Short sellers are increasingly betting against a cluster of AI data center and “neocloud” stocks, amassing roughly $6 billion in paper profits this October as concerns mount that their debt-funded expansion plans are becoming stretched. The bearish positioning is visible in the data: ORTEX, a market-data firm that tracks short selling, estimates short interest at roughly 20% to 26% of the free float—the shares available to trade—across CoreWeave, IREN, Nebius Group, Applied Digital, and Core Scientific.
This is more than a routine pullback. It reflects a growing skepticism about the business models of companies that are borrowing heavily to build out AI infrastructure before the revenue fully materializes. For everyday investors, understanding this dynamic is key to grasping why these stocks have been so volatile.
What the short interest numbers tell us
Short interest is the percentage of a company's tradable shares that have been sold short—investors betting the price will fall. When that figure climbs, it signals that a growing number of market participants expect declines. ORTEX's estimates show that across the five names, roughly $17 billion worth of stock is currently sold short.
CoreWeave, a cloud service provider focused on AI, has seen its short interest rise to 23.7% from 19.6% since the start of October. The stock has fallen almost 30% this month, handing short sellers around $1.7 billion in paper gains. Nebius, another AI infrastructure player, has short interest of about 20%, up from 15.4% earlier this year. Bears are up $2.94 billion in October on that position, though they remain down for the year overall.
The fact that borrowing shares is still cheap is notable. Low borrowing costs make it easier for skeptics to maintain their bearish trades, which can prolong downward pressure on share prices and keep volatility elevated. In other words, the short sellers aren't being squeezed out by expensive fees, so their bets can persist longer.
Why this matters for the companies and their funding
Heavy short interest doesn't just weigh on a stock quote. It can tighten the funding window for these companies. Data-center builders like CoreWeave and its peers often rely on raising outside capital to pay for servers, power infrastructure, and new sites before the revenue from those assets starts flowing. When shares are under pressure, an equity raise becomes more dilutive—meaning the company has to issue more shares to raise the same amount of cash. Lenders may also demand higher interest rates to compensate for perceived risk.
This is a bigger deal for these firms than for a typical software company. Their business model is capital-intensive, and they need continuous access to debt and equity markets to fund expansion. If short sellers keep pressing, refinancing or raising new money could become harder and more expensive. That could slow their growth plans or force them to accept less favorable terms.
The broader backdrop is also relevant. The AI boom has fueled a surge in demand for data centers, but it has also raised questions about whether the pace of building is sustainable. Companies across the supply chain are benefiting, but the financing side is getting more scrutiny. Utilities are seeing demand surge, and some suppliers remain optimistic. Yet the stocks themselves are vulnerable to shifts in sentiment.
What it means for investors
For ordinary investors, the takeaway is about risk. These stocks can offer exciting growth potential, but they come with significant volatility and financing risk. When short interest is high, price swings can be amplified, and negative news can trigger sharp declines. Conversely, if the companies deliver on their growth promises, short sellers could be forced to cover their positions, potentially driving prices higher in a short squeeze.
It's also worth noting that short interest is not a one-way bet. Some investors see these levels as a contrarian signal—if the bears are wrong, the rebound could be powerful. But the current environment suggests that the market is increasingly skeptical about the debt-fueled expansion model.
For those watching the sector, the key metrics to monitor are the companies' ability to secure funding at reasonable costs, their progress toward profitability, and any signs that demand for AI compute is slowing. Market volatility is already affecting new listings, and power grid constraints are a known risk for the industry.
In the end, the short sellers' bet is that these companies have overextended themselves. Whether they're right will depend on how the AI infrastructure buildout plays out. For now, the pressure is on.
