Cathie Wood's ARK: disruptive innovation, 190 stocks and Tesla at the top
The best-known growth investor of the 2020s holds nearly two hundred companies in her ETFs. Five quarters of ARK's filings and the idea that cost curves drive technology.
Who she is and why people copy her
Cathie Wood founded ARK Invest in 2014 and became famous when her flagship fund, ARK Innovation (ARKK), rose sharply in 2020. She then became the target of just as much criticism when the fund lost most of its value from the peak of 2021. ARK publishes its trades every day, so her portfolio is one of the most closely watched in the world.
The quarterly 13F is therefore the slowest way to follow ARK, but it has one advantage: it combines all its funds into one report.
The ideas behind the portfolio
ARK's central idea is that technology gets cheaper along a predictable curve. It refers to Wright's law: for every doubling of the cumulative number of units produced, the cost falls by a roughly constant percentage. When costs fall far enough, new products become available to many people, and the companies at the centre of a new platform can grow very fast for years. ARK groups its holdings around a handful of such platforms, including artificial intelligence, robotics, energy storage, genomics and public blockchains.
The approach accepts huge swings in return for the chance of finding the leaders of the next era. It assumes that the market underestimates how fast the change will come, and it pays high prices for expected growth.
What a Form 13F does and does not show
For ARK the 13F lists the holdings of its exchange-traded funds on the last day of each quarter. These funds are transparent and active: they buy and sell every day, and the daily lists are on the firm's website. A quarterly snapshot therefore misses most of the trading. It also shows only US-listed shares, so any foreign-listed shares a fund holds would not appear.
Because ARK's funds sometimes hold the same company in several of them, the report adds them together, which is the number we use.
The portfolio over the last five quarters
From June 2025 to June 2026 the reported portfolio went from $13.6 billion to $15.4 billion, with a peak of $16.8 billion in September 2025 and a low of $12.9 billion in March 2026. It had 191 positions at the start and 190 at the end. It is the least concentrated of the large-company portfolios here: the effective number of positions rose from 31.8 to 39.3 and the ten largest fell from 49.2% to 39.7%.
Tesla was the largest holding in every quarter: 7.2% in June 2025, 9.5% in September and 7.5% in June 2026. Others at the top change: Coinbase was 6.7% in June 2025 and 3.8% in December, and by June 2026 the top included AMD (5.3%), Space Exploration Technologies (5.0%), Tempus AI (3.8%), Robinhood (3.4%), Crispr (3.4%), Shopify (3.1%) and Twist Bioscience (3.0%).
Ten largest positions at the end of the latest quarter (2026-06-30) and the same names at the start of the period (2025-06-30), as a share of the US-listed stocks in the report (190 positions, $15.4 billion in total).
| Position | 2026-06-30 | 2025-06-30 |
|---|---|---|
| Tesla | 7.5% | 7.2% |
| AMD | 5.3% | 2.8% |
| Space Exploration Technologies | 5.0% | — |
| Tempus Ai | 3.8% | 3.5% |
| Robinhood Markets | 3.4% | 5.3% |
| Crispr Therapeutics Ag | 3.4% | 3.6% |
| Shopify | 3.1% | 4.3% |
| Twist Bioscience | 3.0% | 1.7% |
| Alphabet | 2.7% | 0.2% |
| 10x Genomics | 2.5% | 1.2% |
| Ten largest together | 39.7% | 49.2% |
What changed quarter by quarter
In the quarter to September 2025 Tesla rose by 2.3 percentage points while Coinbase fell by 1.9 and Circle by 1.5, and several digital-asset companies, including Bitmine and Bullish, were added. In the quarter to December the portfolio was almost unchanged: the largest moves were Teradyne and AMD up, and Roblox down by 1.8 points. In the quarter to March 2026 Circle rose by 1.2 points, Twist Bioscience by 0.8, and Roku fell by 1.4 and DraftKings by 0.8.
In the quarter to June 2026 a new position in Space Exploration Technologies appeared at 5.0% of the book, together with Cerebras Systems (1.1%) and X-energy (0.6%), while Roku fell by 2.3 points and Circle by 1.5. The changes are small relative to the size of the book, which is typical of a portfolio of nearly two hundred names.
Changes between consecutive filings. A share is measured as a part of the reported stock portfolio; it can move because of price changes too, so read it together with the text.
| Quarter end | Positions | Opened | Closed | Biggest rise in share | Biggest fall in share |
|---|---|---|---|---|---|
| 2025-09-30 | 191 → 196 | Bitmine Immersion Technologies (2.3%), Brera (1.3%), Bullish (1.0%) | Faro Technologies (0.0%), Ansys (0.0%), Brainsway (0.0%) | Tesla (+2.3 pp) | Coinbase Global (−1.9 pp) |
| 2025-12-31 | 196 → 196 | Weride (0.1%), Netflix (0.1%), Atai Beckley (0.1%) | Reddit (0.2%), 3iq Ether Staking ETF (0.2%), Ibotta (0.2%) | Teradyne (+0.8 pp) | Roblox (−1.8 pp) |
| 2026-03-31 | 196 → 182 | Broadcom (1.0%), Eli Lilly & (0.1%), Biontech Se (0.1%) | The Trade Desk (0.7%), Pagerduty (0.7%), Unity Software (0.3%) | Circle Internet (+1.2 pp) | Roku (−1.4 pp) |
| 2026-06-30 | 182 → 190 | Space Exploration Technologies (5.0%), Cerebras Systems (1.1%), X-energy (0.6%) | Brera (0.1%), Intercontinental Exchange (0.0%), Palo Alto Networks (0.0%) | Alphabet (+1.1 pp) | Roku (−2.3 pp) |
What to take away, and what to doubt
The idea that costs fall along a curve, and that investors underestimate the long-term effect, is thoughtful. The portfolio is also open and diversified inside its theme, with no single position above about 10%. But a theme is not diversification: all these companies depend on the same mood about growth and interest rates, and they tend to fall together, as they did after 2021.
Prices matter, too. A great technology can be a poor investment if the price assumes everything goes right. The report shows what ARK holds, not whether it is a good time to hold it.
Where it meets our bots
The growth-technology theme has several mechanical relatives in our bots. The Nasdaq 100 trend bot holds the technology index only while the market is above its average, and the bot with the triple-leveraged Nasdaq fund, the riskiest we run, shows what the same theme looks like when leverage is added. The bot that buys the ten strongest stocks of the year often holds the same kind of company.
All accounts are virtual and public. The guides below explain the rules and where they fail.
Related strategy guides
Nasdaq 100 and the 100-day moving average
TQQQ with a 200-day moving average filter
Momentum stocks: the ten strongest of the year
Portfolios
Portfolio builder
Set the shares of stocks, bonds, gold or take famous models, mix them, and see how such a portfolio lived through every era of history.
Goal finder
Say what you need — for example, 6 % a year over 10 years with a drop no deeper than −25 % — and see which models and mixes met it, and in what share of historical periods.
Retirement: how much to withdraw
The 4 % rule tested on 60 portfolios: which share of the money you could take out every year, and when it failed.