A more open ecosystem could actually be more profitable.
In the long arc of technological transformation, companies that survive their own hype cycles often emerge as something more durable than what the market once celebrated. Roku, once a hardware curiosity riding the streaming wave to euphoric valuations, has quietly rebuilt itself into a capital-efficient monetization platform — one that investor Cathie Wood, undeterred by a 73 percent decline from peak, continues to hold as a core conviction. The story unfolding in Echo Harbor's broader economic landscape is not one of recovery to former glory, but of a business discovering that openness, recurring revenue, and disciplined capital allocation can be more valuable than scale alone.
- Roku's stock sits 73% below its 2021 peak, yet Cathie Wood has returned to buy it 14 times, signaling a conviction that the market's verdict on the company is fundamentally mistaken.
- The tension between a once-closed advertising ecosystem and the demand for open programmatic access has been resolved — and counterintuitively, opening the platform to rivals like Amazon's DSP pushed gross margins higher, not lower.
- A hidden subscription business, long buried in the financials, has now been disclosed as a fast-growing annuity stream with 40% gross margins, reshaping how investors must value the company.
- With capital expenditures in the single-digit millions and $1.2 billion in tax shields, Roku's path to $1 billion in annual free cash flow by 2028 is less a projection than a structural inevitability in the making.
- The first sequential decline in shares outstanding in company history signals that Roku has crossed from growth-at-any-cost into disciplined value creation — the kind of inflection ARK Invest was built to find early.
Cathie Wood has built her reputation on finding value in wreckage, and Roku represents one of her most deliberate bets. ARK Invest first purchased shares in late 2019 and has traded the position 28 times since, arriving at an average cost of roughly $131 — almost exactly where the stock trades today, despite sitting 73% below its 2021 peak. Wood now holds 3.82 million shares worth approximately $497 million. The question is what she sees that others do not.
Three years ago, Roku was organized around scale rather than profit, building an audience of 80 million streaming households while forcing all programmatic advertisers through its own closed system. The transformation began when CFO Dan Jedda arrived in 2023 and opened the advertising platform to outside demand-side platforms — including a high-profile partnership with Amazon that blends Roku's viewer data with Amazon's shopping intelligence. Rather than cannibalizing profitability, the open ecosystem accelerated it. Advertising gross margins climbed to just over 60%, a jump of 450 basis points year over year.
But advertising is now only half the story. For years, Roku's subscription revenue was buried inside broader financials, invisible to most investors. When the company finally separated the numbers, it revealed a fast-growing recurring revenue stream with 40% gross margins — steady, predictable, and operating like an annuity. Roku owns two subscription products: Frndly, an acquisition, and Howdy, built internally as a low-cost ad-free service that has already expanded onto Amazon's platform and into Mexico, where Roku now operates nearly as many streaming households as in the United States.
What makes this particularly compelling to Wood is the capital efficiency underneath it all. Capital expenditures run in the single-digit millions, meaning nearly every dollar of earnings converts directly to free cash flow. A $1.2 billion tax shield from accumulated losses will protect the company from cash taxes for years as profitability accelerates — making free cash flow exceed EBITDA in the near term, a rare dynamic. Jedda has committed publicly to $1 billion in annual free cash flow by 2028, with analysts projecting further expansion to $1.78 billion by 2030. At current multiples, that trajectory could double the stock within four years.
Roku has already begun buying back shares, and the first quarter of 2026 marked the first time in company history that shares outstanding declined sequentially. Whether the stock ever revisits its 2021 peak near $480 remains uncertain. But the business being built today is more durable, more capital-efficient, and more structurally sound than the one that once commanded that valuation — and that, for Wood, appears to be precisely the point.
Cathie Wood has built a reputation on seeing value where others see only wreckage. Her firm, ARK Invest, first bought Roku in late 2019 and has returned to the stock 14 times since, selling on 14 occasions as well. The average price paid across those trades was roughly $131 per share—almost exactly where Roku trades today. Yet the stock sits 73 percent below the peak it reached in 2021, when optimism about streaming seemed boundless. Wood now holds 3.82 million shares worth approximately $497 million, making Roku her sixth-largest position. The question that naturally follows is simple: what does she see?
Three years ago, Roku was a hardware company, period. It had built an audience—around 80 million streaming households when Dan Jedda arrived as chief financial officer in May 2023—but the business was organized around scale rather than profit. The real transformation came when Roku opened its advertising platform to outside demand-side platforms, the automated systems that major brands and agencies use to buy digital ads. Before this shift, Roku forced all programmatic advertisers through its own system, a choice that limited the market and hurt the bottom line. Now the company works with every major DSP, including a high-profile partnership with Amazon that lets advertisers blend Roku's viewer data with Amazon's shopping information in a privacy-compliant way. Jedda called the Amazon deal "a major unlock," and the financial results validate that language. Advertising gross margins have climbed to just over 60 percent, a jump of 450 basis points year over year.
That margin improvement matters because it directly addresses the fear that haunted investors: that opening the platform to competitors would cannibalize Roku's own profitability. It did not. Instead, the company discovered that a more open ecosystem could actually be more profitable. But advertising is only part of the story now. For years, Roku kept its subscription business buried in the financials, lumped together with advertising revenue. When the company finally separated the numbers, investors discovered something they had missed: a fast-growing, recurring revenue stream with gross margins just above 40 percent. Jedda acknowledged the disclosure was overdue, noting that investors simply did not understand how significant subscriptions had become. The segment operates like an annuity, steady and predictable. Roku owns two subscription products—Frndly, which it acquired, and Howdy, which it built from scratch as a low-cost, ad-free streaming service. Howdy has already expanded to Amazon's platform and into Mexico, where Roku now operates nearly as many streaming households as it does in the United States.
What makes this business model particularly attractive to someone like Wood is the capital efficiency. Roku spends almost nothing on capital expenditures—Jedda said CapEx runs in the "single-digit millions"—which means nearly every dollar of earnings flows directly into free cash flow. The company also carries roughly $1.2 billion in net operating losses on its balance sheet, a tax shield that will protect it from cash taxes for the next couple of years as profitability accelerates. This means free cash flow will actually exceed EBITDA in the near term, a rare and valuable dynamic. Jedda has publicly committed to reaching $1 billion in annual free cash flow by 2028, though he suggested it could arrive sooner. Analysts tracking the company forecast free cash flow expanding from $478 million in 2025 to $1.78 billion by 2030. If the stock trades at 22 times forward free cash flow—roughly its current multiple—it could double within four years.
Roku is already deploying excess cash to buy back shares, and the first quarter of 2026 marked the first time in the company's history that shares outstanding declined sequentially. For Wood, this is the kind of inflection point that ARK tends to identify early and hold through the inevitable noise and skepticism. Whether Roku ever returns to its 2021 peak near $480 remains an open question. But the business the company is building today bears little resemblance to the one that commanded that valuation, and it appears far more durable.
Notable Quotes
The Amazon partnership was described as a major unlock for the company's monetization strategy.— Dan Jedda, Roku CFO
Roku acknowledged that investors did not understand how significant the subscription business had become until it was separately disclosed.— Dan Jedda, Roku CFO
The Hearth Conversation Another angle on the story
Why does Wood keep buying a stock that's down 73 percent? Isn't that just throwing good money after bad?
Because the stock fell for reasons that had nothing to do with the business getting worse. Roku was built as a hardware play, and when the streaming boom cooled, the market punished it. But meanwhile, the actual business was transforming into something much more profitable.
What changed?
The company opened its advertising platform to competitors. That sounds like it should hurt margins, but it actually expanded the market so much that margins went up 450 basis points. And they discovered they had a subscription business worth billions that nobody was paying attention to.
So Wood is betting on profitability, not growth?
Both, but profitability is the unlock. The company spends almost nothing on capital, carries a $1.2 billion tax shield, and is on track to generate a billion dollars in free cash flow by 2028. At current multiples, that could double the stock.
And if it doesn't work?
Then Wood loses money on a $497 million bet. But she's made this kind of call before—finding the moment when a broken stock becomes a cash machine. The question is whether the market believes it yet.
Does it matter if Roku ever gets back to $480?
Not really. The business that earned $480 was fragile. The one being built now is built to last.