In mid-December, as her ARK Innovation ETF's shares (ARKK) were plummeting, Cathie Wood doubled down on her claim that betting on adventurous, world-shaking technologies would prove a bonanza in the years ahead. In an interview with Bloomberg, the ARK Invest founder and CEO Investors entering at such supercheap prices would make double the sumptuous returns of folks who bought near the peak. As Wood tweeted on Dec. 17: "Innovation stocks are not in a bubble. They are in deep value territory."

Nothing could be more anti-value than the Wood portfolio

Value investors from Warren Buffett to Carl Icahn wouldn't recognize Wood's definition of the V-word. Value stocks are generally defined as those that produce lots of earnings for the dollars you're investing, or sell below what an enterprise owns in plants, buildings, pipelines, or patents, after subtracting debt. In reality, you couldn't invent a more concentrated, anti-value, all-in-for-growth platform than the portfolio at ARK Innovation. In the past few months, the market's been moving strongly to beaten-down, unloved shares in such areas as energy and consumer brands, and shunning the shooting stars of tech that sparkled in the stay-at-home economy. Wood's been suffering because she's not only on the wrong side of the trend but also dangerously leveraged to the riskiest and most glamorous of the go-go stocks, from Tesla to Coinbase to Robinhood. "This is extreme loading of growth versus value," says Jay Ritter, the University of Florida economist who's the nation's leading expert on IPOs. "Wood got lucky putting the fund's eggs in companies with extremely high growth when the market was bidding up their multiples. Most value or growth investors don't go to those extremes, because some of those extremes are scary. Her approach can do incredibly well as long as you're getting all the media attention and the party keeps going." Now the long hangover has begun. The onetime revelers are questioning if the time-tested rules no longer apply, as Wood claimed, or if the warning voices were right all along.

Indeed, a rotation from growth to value is inevitable when the former is wildly overpriced and its rival a bargain, and that's what we're seeing today. But Wood is sticking to the manifesto that wowed the media: The stalwart fundamentals such as price/earnings multiples and how new competitors grab profits from the first movers no longer count. The future belongs to the daring innovators propagating narratives such as a looming explosion in cryptocurrencies, autonomous vehicles, and genomics so stupendous that it almost doesn't matter what you pay for the pioneers.

ARK Innovation has suffered a shocking decline

In the early years, Wood looked like a visionary. And even today, folks who bought shares in ARK Innovation five years ago have pocketed returns of well over 250%, waxing the Nasdaq by at least 80 basis points. But since ARK hit an all-time peak of $152 in early February, it has suffered a fast slide, the result of buying and holding shares so pricey that only unachievable earnings growth could keep them at current levels—let alone leave room to expand. The market started recognizing that profits will fall far short of the fantasy targets in the years ahead. Since the summit early last year, ARK has fallen to $82.45 as of midday on Jan. 10, a drop of 46%. One-third of that pullback has occurred since Wood's tweet putting ARK in the "deep value" zone. ARK missed a historic bash: Since the start of 2021, it has surrendered 33% versus a rise of 25% for the S&P 500 and 15% for the Nasdaq.

Most of ARK's largest holdings have endured steep declines since the start of 2021. Of its top 10 stocks, measured by their share of ARK's total investments, the only winner is Tesla, which gained 34%. Coinbase is more or less flat since going public in April, while Unity Software shed 14%, Block 31%, Twilio 34%, Spotify 36%, and Exact Sciences 45%. Its numbers two, three, and four bets Roku, Teladoc, and Zoom—encompassing 18% of the portfolio—tumbled 60%, 67%, and 52% respectively. You'd think a retreat this sweeping might have taken ARK Innovation from extremely pricey to a good buy. But amazingly, by any conventional measure, it's still shockingly overpriced.

It's fascinating to examine how ardently ARK is sticking to Wood's tech-of-the-future picks, even though so many generate puny profits or, more often, big losses. And most of her choices have been around for a long time, enjoying an ample runway to make money, but continuing to fall short.

I presented a summary of my analysis to a spokesperson for ARK; the spokesperson declined to comment on the story.

Looking at ARK as one big company

For this analysis, I'll look at ARK Innovation as if it were one big holding company—call it ARK Inc.—with investments in a number of enterprises. That's the package investors get when they buy the ETF's shares. I calculated the fund's portion in the profits for each of its holdings by applying ARK's percentage of ownership to that company's total earnings. For example, ARK's $900 million investment in Zoom equates to a 1.7% stake. Since Zoom posted GAAP net profits in the past four quarters of $1.1 billion, ARK's share of those profits is $18.7 billion (1.7% x $1.1 billion). Although ARK holds $1.33 billion in Tesla stock and the EV maker is its biggest investment at 8.5% of the portfolio, its ownership position represents just 0.12% of Tesla's over $1 trillion cap. So ARK's share of Tesla's $3.1 billion in trailing profits tallies to just $3.7 million—keep in mind that Tesla sports a gigantic P/E of 350. Likewise, ARK holds 1.9% of Unity Software, which lost $455 million over the trailing four quarters. Hence, our ARK Inc.'s portion of Unity's deficit is a negative $8.6 million.

I added all the gains and losses to establish total earnings for "ARK Inc." The overall profits, what the winners are making versus the market value of ARK's investment in them, and the huge negative numbers for the underwater players and where they occur, point to a fund that's a world removed from a value play, and that must rise at an incredible pace to go from deeply unprofitable to making the big money needed to revive its flagging performance.

On Jan. 4, ARK harbored stocks in 43 companies carrying a market value of $15.6 billion. We could consider that $15.6 billion as ARK Inc.'s market cap. Over the past four quarters, only 10 or fewer than one-quarter of its holdings registered positive GAAP net earnings. That profitmaking group—led by such winners as Roku (6.2% of the portfolio), Coinbase (5.1%), Block (3.75%), and Crispr (2.8%)—posted overall profits of $133.6 million. Now let's consider what the in-the-black contingent sell for compared with their profits. Of the total holdings of $15.6 billion, the plus members account for $5.8 billion, or 37%. The price/earnings multiple for those positive contributors is a lofty 43, well above the recent Nasdaq average of 36. Keep in mind that before the recent selloff, the P/E of ARK's profit winners was far higher.

That multiple of 43, however, is probably understated. Amazingly, the biggest earnings kick comes from the fund's 5% ownership in Coinbase. Since its debut in April, Coinbase has proved richly profitable, garnering almost $3.5 billion and putting ARK's portion at $44 million. In fact, the crypto exchange alone accounts for one-third of the 10 money-spinners' total earnings.

Look for Coinbase's huge contribution to shrink, says David Trainer, CEO of research firm New Constructs. "They're charging exorbitant fees for trading cryptocurrencies," he told Fortune. "People joke on social media that they spent $10 on crypto and paid $20 in fees. Nobody in brokerage or exchange business has margins anywhere close." Trainer contends that competitors will soon be offering similar service at much lower prices, in part because "Coinbase isn't protected by barriers to entry." He posits that Coinbase's future profits will settle at around $1.2 billion a year, well below its current run rate. He's describing just the kind of competitive, profit-crushing dynamic that prevents companies from minting huge margins for long periods, just what Wood's approach depends on. By the way, a drop that sharp in Coinbase's net would push the P/E for ARK's profitable members from 43 to a formidable 55, and further into the danger zone.

ARK’s big problem: a crowd of money losers, mostly medical

The other two-thirds of ARK's investments, measured by their market value on Jan. 3, sits in those 33 lossmakers. Their combined, prorated deficit over the past four quarters is $428 million. So overall, the fund's holdings are generating annual losses of $294 million ($133.6 in profits for the positives minus $428 million in losses for the negatives). Hence, ARK doesn't have a P/E multiple at all. It's generating 2 cents in losses for every one dollar its shares are selling for. A good chunk of those shortfalls flow from adventurous tech bets in such areas as robotics (UiPath), software (Unity), computer hardware (3D Systems), mobile communications (Trimble), and education and training (2U). Based on ARK's ownership percentage, Robinhood is the biggest profit drag, adding $58 million to the losers' overall deficit. Its stock has also been a drag, shedding 60% since its IPO in July.

The class deepest in the red: health care. ARK owns 18 companies valued at $5 billion in the sector, accounting for a third of its portfolio. Seventeen of these players, prized by ARK as spear carriers in a medical revolution, are losing money, measured by GAAP net earnings. Most of the group are striving to develop groundbreaking therapies in biotechnology (Fate Therapeutics, Veracyte, Pacific Bioscience) or genomics (2U, 10X, Crispr). The only profitable holding is Crispr, and it's second only to Coinbase in boosting ARK's plus column. The established titan in gene editing earned $403 million in the past four quarters, bringing ARK's 8.2% share to $33 million.

But the other 17 have dug a hole that Crispr's big contribution couldn't come close to filling. ARK's share of the combined losses in its medical portfolio was $200 million, amounting to around half of the negative group's total hit.

Maybe Wood expects that a few explosive winners in biotech or genomics will bail out ARK. "You expect big skews in returns from these types of companies," notes Ritter. "It's like movies. The vast majority are failures, but you get a couple of blockbusters that can more than pay for the bombs." It could happen at ARK. The rub is that most of Wood's health care investments have been around for a long time. Only five of the 18 had IPOs since the start of 2020. And still, only Crispr is making money.

Where is ARK going from here?

To achieve the 40% annual gains Wood predicts, ARK's stocks would need to produce a total return of 440% by early 2027. To get there, its portfolio companies must grow profits from underground to such gigantic levels that we can assume it won't happen, or even come close to happening. More likely, ARK's still so expensive, still so much the antithesis of the value sphere where the real deals reside, that it will underperform the Nasdaq and other indexes in the years ahead. Wood will probably be better remembered as a cultural phenomenon than a rebel who changed the game. "She's the new model for portfolio managers who talk directly to the people," says Trainer. "She's good at doing things that are popular. But she's also doing a huge disservice to investors."

Wood has caught one of the most powerful currents in investment history. Now the tide's reversing, the white caps are rising, and ARK's heading for a shipwreck.