CoreWeave closed a $3.1 billion delayed draw term loan facility on Monday to fund infrastructure for two large, non-investment grade customers, which follows an $8.5 billion investment-grade DDTL 4.0 facility the company completed earlier this year. CoreWeave Financing DDTL V, LLC issued the DDTL 5.0 facility, while Morgan Stanley and Mitsubishi UFJ Financial Group served as joint lead arrangers and bookrunners. The transaction received ratings of Ba2 from Moody’s and BB+ from Fitch. The pricing for this facility tightened by 50 basis points from initial discussions to arrive at SOFR + 4.50%. CoreWeave also announced a $1.5 billion senior notes offering due in 2031 to support ongoing growth and lean into long-term demand for AI compute capacity. This issuance follows a $2 billion note offering in May. Qualified institutional buyers and non-U.S. persons under Regulation S will receive the 2031 bonds. Moody’s assigned a B1 rating to these new notes, while Fitch rated them at BB-. This massive debt load results in a debt-to-equity ratio of 7.39 against a total debt of $35.1 billion.
Backlog and concentration risk
The company manages a revenue backlog exceeding $104 billion, with 40% of that amount realizable within 24 months. This backlog helps cover debt obligations, but revenue concentration remains a risk because Fitch reported that 65% of Q1 revenue came from only two customers. You should monitor if these major clients reduce their spending. While the backlog remains large, analysts expect gross leverage to reach 6.3x by the end of 2026 as debt climbs to $47.6 billion. The company currently operates 51 data centers providing 1.5 gigawatts of compute. This includes roughly 4.2 GW under contract as of August 11. Rittenhouse Research claims most of the debt load ties to underlying customer contracts rather than speculative CapEx. Specifically, 98% of revenue comes from specific long-term customer contracts lasting four or more years. In Q1 2026, the company reported revenue of $2.08 billion, beating the $1.97 billion forecast, though it missed earnings per share expectations with a result of -$1.40 against the forecasted -$0.91. DA Davidson reiterated a Buy rating and maintained a price target of $175. CoreWeave’s 2032 senior notes fell below $80, reflecting investor unease as the era of easy money fades, according to Wells Fargo’s Luis Alvarado.
| Financing Detail | Value/Spec |
|---|---|
| DDTL 5.0 Facility | $3.1 billion |
| DDTL 5.0 Maturity | ~5.5 years |
| DDTL 5.0 Pricing | SOFR + 4.50% |
| Senior Notes (2031) | $1.5 billion |
| Total Debt | $35.1 billion |
| Debt-to-Equity Ratio | 7.39 |
| Revenue Backlog | >$104 billion |
| Q1 Revenue | $2.08 billion |
Market reaction and expansion
CoreWeave shares rose 4.5% on Monday after the $1.5 billion notes announcement. This follows a nearly 12-15% drop on October 14 when investors feared the heavy debt load. The company lowered its implied average interest rate from 13% to 11% in the second quarter because total debt grew from $8.8 billion to $11.2 billion during that same period. The company also proposed a $2 billion convertible notes offering that gives initial purchasers an option to buy up to an additional $300 million. CoreWeave intends to use a portion of those proceeds to fund the cost of entering into capped call transactions to reduce potential dilution to the CoreWeave Class A common stock.
The company recently secured $4 billion through a joint venture to support expansion in Lancaster, Pennsylvania. This deal involves funds from Blue Owl Capital, affiliates of Chirisa Technology Parks, and Machine Investment Group. This funding targets the Lancaster campus, which will host compute infrastructure for AI applications. The partnership also plans to fund $200 million for local power grid improvements with PPL Electric Utilities. The Lancaster facility will utilize adaptive reuse of existing buildings and advanced, energy-efficient cooling systems. CoreWeave’s infrastructure provides computing up to 35 times faster and 80% cheaper than AWS or Google Cloud. The company also topped inference speed benchmarks for Moonshot AI’s Kimi K2.6 model, delivering 205 tokens per second at $0.7 per million tokens. Barclays analyst Raimo Lenschow set a price target of $140, noting a valuation of 50x CY26E EV/EBIT. Does the massive debt accumulation eventually outpace the revenue from these specialized AI workloads?
