A $59 billion dream: How did the female Buffett fall from grace?

At her most successful moment, she was destroyed by her own success.

Byline: Whirlwind Charge, Deep Tide TechFlow

In February 2021, Cathie Wood—known in the market as the “Wooden Sister,” among other monikers—was standing at the highest point of her life.

With assets under management of $59 billion, Bloomberg had just named her the best stock picker of the year, and a reporter from The New York Times called to ask what she thought about “becoming the Millennial Buffett.” Someone on Reddit turned her photo into a meme with the caption, “She sees the future we can’t see.”

Retail investors flooded in like crazy, and her flagship fund ARKK saw net inflows of more than $1 billion in a single day.

Nobody thought this would end.

Today, the $59 billion is down to less than $14 billion—overall size has fallen by 75%.

The media that once crowned her the “female stock god” has started calling her a “one-hit wonder,” and her fans from back then now wonder how the Wooden Sister—the female stock god who once dominated the scene—was stripped of her mystique and stepped down from her pedestal.

This story is far more complex than “she lost her bets.”

From obscurity to the pedestal

ARK’s early days were not easy.

That was in 2014, when quantitative investing was sweeping through Wall Street and passive index funds were the new favorites of all rational investors. The Wooden Sister insisted on going against the trend, betting on tech companies that were “burning money but had a future”: Tesla, gene editing, industrial robotics, and blockchain.

ARK’s initial AUM was still under $100 million. The Wooden Sister herself paid out of pocket to keep operations running. The old-money crowd on Wall Street looked at these holdings and dismissed them: this wasn’t investing—it was gambling.

She did something nearly unheard of on Wall Street: she published the entire research process, updated her holdings every day, and let anyone see in real time what she was buying and why. The team recorded videos on YouTube explaining the logic behind each investment. In an industry where information asymmetry is the lifeline, this was a kind of near-crazy transparency.

From 2014 to 2020, ARKK’s annualized return was close to 39%, more than three times the S&P 500 in the same period. But nobody cared—the scale was too small, and the market was too noisy.

The real turning point came from a disaster.

In March 2020, US stocks crashed 34% in 33 days, setting the fastest bear market on record. Almost every fund manager was cutting positions, waiting, and praying.

Going against the trend, the Wooden Sister added aggressively. She doubled down on Zoom, Teladoc, and Roku. The logic could be summed up in a single line: a virus won’t destroy tech—it will accelerate it.

She bet right.

ARKK rose 152% for the year.

On Reddit and Twitter, her name appeared in conversations among young people who never used to follow financial news. Retail investors discovered something extraordinary: her holdings were public, so you could copy her moves directly—and she was going up.

Believers began to pour in. By the end of 2020, ARKK had become the world’s largest actively managed ETF. By February 2021, the total size of her funds had surpassed $59 billion—seven years, from nothing, to $59 billion.

She became the female stock god: a hyper-aggressive, female version of Buffett.

A pedestal has an expiration date

In February 2021, ARKK’s net inflow exceeded $1 billion in a single day. Retail investors swarmed in wildly at the top. That was both the peak—and the first bell of the funeral, after which the plot turned sharply downward.

The Federal Reserve began signaling rate hikes. The market’s nerves snapped tight. Once interest rates rose, those high-growth stocks whose valuations were being supported by “future earnings,” would face a devastating repricing.

The companies in ARKK’s holdings were all built on the same model: currently losing money, future profitability, and valuation supported by belief.

Belief is the most fragile asset.

From 2021 to 2022, ARKK fell to nearly 75% down.

Zoom dropped from its $559 peak back to $70. Teladoc fell more than 95% from its high. Roku crashed, Unity crashed…

The retail investors who had been flooding WallStreetBets with rocket emojis under her name watched the numbers in their accounts shrink by half within a single quarter. Post titles changed from “ARKK to the moon” to “I’m ruined.”

The redemption wave arrived on schedule. Panic can feed on itself; outflows forced her to sell holdings at low levels. Those sales pushed net asset value down further, and the decline in net asset value triggered even more redemptions.

Morningstar later did the math: over the decade ending at the end of 2023, because of the huge rush of retail investors at high levels and mass selling at low levels, the ARK funds collectively destroyed more than $14 billion in shareholder value. This number doesn’t measure the decline in fund net asset values—it measures the money that real investors actually lost due to the wrong timing. That’s why ARK was dubbed the “largest wealth-destruction” fund family.

At nearly $50 billion in size, by March 2026 it had shrunk to about $13 billion.

Most of the market’s explanations for the Wooden Sister’s defeat boil down to the same layer: rising interest rates crushed growth stocks—she lost, that’s all.

The real problem is hidden deeper.

Running the playbook like a VC in the secondary market

The Wooden Sister’s holding philosophy has never been “pick the best companies.” Her approach is “buy the whole industry before there’s a winner in the race.”

In gene editing, she held CRISPR Therapeutics, Editas Medicine, and Beam Therapeutics—all of which are competitors with one another—so they were all pressed down together. In autonomous driving, she held Tesla, Luminar, and Aurora at the same time.

This logic has an official name: venture capital, or VC.

The underlying logic of VC is to invest in 100 companies; 95 will go under—it’s fine. As long as at least one of the remaining five produces an Airbnb-like outcome, the whole ledger is a win. A high failure rate isn’t a flaw—it’s a cost that the strategy itself must be able to bear.

That logic is a given in the primary market. Startups aren’t traded in public markets, so there’s no “market consensus” embedded in the price—only your view of the future. The loss of the losers is locked into the books and doesn’t affect other holdings, nor your daily liquidity.

Cathie Wood took this logic, unchanged, and brought it into the secondary market. The issue is that the secondary market has something that doesn’t exist in a VC world: real-time pricing.

For every stock you buy, the price already contains the market’s collective judgment of its future. At the peak, Teladoc’s market cap exceeded $40 billion. That wasn’t because it had already earned $40 billion—it was because countless people believed it could earn that amount later. When that “belief” starts to wobble, $40 billion can evaporate into $2 billion within a few quarters. That loss is real and immediate—no “hundred-bagger” stock can fill the hole.

VC losers don’t show up in the income statement. Secondary-market losers, however, are watching your net asset value drop every single day.

These are two entirely different games. She came into the secondary market carrying a VC script.

So why did she win in 2020?

Because 2020 was an extremely rare special window in human history. In that window, the VC logic briefly came alive in the secondary market.

Reconstruct the conditions back then: the Federal Reserve drove interest rates to zero, making the present value of all future cash flows huge and systematically lifting high-risk assets; the pandemic forcibly moved daily life online, turning demand for Zoom and Teladoc from “optional” overnight into “essential”; and most crucially, at that time, the winners of the AI era, the gene-editing era, and the autonomous-driving era had not yet surfaced.

Nobody knew that Nvidia would be the super winner of the AI era. That kind of uncertainty is the survival soil for a VC-style strategy of casting a wide net. When a sector has no clear winner, it’s reasonable to diversify across the whole industry—even in the secondary market.

The Wooden Sister won. The reason she won was “at this moment, there is no answer,” not “she found the answer.”

It was like taking a limited-time open-book exam. After the exam ended, the paper was collected. Yet she believed it was permanent truth, treating this method like a disruptive investing discovery. The bigger she made the scale, the louder the narrative became.

The cruelest irony

This is the most heartbreaking part of the story—and the true key to understanding the Wooden Sister’s fate.

The AI era truly arrived: Nvidia’s market cap broke through $1 trillion, then $2 trillion, then $3 trillion. This is exactly the future Cathie Wood had been predicting for years: AI would reshape everything.

Early 2023, ChatGPT detonated around the world. Every tech company was scrambling to buy GPUs. Cathie Wood stood in front of TV cameras and said, “We started researching AI back in 2014.”

ARK was indeed one of the earliest institutions to take a systematic bullish view on AI. Their Big Ideas reports, year after year, wrote about how AI would change the world. From the timeline, she was a pioneer.

But pioneers aren’t necessarily the big winners.

Because the way the AI era cashes out is completely the opposite of what VC logic needs. VC logic requires diversifying winners, requires market chaos, requires that nobody knows the answer. The market in 2020 met those conditions. But after 2023, the AI wave wasn’t like that.

Its cash-out is winner-takes-all.

Nvidia monopolized computing power, capturing almost all of the excess profits in the AI infrastructure layer. Microsoft, by betting on OpenAI, locked in the application-layer entry point. Meta, Google, and Amazon carved up the remaining share through their respective ecosystem moats. Excess returns were highly concentrated in these names—and every one of them was a large-cap blue chip.

In 2023, Nvidia rose 239%. The “Magnificent Seven” contributed the vast majority of the S&P 500’s full-year gains.

That’s precisely something the Wooden Sister couldn’t do—or more accurately, something she chose to give up.

In fact, ARK was one of Nvidia’s earliest institutional investors. In 2014, when Nvidia was still viewed by the market as a “gaming graphics card company,” Wood began building a position. If she had kept holding, it would have been the greatest trade in ARK’s history.

But she didn’t hold on.

By late 2022, when Nvidia’s stock price plunged due to a collapse in crypto mining and periodic concerns, ARK started selling heavily. In January 2023, the flagship fund ARKK completely exited Nvidia. The remaining positions in other funds were also gradually reduced over the following year. Wood’s stated reason was: Nvidia was a “stock with a strong cyclicality,” and ARK needed to move capital into more “disruptive” AI targets.

Then ChatGPT detonated the world. Nvidia climbed from the price levels at which she had fully sold—up to a trillion-dollar market cap, then two trillion, then three trillion. According to Business Insider, selling Nvidia too early caused ARK to miss out on more than $1.2 billion in returns.

Her entire methodology was “don’t pick winners—buy the whole theme.” But Nvidia had already been in her hands. She selected the winner—and then, because of her own methodology, sold the winner with her own hands, replacing it with a bunch of mid- to small-cap companies that “might benefit from AI.” UiPath, Twilio, Unity—they’re indeed related to AI, just as a creek is connected to the ocean. But when capital’s flood surged straight toward Nvidia and Microsoft, the creek couldn’t share in the water.

Meanwhile, the losers inside the “VC portfolio” began to show their true faces. Teladoc fell 98%. During the pandemic window it was treated as “the future of telemedicine,” but once the window closed, the market discovered it had neither a monopoly position nor profitability. Today its stock price is under $5, leaving behind an increasingly awkward valuation. Zoom returned to the forgotten corner, becoming the quintessential footnote under the label “pandemic beneficiary.” Roku fell more than 80% from its peak.

In the VC ledger, this is called “expected attrition.” In the secondary market, this is called “your principal is gone.”

By the end of 2025, ARK bought Nvidia back during its pullback. By the end of March 2026, she sold again. Within two days she dumped more than 210,000 shares, worth about $37 million. Buy and sell, sell and buy. Nvidia stayed, in her hands, a “trade,” not a “belief.” And yet—ironically—the price curve that the AI era assigned to this stock is exactly one that requires belief in order to hold.

This is the cruelest irony: she was one of Nvidia’s earliest believers. She predicted a correct future with precision. Then, on the eve of that future cashing out, she refunded the ticket herself—because “this stock is too cyclical; I want to ride a more disruptive ship.”

The hunter becomes the hunted

There’s another thing that made the situation irrecoverable.

A true VC can quietly build a position, or quietly exit—nobody is watching every trade you make. But ARK, as a publicly traded ETF, discloses holdings daily, and each sale is a real-time public signal. When she held a small-cap company for more than 10%—or even 20%—of its float, she couldn’t secretly add more without drawing attention, and she couldn’t quietly leave either. The market watches her every move and runs ahead of time.

With nearly $50 billion in size, she went from hunter to hunted.

VC power comes from being small and fast—completing the layout before consensus forms in the market. When you package VC logic into a public fund of nearly $50 billion, you lose two of VC’s most core weapons at the same time: opacity and flexibility.

Also, her influencer persona—ironically—became a cognitive straitjacket. Let’s call it “anti-consensus addiction.”

Wood’s early success all came from going against consensus. In 2014, when nobody believed in her, she won. In 2020, when everyone panicked, she added, and she won again. Each time it was “the market thinks I’m wrong but in the end I’m right,” it reinforced the same belief loop: consensus is wrong, and I’m right.

In a rising market, this loop is a superpower. In a falling market, it’s a curse.

By 2022–2023, market consensus became large-cap blue chips, earnings certainty, Nvidia, and cash flow. This time, consensus happened to be right. But she had already lost, over eight years of positive feedback, the psychological ability to accept the thought “this time, consensus isn’t wrong.”

The problem is that this “anti-consensus” isn’t only her investment strategy—it’s also her public identity. Big Ideas reports, YouTube livestreams, predictions on Twitter, and regular appearances on CNBC: she transformed herself from “the person who manages money” into “the person who sells stories.”

Stories attract capital. Capital pushes up holdings. Holdings validate the story. The cycle accelerates. In a rising market, this flywheel helps her reach godhood; in a falling market, it pins her down.

Because once you build a brand on “anti-consensus,” you can never truly embrace consensus again.

Sell a “disruptive innovation” stock and the market will say, “she doesn’t believe anymore.” Buy a large-cap blue chip and fans will say, “she changed.” Narrative becomes a golden handcuff. That explains why she repeatedly enters and exits Nvidia—buying is just a quick ride on the momentum, while selling is maintaining the persona. She can’t truly hold Nvidia as a heavy position, because Nvidia is “consensus,” and her entire brand is built on “anti-consensus.” On this stock, brand logic and investment logic collide in a deadly way.

The toolset that made her famous—at her most successful moment—was destroyed by her own success.

Epilogue

In early 2026, the Wooden Sister did something familiar.

She cut back significantly on Roku and Shopify, redirecting the capital into the gene editing track.

ARKK and ARKG together bought nearly 200,000 shares of Beam Therapeutics, added 230,000 shares of Intellia Therapeutics, and also swept 420,000 shares of Pacific Biosciences sequencing equipment, plus 100,000 shares of Twist Bioscience synthetic DNA. From gene therapy to sequencing tools to the synthetic DNA platform, ARKK essentially covered the entire industry chain of this frontier sector.

The familiar recipe: buy the whole industry before there’s a winner in the race.

As always, use VC-style moves to lay out positions in the secondary market.

The Wooden Sister didn’t bet wrong on the future. Gene editing could indeed be the next technology that changes humanity’s fate. AI really did change the world, just as she said in 2014—part of it is cashing out in some way.

But between being right in judgment and truly making money, there’s a long distance. The name of that distance is sometimes called timing, sometimes called structure, and sometimes called temperament.