CoreWeave has had a brutal week. The AI infrastructure company, which builds and operates massive data centers designed to supply computing power for artificial intelligence workloads, saw its shares sink nearly 12 percent on Friday alone, and then slip another 1 percent in after-hours trading. By the end of the week, the stock had fallen more than 25 percent, marking the worst weekly decline since a roughly 26 percent drop back in November 2025. To put that in perspective, this is the kind of sell-off that shakes even patient investors, especially after months of eye-popping volatility. The concern is not really about whether CoreWeave has customers or contracts; it is about how the company is funding its enormous expansion. CoreWeave carries a heavy debt load, and as interest rates stay higher and borrowing costs rise, the market has started to worry that the company’s financial engine could become strained. Part of that anxiety shows up in the credit default swap market, where investors buy insurance against the risk that a company might fail to repay its debt. Those spreads have widened, and that has made investors jittery about every piece of bad news. The stock’s longer-term picture is even more uncomfortable: it is down about 14 percent year-to-date and roughly 57 percent over the past twelve months. That kind of slide stands in stark contrast to broad market indexes, which have been climbing. The Invesco QQQ Trust, for example, is up about 18 percent over the same twelve-month period, while the Vanguard Morningstar Total Stock Market ETF is up around 15 percent. So CoreWeave is not just falling; it is falling while the rest of the market is rising, which makes the decline feel even more severe for anyone watching the AI trade.
But Dan Ives, a well-known tech analyst, says the market is getting this story wrong. In an interview with Bloomberg on Friday, Ives pushed back hard against the panic. He described the sell-off as a “false narrative,” and argued that the recent pressure on CoreWeave, along with other AI infrastructure stocks, looks more like a buying opportunity than a reason to run for the exits. Ives drew a direct comparison to what he called the “SaaS apocalypse” earlier this year, a moment when investors suddenly turned against software stocks on fears that the whole category was overhyped and overpriced. That panic turned out to be overdone, and Ives believes the same pattern is playing out now with debt-heavy data center companies. He acknowledged that higher interest rates could create real challenges for CoreWeave, especially because the company needs continuous access to affordable capital to keep building out its network of GPU clusters and data centers. But in his view, the strength of AI demand and the size of the company’s backlog outweigh those concerns. He said there is a major disconnect between what the market is focused on, which is the debt and the cost of borrowing, and what is actually happening on the ground, which is that customers are still lining up for AI computing capacity at a remarkable pace. For Ives, the sharp decline in AI-related stocks is not a warning signal; it is an opportunity to buy at a moment when the short-term noise has overwhelmed the longer-term story.
To understand why Ives is so confident, you have to look at what he saw on a recent trip to South Korea. He said the trip reinforced his conviction that demand across the AI hardware ecosystem is accelerating, especially in memory chips. Memory is a critical piece of the AI supply chain, because every advanced AI server needs massive amounts of high-bandwidth memory to process large models quickly. If memory demand is surging, that is a strong signal that chipmakers, cloud providers, and data center builders are still scrambling to add capacity. Ives said that memory demand has accelerated by 30 to 40 percent over the last 60 to 90 days, and he believes that kind of growth is the clearest evidence that the AI boom is nowhere near finished. He also pointed to the upcoming earnings reports from hyperscalers, the giant cloud companies that are spending tens of billions of dollars each year on AI infrastructure, as the next big test. If those companies keep raising their capital expenditure guidance, that will confirm that the spending cycle is still accelerating. For Ives, that is the real barometer. He describes the current moment as being in the “third inning” of the AI game, meaning that despite all the hype and anxiety, the boom is still relatively early. In baseball, the third inning is not the final stretch; there is a lot of game left to play. Ives is essentially telling investors not to confuse a difficult stretch with the end of the story. He wants people to focus on demand, not on daily price swings, because demand is what will eventually determine which companies thrive over the long term.
Ives also addressed a legitimate risk that has been weighing on the entire sector: political and regulatory opposition to data center construction. As these facilities consume enormous amounts of electricity, water, and land, local communities and government officials have started to push back. Some projects have faced permitting delays, environmental challenges, and neighborhood resistance. Ives estimated that 5 to 10 percent of planned data centers could ultimately be rejected or delayed because of these political obstacles. That is not nothing, especially when billions of dollars are at stake. But he argued that this resistance will not fundamentally derail the AI infrastructure buildout. Even if some projects get canceled, the scale of what is still coming is enormous. Ives said the market should expect somewhere between 800 and 1,000 data centers to be built over the coming years, and his core thesis is that capital spending in this area will have a massive ripple effect across the entire technology sector. He described the multiplier as being “6 to 7x” relative to the rest of tech. In other words, every dollar spent on data centers will help drive several more dollars of activity through chips, networking, power equipment, cooling systems, software, and countless other categories. That kind of broad economic impact is why Ives remains so bullish, not just on CoreWeave but on the entire AI trade. He is not pretending there are no obstacles; he is simply saying that the size of the opportunity is so large that a few rejected projects will not change the overall trajectory.
Jefferies is also pushing back on the idea that CoreWeave’s problems are as bad as they look. According to a research note reported by TheFly, analyst Brent Thill maintained a Buy rating on the stock with a $150 price target, even as the shares were getting battered. Thill’s view is that the market’s reaction to rising yields on CoreWeave’s unsecured debt has been too harsh. The yield on a company’s debt is essentially the interest rate investors demand for lending to it, and when that yield rises, it signals that lenders are more nervous about the company’s financial health. But Jefferies believes the market is overreacting to that signal and ignoring some important offsets. Specifically, the firm pointed out that CoreWeave has been signing short-duration deals at around $40 million per megawatt, which are high-value contracts that should generate strong cash flow in the near term. The idea is that even if future borrowing costs go up, the revenue coming in from these deals is enough to offset the pressure. Jefferies therefore views the recent decline in CoreWeave’s share price as excessive. The company’s fundamentals, in their view, haven’t changed nearly as much as the stock price suggests. This is a classic situation where Wall Street analysts and the broader market are seeing the same facts but drawing very different conclusions. The analysts see a growing business with strong demand and revenue visibility. The market sees a company that relies heavily on debt, whose funding costs are rising, and whose stock is being hammered every time interest rate fears flare up.
Retail investors, meanwhile, seem to be siding with the bulls. On Stocktwits, a social platform popular with active traders, sentiment around CoreWeave was trending “extremely bullish” at the time of writing. That is striking given the stock’s awful performance and the widespread worries about the company’s debt. It suggests that many individual investors still believe in the AI story and are willing to look past the recent volatility. They may be thinking about the bigger picture: AI is still in a rapid growth phase, companies are still spending heavily on computing capacity, and CoreWeave is one of the key players in supplying that capacity. But this is also a stock that has burned long-term investors badly over the past year. Being down more than 50 percent over twelve months while the broader market is up is a painful experience, and it raises a legitimate question about whether the optimism is justified or just another example of falling in love with a narrative. The truth is that CoreWeave’s future depends on a combination of factors: AI demand, interest rates, competition, the availability of financing, and the company’s ability to execute on its contracts. Bulls like Ives and Jefferies believe the demand is strong enough to overcome the financial pressure, and they think the market is too focused on the debt side of the balance sheet. Bears would argue that debt-heavy infrastructure businesses are especially vulnerable in a higher-rate environment, and that the sell-off is a rational response to real risk. For now, the market is voting with its feet, and CoreWeave is having one of its worst weeks in nearly a year. But the debate is not settled. If upcoming earnings from hyperscalers show that AI spending continues to accelerate, the bulls may look wise for calling this a false narrative. If the spending slows or the financing picture worsens, the sell-off may prove to be just the beginning. Either way, it is a reminder that the AI boom is not just a technology story; it is also a financial story, and the two are not always in perfect harmony.