CORZMarch 2, 2026 at 9:10 PM UTCSoftware & Services

Core Scientific's Q4 2025 Results Tout Expansion, But DeepValue Analysis Flags Unchanged Risks

Read source article

What happened

Core Scientific announced fourth quarter fiscal 2025 results, highlighting progress on existing builds and a scaling colocation pipeline to 1.5 gigawatts. However, the latest DeepValue master report underscores that the company's HPC hosting segment remains 100% dependent on CoreWeave, with significant execution and concentration risks. Despite the positive spin in the press release, SEC filings reveal that delivery dates for key projects like Denton, Texas have slipped due to weather and construction delays, and capex reimbursements are critical for liquidity. The report also notes that colocation revenue growth is inflated by power fee pass-throughs with no markup, limiting gross profit impact. Therefore, this announcement fails to address the core vulnerabilities of customer diversification and on-time delivery that drive the investment thesis.

Implication

The Q4 results emphasize capacity expansion, but the DeepValue report indicates that the business model is still tethered to CoreWeave, with 100% of colocation segment revenue from this single counterparty. High capex commitments of $1.14 billion, with $899 million reimbursable, create timing risks that could strain cash flow if reimbursements lag or delays persist. Without disclosed non-CoreWeave tenants, diversification efforts remain unproven, leaving the stock exposed to counterparty-specific shocks. Previous schedule slippages in Denton projects erode credibility, potentially delaying revenue recognition and increasing capital intensity. Thus, until concrete evidence of tenant diversification and stabilized delivery timelines emerges, the stock's upside is capped by these fundamental weaknesses.

Thesis delta

The announcement does not alter the core investment thesis; it reinforces that execution risks and customer concentration remain critical. While the 1.5 GW pipeline expansion is a positive step, it does not provide the necessary proof points of non-CoreWeave leases or on-time energized MW delivery. Thus, the WAIT rating and monitoring criteria from the DeepValue report remain unchanged, with no shift in the probability-weighted scenarios.

Confidence

moderate