By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel is neutral on APLD, with concerns around execution risks, customer concentration, and the durability of AI demand outweighing the transformative $7.5B revenue contract and 15-year lease duration.
Risk: Customer concentration and the durability of AI demand
Opportunity: The $7.5B revenue contract and 15-year lease duration
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Applied Digital Corp (NASDAQ:APLD) is one of the best data center stocks to invest in according to billionaires. Applied Digital shares have gained around 57% year-to-date and have surged more than 245% over the past 12 months. Analysts continue to see substantial upside potential in the stock.
Applied Digital Corp (NASDAQ:APLD) has secured another major data center deal with an existing hyperscale customer. On May 20, the company said it signed a 15-year lease for its fourth AI-focused data center campus, known as Polaris Forge 3, with this customer. This is the same customer that the company previously signed for the Delta Forge 1 campus.
Applied Digital said the latest contract is expected to generate around $7.5 billion in revenue over the initial lease term. If all renewal options are exercised, the total value could reach $18.2 billion. According to Applied Digital, the latest deal pushes its contracted revenue backlog to roughly $31 billion across four campuses. The backlog could increase to $73 billion if customers extend their leases.
About 65% of Applied Digital’s contracted revenue is now backed by blue-chip customers. These are US-based investment-grade hyperscalers. The Polaris Forge 3 project will provide 300 megawatts of computing capacity and is scheduled to begin operations in 2027. So far, Applied Digital’s total contracted capacity is 1.2 gigawatts.
Applied Digital Corp (NASDAQ:APLD) builds and operates data centers and provides colocation services. The company provides the digital infrastructure and services to power AI, cloud, and blockchain workloads. It has purpose-built infrastructure for training and running AI models.
While we acknowledge the potential of APLD as an investment, we believe certain AI stocks offer greater upside potential and carry less downside risk. If you're looking for an extremely undervalued AI stock that also stands to benefit significantly from Trump-era tariffs and the onshoring trend, see our free report on the best short-term AI stock.
READ NEXT: 10 Best Debt-Free IT Stocks to Buy Now and 10 Best Stocks to Buy According to Billionaire Bill Gates.
Disclosure: None. Follow Insider Monkey on Google News.
Four leading AI models discuss this article
"Backlog and long leases look promising, but the value rests on massive capex and financing with few customers—any delay, cost overrun, or rate shock could drastically erode cash flows."
APLD touts a 15-year Polaris Forge 3 lease for 300 MW with a $7.5B initial revenue and a $31B backlog, lifting contracted capacity to 1.2 GW and blue-chip exposure to 65%. That reads as durable demand from hyperscalers and meaningful optionality. But the economics are fragile: 300 MW of new build implies multi‑billion capex likely financed with debt, so rising rates or refinancing risk could erode margins. The backlog is value, not cash, and renewals depend on customer budgets and competitive dynamics. Execution risks include interconnection delays, construction overruns, and competition from other data-center players. AI demand is still uncertain, and the stock may already reflect optimistic pricing.
The strongest counterpoint is that the backlog is contingent on financing, permitting, and renewals; any rate shock or a major customer's delay could drastically reduce the expected IRR.
"APLD has achieved a critical pivot to hyperscale infrastructure, but the stock's current valuation fully prices in perfect execution of its multi-year, capital-intensive build-out."
APLD’s $7.5 billion revenue contract is undeniably transformative, signaling that the company has successfully transitioned from a speculative crypto-miner to a legitimate infrastructure play for hyperscalers. The 15-year lease duration provides rare cash flow visibility in a volatile sector. However, investors must look past the headline backlog. APLD’s execution risk is massive; scaling to 1.2 gigawatts requires immense capital expenditure and power grid interconnectivity that often faces multi-year delays. While the 245% surge reflects market enthusiasm, the valuation now hinges entirely on the company’s ability to deliver these massive facilities on time without significant cost overruns or power procurement failures.
The company’s massive backlog is contingent on future project completion, meaning that if power grid bottlenecks or construction delays occur, these multi-billion dollar contracts could be renegotiated or terminated, leaving APLD with stranded assets.
"APLD's $31B backlog is optionality, not earnings—the stock's 245% YTD move has already priced in near-perfect execution through 2027, leaving minimal margin for error."
The $31B contracted backlog is real and material, but the article conflates backlog with revenue certainty. A 15-year lease sounds locked, yet hyperscalers have renegotiated or walked away from infrastructure deals before—see the AWS/Nvidia tension over custom chips. The 2027 start date for Polaris Forge 3 means zero revenue contribution for 2+ years. More critically: APLD trades on backlog multiples, not earnings. At current valuation, the stock prices in flawless execution, 300MW delivery on schedule, and zero customer concentration risk. The 65% blue-chip concentration is reassuring but also means one major customer loss cascades. The article omits capex requirements, debt levels, and power availability constraints—all material to whether this backlog converts to actual profit.
If execution falters (power grid delays, supply chain, customer pivot to in-house builds), the backlog becomes a liability, not an asset. Hyperscalers are increasingly building proprietary infrastructure; APLD's moat is narrowing.
"Execution and funding risks on 2027-era campuses outweigh the headline backlog after a 245% run-up."
APLD’s 15-year Polaris Forge 3 lease lifts contracted backlog to $31B, yet the 300 MW campus only starts in 2027 and requires substantial capex the company has not yet funded. With 65% of revenue tied to a handful of hyperscalers and total capacity still just 1.2 GW, any delay, rate renegotiation, or slowdown in AI training demand could erase much of the announced value. Shares already up 245% in twelve months price in flawless execution that history of data-center developers rarely delivers.
The same customer returning for a second multi-billion-dollar campus signals durable, investment-grade demand that could de-risk future financing and support further multiple expansion.
"Backlog durability depends on customer diversification and financing commitments; high hyperscaler concentration could prune IRR if that single customer delays or renegotiates."
Grok’s view that a second mega-campus signals durable demand and de-risks financing overlooks concentration risk. Relying on 1–2 hyperscalers makes the backlog highly rate-sensitive and vulnerable to renegotiation or delays. If that customer pulls back, the entire 1.2 GW backlog and related IRR could compress, regardless of schedule fidelity on Polaris Forge 3. Diversification of customers and verifiable financing commitments would be needed to translate the backlog into durable, repeatable cash flow.
"APLD's real moat is the acquisition of power-ready sites that bypass the standard grid interconnection delays plaguing competitors."
Claude and Grok ignore the regulatory tailwind: the FERC-approved interconnection queue reforms. While they fixate on grid delays, they miss that APLD’s site selection strategy targets locations with existing, underutilized high-voltage infrastructure, effectively bypassing the 'multi-year' bottleneck others assume. This isn't just about construction; it's about securing scarce, energized land. The valuation isn't just pricing execution; it's pricing the scarcity of power-ready sites in a market where the grid is the primary barrier to entry.
"Solving grid interconnection doesn't solve power procurement or demand cycle risk over a 15-year horizon."
Gemini's grid-bypass argument assumes APLD's site selection solves interconnection risk, but doesn't address power procurement itself. Securing energized land ≠ securing stable, long-term power contracts at locked rates. Hyperscalers' AI capex cycles are also cyclical; a 2027 start date means APLD bets on sustained GPU demand through 2042. That's a 15-year bet on a 2-year-old technology cycle. The backlog's durability hinges on power availability AND demand persistence—not just real estate arbitrage.
"Site selection does not solve power-contract or demand-duration risks that could still void the 2027 start and backlog value."
Gemini overstates the FERC reforms and site-selection edge. Energized land still requires locked-in power purchase agreements at viable rates, which hyperscalers negotiate separately and can walk from if GPU demand slows post-2027. The 15-year lease therefore embeds a bet on sustained AI capex cycles that neither interconnection queues nor real-estate scarcity can guarantee, leaving the $31B backlog exposed to the same demand and procurement risks Claude flagged.
The panel is neutral on APLD, with concerns around execution risks, customer concentration, and the durability of AI demand outweighing the transformative $7.5B revenue contract and 15-year lease duration.
The $7.5B revenue contract and 15-year lease duration
Customer concentration and the durability of AI demand