclosed |
|---|---|
| Etched | Name only (highlighted portfolio company) |
| Cognition | Name only |
| Base Power | Name only |
| Imprint | Name only |
| Modal | Name only |
| Rogo | Name only |
| Neko Health | Name only |
| Physical Intelligence | Name only |
| Vanta | Name only |
| Bending Spoons | Name only |
| Skyryse | Name only |
| Arc | Name only |
| Flapping Airplanes | Name only |
| Alpaca | Name only |
| January | Name only |
| Ridgeline | Name only |
| Tegus — acquired by AlphaSense | $20M invested (Positive Sum, 28 Sep 2022); acquired by AlphaSense for $930M (announced 11 Jun 2024, closed 8 Jul 2024). No return/multiple was ever disclosed — none may be stated. The $20M was a top-up/bridge round per Inex One, not necessarily his total position |
| Prefect | Name only |
| Canopy | Name only |
| Blues | Name only |
| Invert | Name only |
| Fortell | Name only |
| Inheritance | Name only |
| Entrepreneurs First | Name only |
| Arcana | Name only |
| Eight Sleep | Name only |
| ID.me | Name only |
| Casa | Name only |
| Retro | Name only |
Patrick O’Shaughnessy files no 13F and his personal holdings are not public. Companies are as listed at positivesum.com/portfolio (accessed 11 Nov 2026). Only quantified item: Tegus — $20M invested (PR Newswire, 28 Sep 2022); no return or multiple was ever disclosed.
Gavin Baker is the founder and CIO of Atreides Management, who ran Fidelity's $17 billion OTC fund for eight years (2009–2017), compounding roughly 19.3% a year net of fees and beating over 99% of Morningstar peers. He buys the physical bottlenecks of the AI buildout — watts, wafers and memory — where supply cannot expand to meet demand, and only when the cheap side of the stack is priced far below the expensive side.
Buys when a company in the AI stack trades at a mid-single-digit earnings multiple while same-cycle peers trade at 40x earnings or more — the gap he quantified in May 2026 as DRAM at mid-single digits versus semi-cap equipment at about 40x and optical names at triple-digit P/E.
Cross-sectional valuationEdge: supply and demand learned covering cement and aggregates — he owns the bottleneck (watts, wafers and memory) only while supply cannot expand to meet demand, tested by sold-out capacity, rising prices and full utilization, with every GPU running at 100% in May 2026.
Physical bottleneckSells / avoids when the AI buildout is funded with debt rather than operating cash flow, or when TSMC expands capacity enough to end the wafer shortage — his single watch-item — and he crystallizes a loss when the thesis is disproved, not on price alone.
Funding & capacity| Holding | Form | Value | % of all-rows total ($14.34B) | % of stock-only total ($11.15B) | Detail |
|---|---|---|---|---|---|
| SPCX — SPACE EXPLORATION TECHN CORP | Stock | $4,670,106,641 | 32.58% | 41.89% | $4,670,106,641 stock (27,332,943 sh) |
| QQQ — INVESCO QQQ TR | Put options only | $2,356,480,000 | 16.44% | 21.14% | $0 stock (0 sh) + $2,356,480,000 options (3,200,000 underlying sh) |
| META — META PLATFORMS INC | Stock + call options | $902,824,313 | 6.30% | 8.10% | $226,876,313 stock (402,770 sh) + $675,948,000 options (1,200,000 underlying sh) |
| CBRS — CEREBRAS SYSTEMS INC | Stock + call options | $842,029,006 | 5.87% | 7.55% | $687,329,006 stock (3,110,086 sh) + $154,700,000 options (700,000 underlying sh) |
| MU — MICRON TECHNOLOGY INC | Stock | $821,120,352 | 5.73% | 7.37% | $821,120,352 stock (711,364 sh) |
| ALAB — ASTERA LABS INC | Stock | $587,736,321 | 4.10% | 5.27% | $587,736,321 stock (1,216,795 sh) |
| CIEN — CIENA CORP | Stock | $337,760,862 | 2.36% | 3.03% | $337,760,862 stock (688,521 sh) |
| CRDO — CREDO TECHNOLOGY GROUP HOLDI | Stock | $307,109,056 | 2.14% | 2.75% | $307,109,056 stock (1,129,285 sh) |
| PANW — PALO ALTO NETWORKS INC | Stock | $299,058,853 | 2.09% | 2.68% | $299,058,853 stock (876,954 sh) |
| U — UNITY SOFTWARE INC | Stock | $292,825,507 | 2.04% | 2.63% | $292,825,507 stock (10,245,819 sh) |
| COHR — COHERENT CORP | Stock | $253,289,581 | 1.77% | 2.27% | $253,289,581 stock (642,101 sh) |
| NVDA — NVIDIA CORPORATION | Stock | $249,970,636 | 1.74% | 2.24% | $249,970,636 stock (1,249,291 sh) |
| AMZN — AMAZON COM INC | Stock | $231,101,376 | 1.61% | 2.07% | $231,101,376 stock (969,629 sh) |
| CRWV — COREWEAVE INC | Stock | $211,507,071 | 1.48% | 1.90% | $211,507,071 stock (2,124,845 sh) |
| GOOGL — ALPHABET INC | Stock | $174,502,342 | 1.22% | 1.57% | $174,502,342 stock (488,296 sh) |
13F-reported holdings (Atreides Management), quarter ended 30 June 2026 (Q2 2026), filed 14 Aug 2026 — top 15 shown. The all-rows total ($14.34B) includes $3.19B of options notional — the QQQ put position is a hedge, not a long holding; stock-only total $11.15B. 13F filings are reported with a lag and cover US-listed long positions only.
James Lavish, CFA, is a macro / Bitcoin strategist — not a stock picker — with 30+ years on Wall Street, co-founder and managing partner of the Bitcoin Opportunity Fund, author of The Informationist since 2021, and a Strive (ASST) board director. His approach in one line: own assets no government can print, because unsustainable sovereign debt leaves currency debasement as the only politically survivable exit.
Buys when bank reserves fall below 10% of GDP while the Fed's reverse-repo facility is effectively drained, the signal that quantitative tightening must end and quantitative easing must follow.
Test: reserves / GDP; RRP balance — Oct 2025: 9.71% and ~$5BEdge: Bitcoin's fixed 21 million supply cap, which no government, central bank, corporation or miner can raise, against gold whose supply grows about 1.5–2% a year.
Test: protocol supply cap of 21M vs gold's annual mining growthPrinciple (not measurable): he has stated no sell price or exit condition for Bitcoin — his $400,000–$500,000 mid-2029 figure is a power-law forecast, not a sell rule, and he expects a 60–70% drawdown after any such peak.
No stated exit test; forecast only| Position or call | Stated size / target | Date stated | Source | Status / outcome (as of 11 Nov 2026) |
|---|---|---|---|---|
| Personal Bitcoin holding | Owns Bitcoin; "significantly overweight Bitcoin relative to gold"; "I'm personally heavily weighted toward Bitcoin" | 19 Oct 2025 | The Informationist, "Why QT is Dead and QE is Coming" — https://www.jameslavish.com/p/why-qt-is-dead-and-qe-is-coming | Holding, stated. No coin count or % ever disclosed. |
| Personal Bitcoin — buying history | Got into Bitcoin "late 20, early 21"; "started buying it late 21… we've clearly been buying it at prices far below here" | 30 Oct 2026 | What Bitcoin Did podcast (transcript) — https://podscripts.co/podcasts/what-bitcoin-did/why-bitcoin-could-hit-500k-in-2029-james-lavish | Consistent with a long-held, accumulating position. No amounts given. |
| Personal gold holding | Owns gold; "the safe, proven play… boring, reliable"; "both deserve to be core holdings" | 19 Oct 2025 | The Informationist — https://www.jameslavish.com/p/why-qt-is-dead-and-qe-is-coming | Holding, stated. No ounces or % disclosed. |
| Gold price point | Gold "over $4K per ounce", "rallied over 70% since I first started warning about liquidity issues in January 2025 when it was around $2,500" | 19 Oct 2025 | The Informationist — https://www.jameslavish.com/p/why-qt-is-dead-and-qe-is-coming | Materialised as stated: gold did trade above $4,000/oz in Oct 2025, up ~60–70% from ~$2,500 in Jan 2025. |
| Avoid: cash and long-term bonds | "Not cash… Not long-term bonds — because negative real yields destroy wealth" | 19 Oct 2025 | The Informationist — https://www.jameslavish.com/p/why-qt-is-dead-and-qe-is-coming | Stated avoidance, not a holding. No cash/T-bill allocation ever disclosed. |
| Bitcoin Opportunity Fund mandate | Invests in "public and private opportunities in the Bitcoin Ecosystem"; "Bitcoin-only, because Bitcoin is superior to all other cryptocurrencies"; raising a $100M fund; accredited investors / qualified clients only | Fund site (live, accessed 11 Nov 2026); launch 16 Mar 2023 | https://www.bitcoinopportunity.fund/ ; https://www.prweb.com/releases/the-bitcoin-opportunity-fund-launches-and-sets-out-to-raise-100m-fund-859738149.html | Mandate only. No portfolio holdings, positions, sizes or returns are public. |
| Fund exposure to Strive (ASST) — via Form 3 | Bitcoin Opportunity Fund vehicles held 759,259 + 231,852 + 120,000 ASST Class A shares, plus warrants at $1.35 | 12 Sep 2025 (SEC Form 3) | SEC Form 3, 12 Sep 2025 — carried from prior audit (~/workspace/audit_working/lavish.md); not re-verified in this pass | Only fund position visible anywhere. Dollar value at the time not stated in the audit; do not compute one without a dated ASST price. |
| Personal Strive (ASST) director shares — SALE | Sold 6,000 shares at $30.75 avg = $184,500 on Mon 5 Oct 2026; directly owned 8,815 shares afterwards (−40.50% of his direct position) | 5 Oct 2026 (disclosed in SEC filing, reported Oct–Nov 2026) | https://www.marketbeat.com/instant-alerts/analyst-strives-asst-buy-rating-reiterated-at-b-riley-financial-2026-11-09/ ; https://www.quiverquant.com/news/Strive+Announces+Up+to+%241+Billion+At-the-Market+Offering+of+15.00%25+Series+A+Perpetual+SATS+Preferred+Stock | New finding this pass. Sale is a disclosed fact; the *reason* for his sale is not established (tax-withholding language in the same reports attaches to CEO Matt Cole's same-day sale, not clearly to Lavish's). Do not present a reason. |
| Bitcoin $150,000–$180,000 in 2026 | "I wouldn't be surprised whatsoever to see Bitcoin at $150,000 or $180,000 next year" (on expected liquidity expansion) | 30 Nov 2025 | The Wolf of All Streets — https://podscripts.co/podcasts/the-wolf-of-all-streets/the-4-year-bitcoin-cycle-is-dead-heres-what-replaces-it-james-lavish ; reported 2 Dec 2025 — https://cryptonews.net/news/bitcoin/32074836/ | NOT materialised. BTC ~$74,265 on 11 Nov 2026 (Finnhub, 19:24 UTC; 10 Nov close $73,796). Would need roughly a doubling in the last 7 weeks of 2026. |
| Bitcoin $400,000–$500,000 by early–mid 2029 (base case) | Base case top ~$400K–$500K; "the bottom [was] in… back in that expansion cycle" | 30 Oct 2026 (origin); repeated BMTV 10–11 Nov 2026 | What Bitcoin Did — https://podscripts.co/podcasts/what-bitcoin-did/why-bitcoin-could-hit-500k-in-2029-james-lavish ; BMTV — https://www.youtube.com/watch?v=wVb51WmQ2Os ; report — https://www.tradingview.com/news/stocktwits:6f3622b83094b:0-bitcoin-fatigue-is-real-james-lavish-says-early-holders-have-moved-on-as-young-investors-turn-to-betting/ | Open (forecast to 2029). BMTV's own description now pins it to "a one-standard-deviation move above the [power-law] mean" — this resolves the audit's 1-vs-3 SD discrepancy for the BMTV repeat; the What Bitcoin Did transcript at ~00:32:39 is garbled ("$4,000, $600,000") and should not be quoted for the SD figure. |
| Bitcoin $200,000–$300,000 floor by mid-2029 | "Would be absolutely shocked" if Bitcoin were not in the $200,000–$300,000 range by then | 30 Oct 2026; repeated BMTV 10–11 Nov 2026 | Same sources as row above | Open (forecast). Note: one secondary summary says "at least $300,000" — use the $200K–$300K range from BMTV's own description. |
| 60–70% drawdown after any such peak | "If we do hit that… I do believe we have another regular drawdown and that's going to be… a 60–70 percent drawdown" | 30 Oct 2026 | What Bitcoin Did transcript — https://podscripts.co/podcasts/what-bitcoin-did/why-bitcoin-could-hit-500k-in-2029-james-lavish | Open (forecast, not a sell rule — he states no exit price). |
| Bitcoin $1M by 2030/2031 (earlier call, lowered) | Earlier base case $1M by 2030/2031; $400K–$500K by 2029 is "comparatively lower"; still says Bitcoin is "going to $1 million per coin" eventually | Stated as his earlier call on 30 Oct 2026; $1M "eventually" per Bitcoin Conference graphics Jul–Aug 2026 | What Bitcoin Did transcript (above); TradingView report 31 Oct 2026 — https://www.tradingview.com/news/stocktwits:6f3622b83094b:0-bitcoin-fatigue-is-real-james-lavish-says-early-holders-have-moved-on-as-young-investors-turn-to-betting/ | Open. Present the 2029 call as a *lowered* base case vs the earlier $1M call. |
| "The 4-year cycle is dead" | Liquidity cycle ~6–7 years replaces the halving cycle | 30 Nov 2025 | The Wolf of All Streets — https://podscripts.co/podcasts/the-wolf-of-all-streets/the-4-year-bitcoin-cycle-is-dead-heres-what-replaces-it-james-lavish | RETRACTED by Lavish himself, Apr 2026: "I thought the four-year cycle was dead. I declared it dead last year. I was wrong" (after BTC hit $126K) — https://cryptonews.net/news/bitcoin/32729062/ (19 Apr 2026, reporting AMBCrypto 18 Apr 2026). By Nov 2026 he again maps highs to the halving / four-year election cycle (BMTV). Do NOT present as his current view. |
| "QT is dead. QE is coming. Probably Q1 2026, possibly sooner" | Liquidity test: bank reserves $2.96T = 9.71% of GDP (below Fed ~10–11% "ample" floor); RRP ~$5B | 19 Oct 2025 | The Informationist — https://www.jameslavish.com/p/why-qt-is-dead-and-qe-is-coming | Directionally correct on timing, with a labelling caveat: the Fed ended QT on 1 Dec 2025 and began reserve-management purchases of T-bills on 12 Dec 2025 (~$40B/month at first). The Fed does NOT call these purchases QE. So: QT did end in Q4 2025/Q1 2026 as he predicted; "QE" is his word, not the Fed's. |
| Silver / hard assets (Aug 2022) | "A portion… allocated to hard assets like gold and silver, and also a small percentage in Bitcoin" | Aug 2022 | Carried from prior audit (~/workspace/audit_working/lavish.md) — not re-verified in this pass | Low weight: 2022-vintage, no size, and no silver position has been restated since. Do not present silver as a current stated holding. |
James Lavish’s Bitcoin Opportunity Fund is private, files no 13F and its holdings are not public. These are positions and calls he has personally stated, as compiled in November 2026.
Venture investor and policy advocate — not a stock picker. PayPal's first product leader, then COO (1999–2002, eBay acquired PayPal for $1.5B); co-founded Yammer, sold to Microsoft for $1.2B in cash in 2012; co-founded Craft Ventures in late 2017 (Fund I $350M; $3.3B AUM at Nov 2023); angel investor in Facebook, Uber, SpaceX, Palantir and Airbnb; White House AI & Crypto Czar, 20 Jan 2025–26 Mar 2026, then PCAST co-chair. His testable framework is Craft's published SaaS screen: buy only viral, recurring-revenue software that grows fast, retains customers and burns little per dollar of new revenue.
Buys when a B2B SaaS company shows compound monthly growth of at least 15% below $1M ARR and at least 10% above $1M ARR, with dollar retention of at least 100% and gross margin of at least 75%.
Craft, "The SaaS Metrics That Matter" — Series A barEdge: Bottom-Up SaaS — a product that spreads virally inside companies like a consumer product before any top-down sale, tested by engagement of at least 40% DAU/MAU and by Yammer's own deed: 0 to 5M users and 85% of the Fortune 500 in under four years.
Craft, "Announcing Craft III" (2021); Yammer, 2008–2012Sells / avoids when the Burn Multiple — net burn divided by net new ARR — rises above 2, or dollar retention falls below 100%, or gross margin stays persistently below 75%.
Sacks, Burn Multiple (Apr 2020); Craft SaaS benchmarks| Holding / stake | Disclosed amount or % | Date of disclosure | Source |
|---|---|---|---|
| PayPal — first product leader, then COO (operating role, not a personal stake) | Company sold to eBay for $1.5B | Oct 2002 | SEC Form 8-K, 3 Oct 2002 (see Thiel §1) |
| Yammer — co-founder (with Adam Pisoni) and CEO | Microsoft acquired for $1.2B in cash; Sacks's personal stake / proceeds undisclosed | Agreement 25 Jun 2012; closed Jul 2012 | Microsoft press release, 25 Jun 2012; TechCrunch, 25 Jun 2012; PCWorld, 19 Jul 2012 (close) |
| Angel investments (personal) | Facebook, Uber, SpaceX, Palantir, Airbnb — names only; amounts and dates undisclosed for each | Listed on Craft portfolio page (accessed Nov 2026) | Craft Ventures portfolio page (craftventures.com/portfolio — lists "David Sacks, Angel" for Airbnb, Affirm, Facebook); Wikipedia, "David Sacks" |
| Craft Ventures Fund I | $350M | Announced Jan 2018 (Crunchbase dates close Oct 2017) | TechCrunch, 15 Oct 2019 ("debut fund had closed with $350 million"); Crunchbase News, 16 Oct 2019 |
| Craft Ventures Fund II | $500M | 15 Oct 2019 | TechCrunch, 15 Oct 2019; Craft Ventures announcement, 15 Oct 2019 |
| Craft Ventures III + Growth I | $612M + $510M = $1.12B combined; AUM $2B at the time | Aug 2021 | Craft Ventures, "Announcing Craft III", Aug 2021 |
| Craft Ventures IV + Growth II | $712M + $608M = $1.32B combined; AUM $3.3B at the time — STALE as of Nov 2026, see §4 | 9 Nov 2023 | Craft Ventures (Medium), 9 Nov 2023 |
| Craft Ventures V | Form D filed 7 Aug 2026 (indefinite amount, no capital sold at filing); target ~$1B *reported*, not in the filing | 7 Aug 2026 (filing); 11 Aug 2026 (target reported) | SEC Form D (EDGAR), 7 Aug 2026; The Information, 11 Aug 2026 (via Angel Investors Network) |
| Craft Funds I/II unicorns | Addepar, Bird, BitGo, ClickUp, Multicoin, Neuralink, Pipe, Reddit, Sourcegraph, SpaceX — amounts undisclosed; "over 30 unicorns backed at an early stage" (firm's words) | Aug 2021 | Craft Ventures, "Announcing Craft III", Aug 2021 |
| Craft portfolio — named roles | Led Series A: BitGo, ClickUp; led seed: Eventbrite; Sacks angel: Affirm, Airbnb — amounts undisclosed | Portfolio page (accessed Nov 2026) | Craft Ventures portfolio page (craftventures.com/portfolio) |
| Personal crypto — divested | Sold all personal cryptocurrency (BTC, ETH, SOL) prior to the start of the administration | 2–3 Mar 2025 | Sacks on X, 2–3 Mar 2025 |
| Crypto-related divestment — Sacks + Craft | Divested over $200M of positions related to the digital-asset industry, of which at least $85M directly attributable to Sacks | Memo 5 Mar 2025 | White House Counsel ethics memo, 5 Mar 2025, reported by Fortune, 14 Mar 2025 |
| Retained crypto-company interests — Craft | BitGo <2.5% and Lightning Labs <1.2% of total assets (retained under ethics waivers); total retained crypto-company interests <3.8% of assets | Memo 5 Mar 2025 | White House Counsel ethics memo, 5 Mar 2025, reported by Fortune, 14 Mar 2025 |
David Sacks is a venture investor and files no 13F; he has no public-stock portfolio. Only publicly disclosed stakes are listed.
Co-founded Netscape — IPO August 1995 at $28, closing $58.25 for a ~$2.9B valuation, sold to AOL for $4.2B in stock in 1998 — and Opsware, sold to HP for about $1.6B in 2007; co-founded Andreessen Horowitz in 2009 with a $300M first fund, now about $90B under management, backing GitHub ($100M in 2012, over $1B returned on Microsoft's $7.5B acquisition) and Coinbase (led the $25M Series B in 2013). He backs software companies attacking industries software has not eaten yet, betting the venture power law — a venture investor and policy advocate, not a public-stock screen.
Buys when a startup is a software company attacking a large industry software has not yet eaten, with an extreme strength on one important dimension — his "strength versus lack of weakness" test (2014) — tolerating other weaknesses.
Principle (not measurable)Edge: the power law — about 15 of ~4,000 venture-fundable companies a year reach $100M in revenue and generate about 97% of all venture returns, so he sizes for the outlier and accepts that most bets fail.
15 of ~4,000; ~97% of returnsSells / avoids when a company has no extreme strength and no software edge over its industry, or is a "check-box" deal — reasonable on every dimension, remarkable on none; he states no numeric sell trigger, and venture positions are held to IPO or acquisition.
Principle (not measurable)| Holding / stake | Disclosed amount or % | Date of disclosure | Source |
|---|---|---|---|
| Facebook (Meta) — personal angel check | $500,000 | 2005 | Agreed across sources; audit note `andreessen.md` §3. (Board member since 30 Jun 2008.) |
| Skype + Apptio — a16z's first two investments | Amounts undisclosed; Skype sold to Microsoft for $8.5B | 2009 (Skype exit May 2011) | Wikipedia, "Andreessen Horowitz" |
| Okta — a16z led Series A | $10M (round total, led by a16z) | Jul 2010 | ReadWrite, 14 Jul 2010; Worth, 1 Aug 2019 |
| Airbnb — a16z led Series B | $112M round at ~$1.3B valuation; a16z ~$60M per reported split (~$60M a16z / ~$40M DST / ~$5M General Catalyst) | 25 Jul 2011 | TechCrunch, 24 Jul 2011 (round announced 25 Jul 2011). Note: the $60M split is *reported* ("we're hearing"), not company-stated |
| Airbnb — a16z stake at IPO | Initial stake grown to ~$1B in value at IPO (Dec 2020, $40.6B valuation) | Dec 2020 | CB Insights, "The Future According To Andreessen Horowitz" (2021) |
| Twitter — a16z | $80M | 2011 | Wikipedia, "Andreessen Horowitz" |
| GitHub — a16z Series A | $100M — largest check the firm had written at the time | Jul 2012 | PitchBook, Jun 2018; Crunchbase News, 4 Jun 2018; Worth |
| GitHub — exit | ~13% stake, worth over $1B, on Microsoft's $7.5B acquisition | 4 Jun 2018 | PitchBook, Jun 2018 ("biggest VC winner… 13% stake worth more than $1 billion"); Recode via Crunchbase News ("over $1 billion") |
| Coinbase — a16z led Series B | $25M round (led by a16z; USV and Ribbit participated) | 12 Dec 2013 | TechCrunch, 12 Dec 2013; Coinbase announcement, 12 Dec 2013 |
| Coinbase — a16z stake pre-direct-listing | Largest institutional holder; stake worth >$6B (Jun 2021, on market cap at the time). A ~$9.7B value by 2021 is also cited — see note below | Apr–Jun 2021 | Crunchbase News, 2021 (>$6B, Jun 2021). The $9.7B figure (Wikipedia, citing 2021 reporting) is an *estimate* of value, not a disclosed stake — do not present it as a filing figure |
| Oculus VR — a16z led Series B | $75M (round total, led by a16z with Spark, Matrix, Formation 8; a16z's portion undisclosed) | 13 Dec 2013 | TechCrunch, 25 Mar 2014; GamesBeat, 13 Dec 2013 |
| Oculus VR — exit | Facebook acquired for $2B ($400M cash + $1.6B stock + $300M earnout) | 25 Mar 2014 | TechCrunch, 25 Mar 2014 |
| OpenAI — a16z | Participant in the $300M+ share sale / tender at a $27–29B valuation; individual check undisclosed (a16z was also an earlier investor per PitchBook) | 28 Apr 2023 | TechCrunch, 28 Apr 2023 |
| Flow (Adam Neumann) — a16z | $350M — largest single a16z check at the time; valued Flow >$1B pre-launch | 15 Aug 2022 | New York Times, via TechCrunch, 15 Aug 2022 |
| Flowcarbon (Adam Neumann) — a16z led | $70M round | May 2022 | TechCrunch, 15 Aug 2022 |
| Anduril — a16z co-led Series H | $5B round at a $61B valuation, co-led with Thrive Capital; a16z's portion undisclosed | 13 May 2026 | Anduril Industries announcement, 13 May 2026 |
| Roblox — a16z led Series G | $150M (round total, led by a16z) | 2020 | Wikipedia, "Andreessen Horowitz" (via Almanac) |
| Figma — a16z | $50M Series D (round total) | 2020 | Wikipedia, "Andreessen Horowitz" (via Almanac) |
| Clubhouse — a16z | $12M Series A (2020; $10M primary + $2M secondaries) and $100M Series B (2021); round totals | 2020–2021 | Wikipedia, "Andreessen Horowitz" (via Almanac) |
| Slack — a16z | Investor; amount undisclosed; acquired by Salesforce for $27.7B | Jul 2021 (exit) | a16z portfolio (Wikipedia) |
| Fund scale (context, not a holding) — Fund I $300M (2009); Fund II $650M (Nov 2010); $9B raised Jan 2022; $15B+ raised Jan 2026 across five funds (Growth $6.75B, Apps $1.7B, Infrastructure $1.7B, American Dynamism $1.176B, Bio+Health $700M, +$3B other strategies); AUM >$90B at Jan 2026 | — | 9 Jan 2026 | TechCrunch, 9 Jan 2026; a16z announcement (Ben Horowitz blog, 9 Jan 2026). Later: Form ADV 31 Dec 2025 shows $106.5B regulatory AUM; firm states "over $100B" (30 Apr 2026) — Dealroom, Aug 2026 |
| Crypto + American Dynamism funds (context) — Crypto I $300M (2018), II $515M (2020), III $2.2B (2021), IV $4.5B (2022); American Dynamism I $600M (Apr 2024), II $1.176B (Jan 2026) | — | 2018–2026 | a16z fund history (Wikipedia; audit note `andreessen.md` §1) |
Marc Andreessen / Andreessen Horowitz file no 13F for these stakes. Only publicly disclosed investments are listed; “amount undisclosed” means the investment is public but its size is not.
Co-founded Oaktree Capital Management in 1995 with Bruce Karsh, after running high-yield and distressed debt at TCW (1985–1995) and Citicorp (1969–1985); Oaktree stated $209B under management as of June 30, 2025. He has written around 160 public investor memos since his first in October 1990 — "bubble.com" (January 2000) warned of the dot-com bust — distilled into "The Most Important Thing" (2011) and "Mastering the Market Cycle" (2018). He buys credit at a discount to value when others are fearful, and waits when they are euphoric — a credit and distressed-debt investor, not a public-stock screen.
Buys when price is below his estimate of intrinsic value, with a margin of safety — most often the bonds and loans of troubled companies, bought at a deep discount on the bet that recovery value exceeds price.
Principle (not measurable)Edge: second-level thinking — a view that is both different from the consensus and correct, asking what is already in the price; and risk as the probability of permanent loss, not volatility, so risk is highest when it is least perceived.
Principle (not measurable)Sells / refrains when the pendulum swings to euphoria — prices above value, thin risk premiums, widespread belief there is no risk — and prepares rather than predicts, waiting for the cycle to turn; he states no numeric sell trigger.
Principle (not measurable)Peter Lynch managed Fidelity's Magellan Fund from 1977 to 1990, compounding about 29.2% a year and growing its assets from about $18 million to about $14 billion, beating the S&P 500 in 11 of those 13 years. He buys understandable fast-growing companies at a P/E no higher than their growth rate — a PEG of 1 or less — and holds them until the story changes.
Buys when a fast grower — earnings growing 20–25% a year — trades at a PEG below 1.0, meaning its P/E is below its growth rate; a PEG around 0.5 is very attractive. The PEG test applies only where growth is 15% or more, never to slow growers.
PEG < 1.0, growth 20–25% (1989)Edge: the tenbagger — small, profitable, understandable companies in dull, overlooked industries, bought before Wall Street notices them; his Dunkin' Donuts stake returned 10–15 times his money.
10–15x on Dunkin' Donuts (1994)Sells / avoids when a fast grower's P/E rises to twice its growth rate (PEG ≥ 2), when its growth slows sharply, or when the story behind the buy breaks — and avoids companies loaded with bank debt, which is due on demand and decides who survives a crisis.
PEG ≥ 2 = sell (fast growers)Columbia professor who wrote Security Analysis (1934, with David Dodd) and The Intelligent Investor (1949, revised 1973); taught Warren Buffett at Columbia and ran the Graham-Newman partnership; his best investment was GEICO. He invented security analysis and the margin of safety — "the central concept of investment" — and buys only at a price far below value, never for growth at any price.
Buys when a stock passes the defensive screen — current ratio of at least 2, long-term debt below net current assets, positive earnings in each of the last 10 years, dividends for 20 years, earnings growth of at least a third over 10 years, P/E no more than 15 on 3-year average earnings and P/B no more than 1.5 (P/E x P/B no more than 22.5) — or, for the enterprising investor, when its price is below two-thirds of net current asset value (current assets minus all liabilities).
P/E ≤15, P/B ≤1.5; net-net ≤⅔ NCAVEdge: the margin of safety — "Confronted with a challenge to distill the secret of sound investment into three words, we venture the motto, MARGIN OF SAFETY" (The Intelligent Investor, ch. 20) — a gap large enough that "it is enough to assume that future earnings will not fall far below those of the past."
Margin of safety (1949)Sells / avoids when price rises above the Graham Number — the square root of (22.5 x EPS x book value per share) — or when a holding fails the financial-strength tests; he never pays up for growth and never buys on a forecast.
Graham Number = √(22.5 × EPS × BVPS)Wharton MBA; founded Gotham Capital in 1985 (reported roughly 40% a year net for about 20 years — firm-reported, not audited); Columbia lecturer; founded the Value Investors Club; author of The Little Book That Beats the Market (2005) and You Can Be a Stock Market Genius (1997). He buys cheap, good businesses by formula — ranking every stock on earnings yield and return on capital — and lets the ranking, not judgment, decide.
Buys when a stock ranks at the top of the combined Magic Formula rank — earnings yield (EBIT divided by enterprise value) plus return on capital (EBIT divided by net working capital plus net fixed assets) — taking the top 20–30 names, equal-weighted, excluding financials, utilities, foreign ADRs and micro-caps below $50M market cap.
Top 20–30 by combined rankEdge: cheap plus good beats either alone — his 1988–2004 backtest returned 30.8% a year against 12.4% for the S&P 500 (his published backtest on an academic dataset; it ignored taxes, slippage and survivorship, and live implementations since 2005 have been less dramatic).
Backtest 30.8% vs 12.4% (1988–2004)Sells / avoids when a holding falls out of the top ranks at the yearly rebalance — hold one year, re-rank, rebalance — and never owns financials or utilities, whose accounting makes EBIT and return on capital incomparable.
Rebalance yearlyFormer top-rated City of London bank analyst, fired by UBS in 1992 for Accounting for Growth, which exposed his employer's clients' accounting tricks; launched the Fundsmith Equity Fund in November 2010. From inception to 31 July 2026 the fund returned 593.4% against 525.4% for the MSCI World Index — but it trailed the index over the last five years (5-year return 8.4% to 31 July 2026), and in July 2026 Smith broke from "do nothing" with 51% portfolio turnover in a half-year, rotating into momentum names. He buys good companies, doesn't overpay, and does nothing — except, lately, when he does.
Buys when a company sustains a high return on capital employed — his portfolio averaged 32% ROCE and 64% gross margin in 2024, against 16% and 45% for the S&P 500 — converts profits into cash (cash conversion about 96%), carries little debt, grows by reinvestment rather than acquisition, and trades at a reasonable free-cash-flow yield (4.3% in June 2026).
ROCE ~32%, gross margin ~64% (2024)Edge: quality compounding plus doing nothing — turnover ran about 3% a year for most of the fund's life, so compounding, not trading, did the work — tested by the portfolio's published aggregates against the index every year.
Turnover ~3%/yr (historically)Sells / avoids when return on capital fades, cash conversion breaks, debt rises, or growth needs acquisitions — and avoids banks, commodities, utilities and telecoms, whose returns on capital are structurally low. Note the July 2026 break: 51% turnover in six months, selling Unilever, Nike and LVMH for TSMC, AppLovin and GE Vernova, which his letter attributed to a market "dominated by momentum and passive investing rather than fundamental factors."
ROCE fades / cash stops convertingFounder of Investor's Business Daily (1984) and author of How to Make Money in Stocks (1988); he distilled CAN SLIM from studying the biggest winning stocks of previous decades. He buys market leaders breaking out — strong and accelerating earnings, a new high, rising volume, institutional buying, in a market uptrend — and cuts every loss at 7–8%, no exceptions.
Buys when a stock passes CAN SLIM — current quarterly EPS up at least 25% year on year and accelerating, annual EPS growth of at least 25% over three to five years with ROE of at least 17%, a new product, management or price high, rising volume on up days, a Relative Strength rating of 80 or higher, growing institutional sponsorship, and a market in a confirmed uptrend.
CAN SLIM: EPS +25%, RS 80+Edge: leaders, not laggards — the strongest stock in a leading industry group, bought at the breakout, before the bulk of institutions arrive; the market direction (M) is the veto that overrides the other six letters.
Leaders at breakouts; M is the vetoSells / avoids when a position falls 7–8% below its entry — cut, no exceptions, never average down — or when the market turns down, or the earnings story breaks; he never buys a laggard because it looks cheap.
Cut at −7–8%, no exceptionsSelf-taught trader; won the U.S. Investing Championship in 1997 (155%, reported) and again in 2021 (334.8%, audited, $1M category); author of Trade Like a Stock Market Wizard (2013) and Think & Trade Like a Champion (2016). His SEPA method (Specific Entry Point Analysis) buys only Stage 2 uptrends — a stock passing all eight Trend Template rules, with accelerating earnings and sales, at a precise breakout entry — and cuts every loss at 7–8%.
Buys when a stock passes all eight Trend Template rules — price above the 150-day and 200-day moving averages, the 150-day above the 200-day, the 200-day rising for at least a month, the 50-day above the 150-day and 200-day, price above the 50-day, price at least 30% above its 52-week low and within 25% of its 52-week high, and a Relative Strength rating of 70 or higher — plus accelerating earnings and sales, a catalyst, and a breakout from a tight, low-volatility base (VCP) on rising volume.
All 8 template rules + VCP breakoutEdge: Stage 2 only — the markup phase of the four-stage cycle — with a precise entry at the pivot, so the risk is defined before the trade; "the trend template is a gate: a stock either passes all eight at once or it is off the list."
Stage 2 markup onlySells / avoids when a position falls about 7–8% below entry, when the price breaks below its key moving averages, or when the market turns; he keeps the average loss well below the average gain and never holds a Stage 4 decline.
Cut at −7–8%; exit on trend breakTwenty-three investors, from venture to macro to quant to value — and they agree on almost nothing specific. What follows is what survives the test.
One tension: Lynch is the nearest counter-voice — fully invested, and corrections are buying opportunities. And no large cash percentage is prescribed here: Klarman walked back holding 30%+ cash as "almost certainly" a mistake.
How would this portfolio have performed? Start from equal weight, change the weights, pick a time frame, and compare against an index or a single holding — plus the baskets built from each testable investor framework.
Loading price history…
| Portfolio | Total return | CAGR | Max drawdown | Volatility (ann.) | Sharpe (rf 0%) |
|---|---|---|---|---|---|
| Computing… | |||||
Uncheck to remove a holding. Weights are percentages and are normalised to 100% — the normalised column is what the backtest uses. The dividend column is the holding’s stated annual rate; the returns already include reinvested dividends.
| ✓ | Holding | Weight % | Normalised | Data since | Dividend |
|---|---|---|---|---|---|
| Loading holdings… | |||||
Each basket is an equal-weight portfolio of the stocks that pass that investor’s testable rule, using the same audited rules as the scoreboard above. Toggle a basket to overlay it on the chart.
Method & honesty: total return from weekly adjusted closes (Yahoo Finance), so dividends are reinvested — including the preferreds’ ~12% (STRC) and 15% (SATS) payouts. $10,000, rebalanced weekly to your weights among holdings that have started trading; before a holding’s first trading date it holds cash at 0%. Price history last updated loading…, refreshed each weekday. Past performance is not a prediction of future results.
Every acronym and investing term used on this page, defined in one place. The same definitions appear in a small popup when you hover (or tap) a dotted term anywhere above.