By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel is bearish on TeraWulf (WULF) and Applied Digital (APLD), citing execution risks, equity dilution, and financing concerns related to their multi-year data center builds and power-as-a-service models.
Risk: Equity dilution due to potential massive secondary offerings and refinancing risks.
Opportunity: None identified.
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 →
Access to sufficient computing power has become a major constraint for artificial intelligence systems. This explains why hyperscalers are not just rushing to build their own data centers, but also sealing long-term deals for more compute with neocloud companies like Applied Digital (NASDAQ: APLD) and TeraWulf (NASDAQ: WULF).
TeraWulf made the news recently for the 20-year, $19 billion deal it just inked with Anthropic. That agreement covers 401 megawatts of critical IT load, which will become available in waves. The full 401 megawatts should be online by early 2028.
Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now, when you join Stock Advisor. See the stocks »
That announcement earned TeraWulf a price target adjustment from Morgan Stanley's analyst, who bumped it to a Street-high $72. That implies that the stock will more than triple from current levels in the next 12 months. It isn't just good news for TeraWulf. It points to broader tailwinds that will also lift Applied Digital.
When a deal like the Anthropic one is announced, it doesn't translate into immediate revenue. Neocloud companies are investing heavily into building AI data centers and have multiple construction projects underway. That's why the full 401 megawatts that the AI giant is contracting for won't be available until early 2028.
Applied Digital touted in an investor presentation that it has 3 gigawatts of active pipeline projects, while TeraWulf only has 2.3 gigawatts in its portfolio. Securing more gigawatts of electricity to power future data centers increases a company's earnings potential, so Applied Digital has the edge in that regard.
However, anytime a company adds a new data center site, it isn't small. Those sites often have hundreds of megawatts. TeraWulf or Applied Digital can suddenly come out with an announcement saying that they got another AI data center site, which can either close or expand the gap by a meaningful margin. When it comes to the quantity of gigawatts, Applied Digital is currently ahead, and that gives them a higher ceiling.
Although TeraWulf has fewer AI data centers, it does have an edge over Applied Digital when it comes to power. TeraWulf makes it a point to own its power, while Applied Digital signs long-term electricity supply agreements with utility companies.
Applied Digital's approach is cheaper right now and lets it complete AI data centers sooner. It also requires its customers to bring their own AI chips and servers, while TeraWulf provides computing hardware in its facilities. These differences make it easier for Applied Digital to realize more revenue at a faster rate, but its business model also makes it dependent on the electric grid. Requiring customers to bring their own hardware also lowers how much Applied Digital can charge for each megawatt of critical IT load.
An overstrained electric grid can cause issues, and when Applied Digital renegotiates utility leases when they expire, the company may have to pay much higher prices. That scenario is especially possible as a growing number of AI data centers will be competing for the same power supply.
TeraWulf develops on-site power generation assets at its data centers. This strategy means it takes a little longer for its data centers to be completed, but it also ensures that TeraWulf won't have to rely on the power grid. It incurs higher costs now for more control over future costs and power availability. In the long run, it is much better to own power generation capacity than to lease it.
TeraWulf has 923 megawatts of critical IT load contracted to clients. Anthropic makes up almost half of that total. TeraWulf is aiming to support 250 megawatts to 500 megawatts of additional critical IT load signings per year, which could result in meaningful net operating income growth once the sites are fully developed.
TeraWulf is targeting an 85% net operating income margin on contracts, showing that profits can scale quickly as well.
Applied Digital has 1.41 gigawatts of contracted critical IT load. Once again, Applied Digital has a slight edge, but a single announcement from either of these companies can meaningfully close or expand the gap. For instance, TeraWulf's contracted critical IT load jumped from 522 megawatts to 923 megawatts on a single Anthropic deal.
Applied Digital also signs long-term deals with tech giants. The company recently secured a 15-year take-or-pay lease with an unnamed, high investment-grade hyperscaler that is based in the U.S.
The deal covers 210 megawatts of critical IT load for approximately $5.2 billion over 15 years. The contract's value can reach $12.7 billion if all renewable options are exercised over a 30-year term.
Applied Digital has an edge when it comes to total gigawatts and contracted critical IT load. However, a single deal from TeraWulf could close these gaps. The main advantage of TeraWulf is that it owns its power, which could matter a lot in the years ahead.
Before you buy stock in TeraWulf, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and TeraWulf wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $395,679! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,294,805!
Now, it’s worth noting Stock Advisor’s total average return is 929% — a market-crushing outperformance compared to 211% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of July 13, 2026. *
Marc Guberti has no position in any of the stocks mentioned. The Motley Fool has no position in any of the stocks mentioned. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"TeraWulf’s headline deal is real but the $72 target embeds flawless execution on power ownership that history suggests is optimistic."
The Anthropic 20-year, $19B deal for TeraWulf (WULF) validates surging AI compute demand and justifies Morgan Stanley’s $72 target (implying >3x from ~$22 levels). Applied Digital (APLD) leads on pipeline (3 GW vs 2.3 GW) and contracted load (1.41 GW vs 0.923 GW), plus faster revenue ramp via customer-supplied hardware and cheaper grid PPAs. Yet TeraWulf’s on-site power ownership offers superior long-term cost and availability control as grids strain. Both stocks trade at high multiples on pre-revenue megawatts; execution risk on multi-year builds remains acute.
The article glosses over massive capex needs, execution risk on 2028 full delivery, and potential hyperscaler concentration—Anthropic is nearly half of WULF’s contracted load. Grid-independent power may prove far costlier than forecast if renewable or nuclear projects slip, while APLD’s leasing model could face sharp electricity price resets.
"The market is conflating contracted gigawatt potential with actual cash-flow generation, ignoring the massive capital intensity and execution risks required to bring these sites online."
The market is currently pricing WULF and APLD as pure-play infrastructure winners, but investors are underestimating the execution risk inherent in the 'power-as-a-service' model. While WULF’s ownership of power assets provides a long-term moat against grid congestion, the capital expenditure required is immense, potentially leading to massive equity dilution. Conversely, APLD’s model is faster to scale but leaves them exposed to utility rate hikes and potential margin compression as the power market tightens. The $19 billion figure for the WULF-Anthropic deal is a headline-grabber, but investors must discount that heavily for time-value and the high probability of cost overruns in site development.
If these firms successfully secure long-term, fixed-rate power purchase agreements or generation assets, they essentially become high-yield utility proxies with AI-driven growth premiums, making current valuations look cheap in hindsight.
"TeraWulf's 3x upside case rests on a single customer (Anthropic) representing 50% of contracted load and assumes power ownership beats capital efficiency—both unproven and high-risk."
The article frames this as a pipeline race, but misses the critical leverage question. TeraWulf's Anthropic deal is 50% of its 923 MW contracted load—massive concentration risk if Anthropic delays, pivots to in-house capacity, or renegotiates. Applied Digital's 1.41 GW is more diversified. More importantly: the article assumes power ownership is unambiguously superior, but ignores that TeraWulf's on-site generation requires massive upfront capex, longer build timelines, and execution risk on power assets—not just data center construction. Applied Digital's utility-lease model is capital-efficient and de-risks delivery. The Morgan Stanley $72 target (3x upside) appears anchored to the Anthropic deal size, not to profitability or cash flow—a classic deal-announcement euphoria trap.
If grid constraints actually materialize and power becomes the binding constraint rather than capex, TeraWulf's owned generation could command a structural premium that justifies the higher capex and longer timelines. Applied Digital could face lease renegotiation shocks that crater margins.
"Big contracts and pipelines don’t guarantee near-term profitability; energy costs, execution risk, and potential hyperscaler shifts to in-house compute could truncate upside."
Opening take: the Anthropic deal is eye-catching but may not translate into near-term cash flow. The article emphasizes gigawatt pipelines and contract signs, yet revenue recognition is years away and depends on continued AI demand. Economically, APLD’s model relies on customers supplying hardware, which compresses margins and increases customer bargaining power; WULF’s on-site generation raises capex and balance-sheet risk even as it reduces grid exposure. Valuation may assume a hyper-accelerated AI buildout that could be disrupted by higher interest rates, supply chain constraints, or hyperscalers choosing in-house capacity. In short, the upside hinges on a fragile, long-dated cycle with meaningful execution and energy-cost risks.
The strongest counter: even massive signed deals may never translate into meaningful near-term profits if AI demand cools, hyperscalers accelerate in-house compute, or grid and financing costs bite harder than anticipated.
"Refinancing walls in 2026-28 pose unmentioned systemic risk to both WULF and APLD leverage profiles."
Claude correctly flags WULF's Anthropic concentration, yet nobody has noted the second-order financing risk: both firms' multi-year builds coincide with potential 2026-2028 debt maturity walls and tighter bank lending standards on AI-exposed credits. If hyperscalers slow capex, refinancing costs could spike 300-400bps, wiping out any owned-power margin edge.
"Refinancing risk will be mitigated by equity dilution, which is a more immediate threat to shareholder value than debt service costs."
Grok, your focus on debt maturity walls is critical, but you overlook the equity-linked nature of these firms. Both WULF and APLD are essentially 'growth-at-any-cost' vehicles that will likely solve refinancing risks through massive secondary offerings rather than credit markets. This creates a permanent dilution trap for retail holders. While you worry about interest rate spreads, the real risk is that these companies prioritize capacity expansion over EPS, ensuring they remain cash-flow negative long after the initial hype cycle fades.
"WULF's long-term Anthropic contract may actually reduce refinancing risk vs. APLD, but only if power assets deliver on cost and timeline."
Gemini's equity-dilution thesis assumes secondary offerings are inevitable, but ignores that WULF's Anthropic deal locks in 20-year revenue at fixed rates—precisely the collateral that enables non-dilutive debt financing. APLD faces worse dilution risk because it lacks that anchor. The real question: does WULF's power-asset moat justify the capex, or does refinancing risk still crater returns before the moat compounds? Grok's 300-400bps spread widening is plausible but timing matters enormously.
"Anthropic concentration could become a financing trigger, not just volume risk, forcing tighter terms and potential dilution."
Claude's concentration risk is real, but the financing knock-on is the missing twist: Anthropic isn't just a load, it's a counterparty credit trigger. If Anthropic delays or renegotiates, lenders reprice project-financing around that risk, tightening terms and potentially forcing reserve draws or covenants that ripple into equity needs. That offsets WULF's moat and could hasten dilution or capex constraints, making the 'power-as-a-service' thesis a financing risk as much as an operational one.
The panel is bearish on TeraWulf (WULF) and Applied Digital (APLD), citing execution risks, equity dilution, and financing concerns related to their multi-year data center builds and power-as-a-service models.
None identified.
Equity dilution due to potential massive secondary offerings and refinancing risks.