DDTL

SpecialTip2720

lmfao I knew wsb was regarded but does anyone actually read the reports or listen to the calls? All this info takes 5 minutes to verify. Cash flow was $2 billion, interest was $846 million. Their weighted-average cost of debt across their entire debt stack is 7.9% (down 250bps YoY), newest DDTL is 7-7.5%. CoreWeave’s high-yield corporate bonds are **unsecured** \- bond holders have no lien to reposes data centers. DDTLs get paid automatically. Customers pay their monthly compute bill, which goes straight into a locked SPV waterfall account that pays the lender's interest and principal first before CoreWeave even touches it.

SpecialTip2720

Personally not got a penny invested in Coreweave as their aggressive debt leveraged approach turns me off, but some of the talk in this thread is fundamentally misunderstanding their financials & business strategy. **Accounting**: You do not pay interest out of GAAP Operating Income. It is calculated after already deducting $1.89 billion in quarterly non-cash Depreciation & Amortisation, they pay out of operating cash flow. "thats dwarfed by the interest payment alone, not to mention the capex" - double counting GPU expenditure. **Gpu depreciation**: If GPUs became worthless waste at the end of their depreciation schedule, customers wouldn't renew them. They just announced on the call they're re-contracting expiring older clusters at a meaningful premium, expecting $10 billion in annualised incremental revenue by 28, with $3bil in prepayments by customers to lock contracts in. **Debt**: Coreweave uses a success based capex model. They only buy GPU's when a customer signs a legally binding contract then draws a DDTL against it. The customers payments are locked into a process that pays the interest and amortises 100% of the loan principal before contract ends. **Risk:** If their customers are unable to pay or go under themselves then they're in trouble. This could be said for any business, but I think the sentiment and what people are really picking up on with bearish takes on here is a fundamental mistrust in the whole AI narrative. That I dont have an answer for, and I'm unsure of myself. The valid thing to point out here is how aggressively exposed Coreweave are to a shift/break in the AI expansion as it stands today. They have leveraged debt aggressively and bet the farm on the AI ecosystem and as a result - their customers liquidity. In my mind this is the valid critique, Coreweave is in a position where they have no fundamental flexibility to weather a severe downturn in the market and this is where the real risk lies. As it stands, their financials are fine and they do make sense - but only in the case where the AI trade continues to be robust. Their financials aren't fundamentally broken, just brittle. It's up to investors to decide what chance there is of slowdown in the AI trade and whether the potential upside is worth the risk.

Ishowspood

That is unsecured debt. In the earnings call they said they won’t tap into it. The new DDTL Loan upcoming for them is 7-7.5% far less than 15-17%

chillinathid

Coreweaves cost of financing has actually been going down over time. Their investment grade DDTL bonds keep having better rates.

Big-Pomegranate3243

**Executive summary** * Q3 revenue reached $3.661B, up more than $1B sequentially; adjusted operating margin reached 10%, a quarter ahead of plan. * Revenue backlog was $132.5B as of September 30, including remaining performance obligations and other estimated revenue under committed contracts, subject to delivery and service availability. **Financial highlights** * Q3 revenue was $3.661B vs. $1.365B YoY; operating income was $207M vs. $52M, with GAAP operating margin of 6% vs. 4%. * Net loss was $709M vs. $110M; diluted net loss per share was $1.27 vs. $0.22, and net loss margin was 19% vs. 8%. * Adjusted EBITDA was $2.256B vs. $838M, with adjusted EBITDA margin of 62% vs. 61%; adjusted operating income was $379M vs. $217M. * Adjusted net loss was $529M vs. $41M; adjusted net loss margin was 14% vs. 3%. * Q3 operating cash flow was $1.975B; purchases of property and equipment, including capitalized internal-use software, were $9.441B. **Outlook and guidance** * FY2026 revenue guidance increased to $13.0B–$13.3B; adjusted operating income guidance rose to $1.275B–$1.375B, a 26% midpoint increase vs. prior guidance. * Year-end active power is expected to exceed 1.85 GW; FY2026 capex is now expected at $37B–$39B, reflecting faster infrastructure deployment and customer delivery. * Exit-2026 annualized run-rate revenue guidance increased to $20B–$21B; the measure is last-month quarterly revenue multiplied by 12. * Q4 2026 interest expense is expected to be $1.05B–$1.15B. **Segment performance** * Active power topped 1.7 GW after adding more than 200 MW in Q3; contracted power grew by more than 900 MW to over 4.6 GW. * Contracted power across India, Indonesia, Spain and other international markets exceeded 1.3 GW; the provider portfolio and powered-land footprint also expanded. * Customer wins included Canva, Mirendil, Parallel Web Systems, Serve Robotics, Substack and Wayve; relationships expanded with OpenAI, Cohere, Perplexity and others. * Multi-rack NVIDIA Vera Rubin NVL72 deployments came online; Cognition’s first production workloads achieved 4.8x greater token throughput than GB200 NVL72. * CoreWeave Forge launched for AI development workflows; performance milestones included a third consecutive SemiAnalysis Platinum ClusterMAX rating and top cloud-provider server throughput in MLPerf Inference v6.1 using NVIDIA GB300 NVL72. **Key financial ratios and metrics** * GAAP operating margin was 6% vs. 4% YoY; net loss margin was 19% vs. 8%. * Adjusted EBITDA margin was 62% vs. 61%; adjusted operating income margin was 10% vs. 16%. * Nine-month revenue was $8.314B vs. $3.559B; nine-month net loss was $2.075B vs. $715M. * Cash, cash equivalents and restricted cash totaled $6.404B at September 30, 2026, vs. $4.130B at December 31, 2025. **Risk factors and uncertainties** * Execution, customer retention, continued AI infrastructure demand, power access, storage capacity, supplier and data-center partner continuity, and internal controls could affect results. * Substantial indebtedness, capital needs and access to financing, alongside geopolitical conditions, inflation, tariffs, trade controls, interest rates, banking-sector instability and FX, pose risks. * Revenue backlog and expectations for customer prepayments, deposits, contracted power and future guidance depend on estimates and delivery or service-availability conditions. **Significant events and developments** * Launched Physical AI Field Engineering and established the CoreWeave Partner Network, with production-validated integrations including Grafana Labs, CrowdStrike and VAST Data. * Partnered with Leidos on secure AI cloud capabilities for defense and intelligence; selected with Parallel Works to deliver a managed AI cloud for DARPA biological research. * Signed a multi-year agreement with Solidigm for priority access to enterprise SSD capacity; announced support for NVIDIA Vera CPUs for AI-agent workloads. **Capital allocation and financing** * Raised approximately $4.2B through an upsized convertible senior notes offering, established an ATM program for up to 35M Class A shares and closed a $2.6B DDTL 5.5 Facility. * The DDTL 5.5 Facility is the first financing facility designed to support shorter-duration customer contracts; management cited operating cash flow, customer-contract-backed financing and expected prepayments as funding sources. * Approximately $9B in customer prepayments and deposits was expected over the next three months; FY2026 capex guidance is $37B–$39B. * Balance-sheet recourse debt was $9.183B current and $32.417B non-current at September 30; non-recourse debt was $1.389B current and $3.884B non-current. **Notes to the financial statements** * Adjusted results exclude items including stock-based compensation and acquisition-related costs; adjusted net loss also reflects debt-extinguishment costs, acquired-intangible amortization, fair-value adjustments and other items. * Q3 GAAP results included a $27M gain on fair-value adjustments; adjusted EBITDA reconciliation removes the gain, while adjusted net loss reconciliation also includes a $14M debt-extinguishment adjustment. **Market and industry conditions** * Demand from AI labs, hyperscalers and enterprises supported customer additions, expanded relationships and rapid growth in active and contracted power capacity. * Infrastructure availability remains central to growth: power, powered land and enterprise SSD access are being expanded, while infrastructure delivery is expected sooner than planned.