The Call: Buy From Hold, but $5 Comes Off the Target
Ramnani based the upgrade on enterprise adoption of open AI models, sovereign AI demand and CoreWeave’s scale advantage over other neocloud providers. Truist estimates that CoreWeave could reach 1.7 gigawatts of active power and about $18 billion of annual recurring revenue by year-end.
The valuation gap was central to the call. Truist estimated CoreWeave at about 2.8 times expected 2027 sales, compared with 4.8 times for Nebius and 5.4 times for IREN, despite CoreWeave’s larger operating footprint.
The lower target keeps the risks inside the note. Truist expects CoreWeave to spend $35.1 billion on capital projects in 2026 and $60.1 billion in 2027, with deeply negative free cash flow while the company builds capacity.
Why a Lower Target Can Still Produce an Upgrade
Analyst ratings reflect the expected return from the current share price, not only the absolute value of the target.
A move from $131 to $126 says Truist’s estimate of CoreWeave’s future equity value weakened slightly. Moving the rating to Buy says the market price weakened much more.
That distinction matters because the debate surrounding CoreWeave is no longer whether demand for AI computing exists. It is whether the company can finance capacity, activate it on schedule and turn contracted demand into cash before interest and depreciation absorb too much of the benefit.
FinanceFeeds previously framed the potential outcomes in its CoreWeave $303 bull case and $36 bear case. Truist’s call sits between those extremes: multiple compression has created upside, but the balance-sheet argument remains unresolved.
CoreWeave Has the Contracts — and the Capital Bill
CoreWeave ended March with $99.4 billion of revenue backlog, including a new $21 billion Meta commitment and a multi-year agreement with Anthropic. It had surpassed 1 gigawatt of active power and secured more than 3.5 gigawatts of contracted power.
The cost of delivering that contract book is substantial. CoreWeave paid $7.7 billion for property and equipment in the first quarter alone, up from $1.4 billion a year earlier.
The company reported $25.1 billion of debt principal at March 31, including $7.5 billion classified as current debt. It had $11.1 billion of total liquidity, but that figure included $8.8 billion available through existing borrowing facilities alongside its cash and securities.
That is why the upgrade cannot be treated as a verdict that CoreWeave’s fundamentals are uncomplicated. Its backlog provides unusual revenue visibility, but the company must fund GPUs, data centers and power before much of that contracted revenue can be recognized.
AMD and Anthropic Confirm the Demand Signal
The sector demand case strengthened this week when AMD agreed to supply Anthropic with up to 2 gigawatts of Instinct MI450 systems beginning in the first half of 2027. AMD also agreed to invest up to $5 billion in Anthropic, with the investment tied to deployment milestones.
The agreement does not automatically become CoreWeave revenue. Anthropic plans to deploy some systems directly and obtain additional capacity through cloud providers and other AI infrastructure companies. It does, however, confirm the scale of the computing requirement Truist is underwriting: AI developers are still committing to infrastructure measured in gigawatts and tens of billions of dollars.
ARK Invest made a similar allocation choice before the AMD announcement. Its July 17 trades included approximately $19.4 million of SpaceX and $8.4 million of CoreWeave purchases, bringing the combined buying to nearly $28 million, while it sold about $11.8 million of AMD shares.
ARK states that its daily trade files are published after execution and are informational, unofficial and unreconciled rather than complete fund accounting records.
The Levels and the Next Catalyst
CoreWeave traded between $80.67 and $85.30 on Thursday and was at $81.83 at 11:20 a.m. ET. The low-$80s are therefore the immediate reference area, while Truist’s $126 target represents the 12-month valuation case rather than a near-term technical level.
As of Thursday, CoreWeave had not added a second-quarter earnings call to its investor-relations calendar. The results will be the next major test once scheduled.
Investors will be watching whether backlog continues to rise, how quickly contracted power becomes active capacity and whether revenue growth begins to outrun depreciation, interest expense and capital spending.
The upgrade says CoreWeave became cheap enough to buy. The reduced target says it did not become safer.