DVB
Deep-Value Investor, Ex-Tech Executive, Agentic Equity Research Builder
New Era Energy & Digital's data-center plan looks risky because building with borrowed money from nothing means high interest costs and squeezed margins. As a small player with no experience, it faces liquidation risk if capacity is underutilized after a market change. Even if NUAI puts up capital for its joint venture with Stream, its revenue share looks small, raising the question of what NUAI actually contributes. x.com x.com x.comÂ
New Era Energy & Digital is a high-risk speculative play rather than an asymmetric bet, because it is essentially a shell company with MOUs and lacks the capital to build data centers. About 100MW needs $200-300M, yet it holds only about $2M of cash and cannot borrow like established players, and relationships with ERCOT are not enough. Without that capital it is valued like a natural gas consultant, not a data center operator. x.com x.com x.comÂ
New Era Energy & Digital's shifting data-center story is not credible. @AtlasShrug1 notes the partner keeps changing, from PDI to Stream and a hyperscaler to talk of Anthropic via Theseus, with no LOI. The Sharon AI JV, what NUAI paid and what it got, says everything about the company. The PPA gives Vistra 5% equity for the right to buy power, and new CEO Nelson has been involved with Gray and New Era Helium since 2024, so he is not exactly new. x.com x.com x.comÂ
New Era Energy & Digital's new colocation deal does not change the bearish view: its economics are good, not great, and revenue is a year out. With 207MW of gross power, only about 150MW of IT load and $1.5B of capex, the deal implies roughly $75M a year of profit on a 10-year payback, beginning in the second half of 2027 at the earliest. x.com x.comÂ
New Era Energy & Digital will have leverage because it has locked in power that data centers cannot be built without. It secured up to 207 MW for 20 years, expected in Q3 2027, when power will be extremely constrained. With the power agreement signed, the next step is securing a tenant and seeing how far the company can take it. x.com x.comÂ
Nebius has pricing power because its hard-to-build AI infrastructure and software serve enterprise clients who cannot find capacity elsewhere. It owns some data centers and land but has mostly operating leases; the share of active compute at owned sites is unspecified. It looks like the best infrastructure play despite slight skepticism about its path to profitability and valuation, and would work even if capex slowed 20% or the cycle paused. x.com x.com x.comÂ
Heavy AI-related borrowing with little to show for it is pressuring Oracle, IREN and Nebius shares. These stocks are beginning to trend lower. x.comÂ
NUAI could become very valuable by delivering the power AI data centers increasingly struggle to secure. Its 493-acre TCDC campus near Odessa, Texas plans 207 MW, then 550 MW, and up to 1.4 GW long term. The first two phases use behind-the-meter generation without ERCOT interconnection, in a Permian Basin site with gas, fiber and water. Permits are advancing, and power delivery could start by end-2027. It still needs a credible anchor tenant to prove demand, making it a small, speculative bet. … Show more
Oracle looks buyable for now because the large financing and equity raises feared for next year no longer look imminent. New Mexico commissioners say there were no talks to alter Project Jupiter's power agreement and doubt a majority will approve a change; Stargate may be over a year behind. PIMCO has left AI data-center bonds, which still threatens funding. This is probably a bottom until the AI bubble pops; wait until $200 to short. x.com x.com x.com x.com x.com x.com x.com x.comÂ
CoreWeave, Nebius, Oracle and IREN cannot count on Nvidia to absorb their debt risk because its AI funding creates no obligation to rescue them. If the AI trade unwinds and they become desperate for cash, their assets could become acquisition targets at a fraction of peak valuations; Nvidia might acquire them cheaply instead of bailing them out. x.com x.comÂ
New Era's land value should be discounted for regulatory risk until its power is truly secured, though the base case still values that land at about 6x Northland's $20M. Signed power isn't secured power: even with a Vistra contract, behind-the-meter colocation must clear ERCOT and PUCT review under SB6, guided at about 180 days but likely longer. The JV will likely be signed before approval, so a counterparty will price in that risk, and a drawn-out process could cause costly delays. x.com x.comÂ
AI cloud infrastructure stocks risk falling together because neocloud providers depend on the same debt funding. Musk says debt markets have almost shut off financing for companies dependent on AI data-center revenue, though he is a chip buyer and balance sheets differ; Nebius remains up over the year. x.comÂ
