$APLD Stock: Why I'm Bullish on Applied Digital and Why I Plan to Buy It

I am bullish on $APLD because the story has moved from "interesting AI data center idea" to "contracted AI factory platform with real hyperscaler demand behind it." That does not make it safe. It does not make it cheap at any price. It does mean I am paying attention, and I plan to build exposure carefully instead of treating this like a meme trade.

This is my personal investing thesis, not a recommendation for you to buy the stock. Applied Digital is volatile, capital intensive, and still has real execution risk. But if I am looking for a public-market way to play the AI infrastructure bottleneck, APLD is one of the few names where the bull case is tied to something concrete: power, land, construction, long-term leases, and customers that need capacity badly.

The latest market-data check I used showed APLD at $37.30 on the June 30, 2026 close, with after-hours trading around $37.43, a 52-week range of $9.02 to $50.73, market cap around $10.79 billion, and short interest near 28.8% of float. That matters because the setup is not a sleepy value stock. It is a high-beta AI infrastructure name with a crowded debate around it.

The AI trade started with chips. Then it moved to servers, networking, cooling, and power. The next stage is simple: whoever can deliver energized data center capacity at scale gets paid. AI models do not run on press releases. They run on power, cooling, real estate, fiber, permits, equipment, and operators who can bring it all online.

That is where Applied Digital fits. The company designs, builds, and operates high-performance data centers and colocation services for AI, cloud, networking, and blockchain workloads. The old version of the story was crypto hosting plus ambition. The new version is HPC hosting and AI factory campuses leased to hyperscale customers.

The reason I am bullish is not just that AI demand is huge. Everybody knows that. The reason is that APLD appears to have something the AI market is desperate for and cannot manufacture overnight: large-scale power-connected campuses with long-term contracts attached.

The biggest reason I am updating the APLD thesis is the contract base. In April 2026, Applied Digital announced a 300 MW Delta Forge 1 lease with a U.S.-based high investment-grade hyperscaler. The company said that lease represented about $7.5 billion of total contracted value over an estimated 15-year term and pushed total contracted lease revenue above $23 billion.

Then in May 2026, Applied Digital announced another 300 MW lease at Polaris Forge 3. That release said total contracted lease revenue had reached $31 billion across four AI Factory campuses, or $73 billion if all renewal options are exercised under their terms.

Then in June 2026, APLD signed a 210 MW Delta Forge 2 lease. The company said that pushed the contracted portfolio to five AI Factory campuses, 1.4 GW of critical IT load, about 2.15 GW of grid-connected utility power, and roughly $36 billion of total contracted base-term lease revenue, or $86 billion if all renewal options are exercised.

That is the line that made me stop treating APLD as a cute AI infrastructure side bet. A company with roughly $36 billion of contracted base-term lease revenue is a very different animal than a company trying to pitch a dream deck.

Early in the APLD story, the obvious criticism was customer concentration. CoreWeave was the big name. That is still important, and I still track it. But the newer announcements point to a broader customer base, including U.S.-based investment-grade hyperscalers.

The May release said approximately 65% of contracted revenue was backed by U.S.-based investment-grade hyperscalers. The June Delta Forge 2 release raised that language to approximately 70% of contracted revenue backed by U.S.-based investment-grade hyperscalers.

That does not remove risk, but it changes the quality of the backlog. If your leases are backed by stronger credits, lenders care. Investors care. Your cost of capital can improve. The whole model becomes more financeable.

CoreWeave is still central to the story because it proved there was real demand for what Applied Digital was building. APLD's fiscal third-quarter 2026 update said the first 100 MW HPC data center at Polaris Forge 1 was operational, the second 150 MW building was being built and expected to come online in calendar 2026, and the third 150 MW facility was anticipated in calendar 2027.

The same update said APLD had entered agreements with CoreWeave intended to improve the credit quality of tenants under the 100 MW and 150 MW leases at Polaris Forge 1. That included restructuring through a CoreWeave SPV subsidiary, springing guarantees from CoreWeave Inc., and a $50 million letter of credit tied to one lease.

That is not meme-stock stuff. That is credit plumbing. It matters because these assets are expensive, the debt stack is real, and better credit support can lower the market's fear that APLD is just building expensive boxes for fragile customers.

In fiscal Q3 2026, Applied Digital reported revenue of $126.6 million, up 139% from the prior-year quarter. The company also reported adjusted EBITDA of $44.1 million. The important detail was the HPC Hosting business: revenue from HPC Hosting totaled $71.0 million, including base rent, tenant fit-out services, power pass-through arrangements, and other ancillary streams.

That matters because the first 100 MW facility only represented the beginning of the contracted platform. Management said that initial 100 MW was about one-sixth of contracted capacity and one-tenth of what was operating or being built at the time. If the next buildings come online on schedule, the model should become easier to judge by cash flow instead of vibes.

The company was still not cleanly profitable on a GAAP basis. Net loss attributable to common stockholders was $100.9 million for the quarter. That is exactly why I do not want the article to sound like a victory lap. The bull case is execution from here, not pretending the income statement is already perfect.

The market is good at pricing what already exists. It is worse at pricing infrastructure that is partially built, partially leased, heavily financed, and still ramping into service. That is the opportunity and the danger with APLD.

If the company delivers buildings on schedule, converts contracted revenue into durable rent, lowers financing costs as assets stabilize, and keeps adding investment-grade hyperscaler customers, the market may eventually stop valuing it like a risky transition story and start valuing it more like a scarce AI infrastructure platform.

That rerating is the upside. Not a magic price target. Not a promise. Just the possibility that APLD moves from "story stock" to "contracted infrastructure owner/operator" as the financials catch up to the leases.

I do not want to chase APLD emotionally after a green candle. I want to build the position around a thesis checklist. My plan would be to scale in, keep the position small enough that a drawdown does not force a bad decision, and add only when execution confirms the story.

That is the ADHD-friendly version of the trade plan: small first, evidence-based adds, no heroic all-in move. The goal is to participate if the company executes, not to turn one exciting ticker into a personality test.

One: power is the bottleneck. Chips are useless without power and cooling. APLD's value is tied to sites where power can actually be delivered.

Two: the contracts are now real enough to matter. Management is not only talking about demand. The company has announced long-term take-or-pay leases across multiple campuses.

Three: customer quality is improving. More contracted revenue is tied to U.S.-based investment-grade hyperscalers, according to company releases. That helps the financing story.

Four: the first HPC revenue stream is visible. Fiscal Q3 showed meaningful HPC Hosting revenue and adjusted EBITDA, while the next buildings are still coming.

Five: the stock is controversial. High short interest and high volatility can make the ride ugly, but controversy is part of why the upside can exist if execution keeps improving.

The first real risk is execution. These campuses are massive construction projects. Delays, cost overruns, equipment problems, power interconnection issues, or commissioning failures can damage the thesis quickly.

The second risk is financing. As of February 28, 2026, Applied Digital reported $2.1 billion in cash, cash equivalents, and restricted cash, plus $2.7 billion in debt. The company has also announced large debt facilities and project financings. That capital stack needs discipline. If debt gets too expensive or equity dilution becomes heavy, shareholders can still get hurt even if the buildings eventually work.

The third risk is customer concentration. Better customer quality helps, but these are still huge leases tied to a concentrated group of hyperscale AI buyers. If AI capex slows, if a tenant's economics weaken, or if a customer tries to renegotiate, the stock will feel it.

The fourth risk is valuation. APLD had already run hard by the latest quote check, with MarketWatch showing the stock up more than 250% over one year and still trading with a high beta. A great business story can still become a bad entry if the price already discounts too much perfection.

The fifth risk is the AI cycle itself. If AI infrastructure spending cools, GPU demand slows, or hyperscalers pause new capacity, APLD's story gets less urgent. This is not a defensive stock. It is tied to a massive capex cycle.

I would cool off fast if buildings slip materially, if financing costs rise faster than lease economics, if customer-credit quality weakens, if management starts overpromising, or if the stock prices in the entire bull case before the cash flow arrives.

I would also back away if the company starts leaning too heavily on promotional language instead of measurable milestones. I want megawatts online, leases funded, revenue recognized, debt refinanced on better terms, and customers paying. Everything else is noise.

That checklist keeps me honest. The story is exciting, but the stock does not care about my excitement. It cares about execution, financing, and whether the AI infrastructure cycle keeps pulling capacity forward.

For me, APLD is not a full-size core holding. It is an aggressive growth position. I would rather own a smaller amount with conviction and patience than own too much and panic on the first 30% drawdown.

If someone cannot emotionally handle a sharp drop, they should not size this like a safe compounder. I am bullish, but bullish does not mean reckless. The whole point is to survive long enough for the thesis to play out.

I am bullish on APLD because the company sits at the intersection of the AI power bottleneck, hyperscaler urgency, long-term lease demand, and infrastructure scarcity. The 2026 announcements made the thesis more real: five AI Factory campuses, 1.4 GW of contracted critical IT load, and roughly $36 billion of contracted base-term lease revenue according to company releases.

That is why I plan to invest in it carefully. Not because it is guaranteed. Not because the ticker is hot. Because if Applied Digital executes, the market may eventually have to treat it less like a speculative AI story and more like a scarce AI infrastructure platform with long-duration contracted revenue.

My move is simple: respect the upside, respect the risk, scale in with discipline, and let the company prove the thesis one quarter at a time.

Applied Digital fiscal Q3 2026 results for revenue, adjusted EBITDA, HPC Hosting revenue, debt/cash figures, construction timing, and risk language.

Delta Forge 1 lease announcement, Polaris Forge 3 lease announcement, and Delta Forge 2 lease announcement for contracted capacity, lease value, and hyperscaler-credit details.

Applied Digital revolving credit facility announcement for the $550 million credit facility and financing context.

MarketWatch APLD quote page for the latest quote, market cap, 52-week range, beta, short-interest, and performance context checked before writing.

*Disclaimer: MentorSurge is not a financial advisor and this is not financial advice. This post is for educational and entertainment purposes only. Nothing here is a recommendation to buy or sell any security. This is a first-person investing thesis, not personalized advice. APLD is volatile, capital intensive, and risky. Investing involves substantial risk of loss. Numbers cited were checked before publication and can change quickly. Always do your own research and consult a licensed professional before making decisions with real money.*