After crashing in 2022, the Ark Innovation ETF has tripled in the last three years. Fund manager Wood is confident the gains are here to stay, dismissing fears of a bubble.
Among all the growth-minded investors who have seen their portfolios soar through this year’s AI craze, one familiar name has outperformed the rest.
Cathie Wood’s Ark Innovation ETF (ARKK) is up 87.1% in the last year, outperforming every other ETF and mutual fund tracked by the American Association of Individual Investors as of the end of September aside from single-stock funds. Its gains have been largely fueled by AI-related stocks like Palantir Technologies, Advanced Micro Devices, Tempus AI, and its longtime largest holding, Tesla, which Wood tells Forbes is the “largest AI project on earth,” marveling at its work developing robotaxis.
The last time Wood was riding this high following her flagship fund’s 157% return in 2020, her idealism and stubborn conviction soon backfired, leading to a 14% loss in 2021 and a dreadful 67% crash in 2022. Even after it has tripled since then, it remains 42% below its February 2021 peak. The flagship fund has $8.3 billion in assets under management, down from $17 billion at the end of 2020, signaling that most investors abandoned her fund during its slide. Wood dismisses any notion that she is being swept up in another bubble like the Covid-19 stock surge.
“The companies investing in AI are some of the most profitable companies in the world,” says Wood. “The reasoning models are astounding people in terms of how much more they can do if you give these models time. I think there were a lot of people expecting at some point the performance would level out. It’s not leveling out at all.”
Almost every AI stock has performed well this year, but Wood, Ark’s 69-year-old founder, CEO and chief investment officer, has had a nose for the biggest winners even relative to their peers. Ark owns more AMD than larger competitor Nvidia, two semiconductor stocks that make up 4.0% and 1.1% of ARKK’s portfolio, respectively, a prescient choice with AMD doubling in value this year compared to Nvidia’s more modest 36% gain. Wood says AMD’s lower valuation at less than $400 million, while Nvidia’s market cap is $4.4 trillion, makes its growth prospects more appealing, and its superiority in chips with more expansive memory is becoming more of a differentiator.
Palantir has done far better, gaining 337% since last November. Its data analytics technology helps government agencies and commercial customers blend and find patterns in massive datasets, and its 12-month sales have grown 39% year over year to $3.4 billion, with $763 in net profit. But Palantir is the poster child for the AI stock bubble, according to most value oriented investors. Its $430 billion market cap is a jaw-dropping 126 times its sales.
Though Wood has taken profits in Palantir, selling 70% of Ark’s stake since August 2024, it still makes up 4.2% of its flagship fund as its ninth-largest holding.
“If Palantir were not at its current valuation, given its position in the platform-as-a-service space, and we think they effectively own that space, it would be right up there with Tesla [in our portfolio],” says Wood, referencing her firm’s 11.9% holding in the electric vehicle giant. “There’s nothing like it out there.”
That’s high praise from Wood, who has loved and owned Tesla for almost a decade. It was a key contributor to her 2020 performance with its 731% gain that year, then it mirrored her fund’s crash in 2022, losing 68%. Last year, Ark raised its price target for the stock to $2,600 per share by 2029, implying a market value of around $9 trillion. Tesla shares currently trade for $443 and has a market capitalization of $1.4 trillion. Ark estimates that by 2029 86% of Tesla’s earnings will be attributable to its robotaxi business, which it just launched in Austin, Texas this June.
“EVs are one and done—you sell a car and hope the customer comes back in five years, and they’re very low margin,” says Wood. “When analysts look at what robotaxis are, they have to use a different model. It’s more of a subscription or recurring revenue model, and it’s very high margin.”
With $1 billion in ARKK invested in Tesla, the holding is twice the weight of its second largest position, Coinbase, but aside from that investment, Wood generally doesn’t invest very much in $1 trillion market-cap tech giants. Amazon, Meta and Nvidia are all in her portfolio, but not in the top 15.
Avoiding mega tech stocks but doubling down on the smaller-cap stocks turned out to be a grave mistake for Wood in 2022. Virtual health business Teladoc Health lost 88% in 2021 and 2022. Game development firm Unity Software fell 80% in 2022. Bets on biotech and life sciences firms like Ginkgo Bioworks, Exact Sciences, Beam Therapeutics and Intellia Therapeutics all floundered.
“We had no idea that we were going to run into a buzzsaw. Many people think it was interest rates, and sure, all long duration assets got killed,” Wood reflects. “But our bigger problem was the supply chain shocks that lingered for so long. Because what drives our models is unit growth. The more unit growth, the faster costs can decline with new technologies.”
In hindsight, Wood concedes the “correct thing to do” would have been to find shelter in larger-cap innovators that were more resilient to supply chain issues and rising interest rates, but she thinks today’s policy environment is far more favorable to her strategy. A longtime Donald Trump supporter, she applauds his move in January to rescind parts of a Joe Biden executive order that aimed to set some standards for AI regulation. New depreciation schedules in Trump’s One Big Beautiful Bill Act have also eased corporate tax burdens.
“The amount of deregulation that is taking place in this administration is astonishing,” she gushes. “I don’t like tariffs, but I would take tariffs if you also gave me what this administration has given with deregulation and much lower tax rates.”
What investors think of Wood depends largely on when they invested. ARKK’s 31.8% three-year annualized return as of September 30 is better than the S&P 500 Index’s 24.9% figure, but its -0.8% five-year number looks terrible compared to the market’s 16.5% annual gains. But rewinding further, a 15.3% annual return since inception in 2014 is back to outperforming the market by two percentage points. It trails the Nasdaq 100 Index’s 16.9% annualized returns in that span, but it’s narrowing that gap.
It’s hard for any investor to come back from an 80% loss from peak to trough, but the Nasdaq similarly lost close to 80% during the dotcom bust, and has gained about 2,000% in the quarter century since. Wood is hopeful that decades from now, its 2022 collapse will similarly look like a blip on ARRK’s long-term stock chart.