CoreWeave stock drops amid technical sell signal, strong revenue
CoreWeave shares fell to $82, more than 40% below the year-to-date high, forming a head-and-shoulders pattern with a $64 neckline while revenue rose 112% to $2.6 billion.
CoreWeave shares declined to $82, a drop of more than 40% from the year-to-date high and the lowest level since Aug. 3. The stock chart displays a head-and-shoulders formation with a $64 neckline.
The head of the pattern is at $137.75, reached on May 6, with the left and right shoulders near $115 and $117. The share price has moved below its 50-day exponential moving average, a common technical indicator that traders use to gauge momentum.
A sustained break below the $64 neckline would point to additional downside, with an initial technical target near $50. The pattern would be invalidated if the stock climbs back above roughly $117.30.
CoreWeave reported revenue of $2.6 billion, a 112% increase year over year. Adjusted EBITDA was $1.5 billion. The company said its backlog exceeds $104.2 billion and listed customers that include OpenAI, Anthropic, Microsoft, Meta Platforms and Mistral, and it provides services to high-frequency trading firm Jane Street.
Balance-sheet items and costs present offsetting factors. Total debt rose to nearly $30 billion at the end of the last quarter, and operating lease liabilities exceed $15.7 billion. Hardware and storage costs have increased, contributing to higher capital and operating expenses.
Competition in GPU-focused data-center capacity has expanded. Large agreements and financing arrangements involving cloud and AI firms have increased options for customers; reported deals include multi-billion-dollar arrangements involving Anthropic, Lambda and Nscale, and a major financing package tied to Nvidia for OpenAI. Other companies expanding capacity include RIOT Platforms, Mara Holdings, IREN and Nebius.
Short interest in CoreWeave is about 15%, reflecting elevated bearish positions. CoreWeave operates GPU-optimized data centers that support machine-learning training and inference for large AI models.








