APLD 3Q25 - AI, ever heard of it? Bull case $70. Bear case $18

Cheat Sheets presented by stocktwits

October 09, 2025

Applied Digital Investment Report - Q3 FY2025 Beat & $11B CoreWeave AI Infrastructure Deal

APPLIED DIGITAL CORP. (APLD)

🖥️⚡ Building AI Data Centers | $11B Revenue Locked In, 700MW Building Now, +24% After Hours

Sales +84% | Actually Building (Not Just Talking) | $20-25B Funding Secured

💰 Market Cap: $9.75B | 🏢 205 Employees | 🌍 AI Data Center Builder
👨‍💼 CEO Wes Cummins | 🎯 AI Data Centers | 🏆 Best Data Center Americas 2025
$36.25
📈 +$6.96 (+23.76%) After Hours
~+900% YTD
Price Target Scenarios

18-Month Targets (2026) - From Current $36.25

$70.00
Bull Case (+93%)
$1B+ Profit Target Hit
Multiple Customers
📊 Math:
• 2026 Revenue: $900M × 15x multiple = $13.5B / 195M shares = $70
🚀 What Needs to Happen:
All 400MW online by end 2026 • PF2 signed + building on schedule • 2-3 new customer deals close • 4GW pipeline → 2GW building • Everything done in 12-14 months • Financing closes as planned • $500M+ profit from CoreWeave + PF2 doing same • Path to $1B profit clear • AI spending stays $300B+ per year
$48.00
Base Case (+32%)
CoreWeave Works
Steady Progress
📊 Math:
• 2026 Revenue: $700M × 13.5x = $9.5B / 195M shares = $48
⚖️ What Needs to Happen:
First 100MW online end 2025 • PF2 signed Q2 FY2026 • 1-2 more deals from pipeline • 700MW done on 12-14 month schedule with small delays • Financing closes this quarter • Revenue ramps as expected • $500M+ profit visible • Cloud business sells • AI demand stays strong
$18.00
Bear Case (-50%)
Delays Hit
AI Slows Down
📊 Math:
• 2026 Revenue: $350M × 10x = $3.5B / 195M shares = $18
🔻 What Could Go Wrong:
First 100MW late past Q1 2026 • 700MW hits delays at multiple sites • Wes's 2027-28 shakeout includes APLD • PF2 doesn't close or takes 6+ months • No new deals happen • CoreWeave slows down • Too many data centers get built • Financing worse than expected • Labor/supply problems • AI spending drops $350B to $200B • Competitors build their own • Stock crashes on any miss
Bottom Line
🏆
Actually Building NOW
700MW building RIGHT NOW. Not future plans - dirt moving today. $1.6B already spent. First 100MW online end 2025. 4GW pipeline (6-12 months out). Supply chain locked in 2 years ago. 4 big tech companies approved design. Real execution.
🎯
Money Problem Solved
$5B Macquarie deal unlocks $20-25B total. No more selling stock to raise cash. Project loans coming this quarter for PF1. Between Macquarie + project finance, PF1 fully funded. Can scale to massive size. Biggest question = answered.
📈
Pipeline Exploding
PF2 lease "very soon" with mega tech company. Talking to 2 MORE for 2 NEW sites. 50+ sites shown last 4 weeks. Others have power, need APLD to build. Revenue $64.2M Q1 (+84%). Build-out revenue jumps next quarter per CFO. Then lease money starts. $500M annual profit from CoreWeave alone. $1B profit in 5 years clear. Stock +23.76% after hours.
⚠️
High Risk = High Reward
Still losing money. 700MW = huge risk. First 100MW MUST hit end 2025. Only CoreWeave customer until new deals close. Wes warns 2027-2028 "shakeout" coming - could hit them. Stock +23.76% after hours to $36.25 = up 900%+ YTD. Valuation at $9.75B stretched. 6.59 beta = wild swings. Not for scared money. Growth with binary outcomes.

📊 What Just Happened: They're Executing

Revenue +84% • Building 700MW NOW • Got $20-25B to Build More • Another Deal "Very Soon" • 50+ New Sites Shown • First Goes Live End 2025 • Stock +24% After Hours to $36.25 🚀

💰
$64.2M
Q1 FY2026 Revenue
+84% YoY
⚡
700MW
Under Construction
Multiple Sites
🎯
$0.5M
Adj EBITDA Q1
Near Breakeven
🚀
4GW
Active Pipeline
6-12 Months Out

📞 Key Earnings Call Points

🎊
Revenue Crushing, Stock Rips +24%
$64.2M sales (+84%). $26.3M from CoreWeave setup (low margin but proves trust). Building 700MW across multiple sites NOW. First done end 2025, starts making money. Almost breaking even. Revenue jumps next quarter, then real money when first site goes live. Market loved it = +24% after hours to $36.25.
🚨
Got $20-25B to Build
Pulled $112.5M from $5B Macquarie deal. That $5B unlocks another $15-20B = $20-25B TOTAL. Translation: Can build massive without selling more stock (no dilution). Getting project loan this quarter at good rates. Money problem = SOLVED. Can scale big now.
💎
Next Deal "Very Soon"
Breaking ground on campus 2 (PF2). 300MW starting, grows to 1,000MW. $3B to build. First online late 2026, full power 2027. CEO says lease with investment-grade mega tech (Google/Microsoft level) closing "very soon." Similar money to CoreWeave. ALSO talking to 2 MORE giants for 2 MORE locations. Demand crazy: need 200MW sites growing to 1 gigawatt. Some want 10+ gigawatt sites!
🔬
Pipeline Going Crazy
4 GIGAWATTS lined up (6-12 months). Last 4 weeks = 50+ new sites shown. Secret sauce: locked supply chain 2 YEARS AGO (transformers, generators, hard stuff). Everyone else scrambling, APLD ready. Can build in 12-14 months vs 24 for others = HUGE edge. 4 major tech companies approved design. Getting flooded from people who grabbed power but can't build - APLD steps in. Can do 1-2+ more Dakota projects at once, expanding to other states.

📍 CEO Wes - Big Picture: Big Tech dropping $350B on AI data centers THIS YEAR. Apollo moon program cost $150B over a decade. Interstate Highway $500B over 30 YEARS. AI companies spending more than that in ONE YEAR. Wes: bottleneck isn't chips anymore - it's "not enough data centers to put chips in." APLD "uniquely positioned" - started building 2023 before others caught on. Supply chain locked early, hired best people. "We're not just watching, we're making it happen."

🌍 The Math: CoreWeave alone = ~$500M per year when running. Company goal: $1B per year in 5 years (profit, not sales). First campus (PF1) grows to 1.4 gigawatts by 2028-2030. Second (PF2) hits 1+ gigawatts. Building North Dakota: cheap power, free cooling 200+ days/year, helpful government. South Dakota has power ready 2026, just needs tax break. Government estimates 40-50 gigawatts SHORT. Google's Eric Schmidt says 90+ gigawatts short. APLD has YEARS of demand.

💰 How They Pay: Macquarie deal is key. Huge bank that knows data centers, has big tech relationships. $5B means build without selling stock and diluting shareholders. Project loans at good rates (4-4.5% above base). Loan structures getting better as tech leases prove out. Already $1.6B property/equipment built. This is REAL.

🔮 Why APLD Wins: "Flooded with opportunities" - 50+ sites in 4 weeks. Super picky: only great locations spread out geographically matching supply chain. New trend: others grabbed power/land but can't build these specialized centers. APLD builds for them. Demand shifting BIGGER - companies want gigawatt+ in one spot. Common ask: "200MW fast, but make it grow to gigawatt." Some want 10X that. Wes predicts 2027-2028 "shakeout" where newer companies who don't know what they're doing hit delays, proven builders (like APLD) get more business. Personal: Wes from small town Idaho, saw cities get opportunities while rural areas left behind. That's why investing in Ellendale and Harwood - bringing jobs, building community, minimizing environment impact. "This is the legacy I want."

🚀 Bull Case
💎
$11B + More Deals Coming
$11B CoreWeave locked, ~$500M annual profit. PF2 with investment-grade customer "very soon" per Wes. Talking 2 MORE for 2 NEW sites. 4GW pipeline (6-12 months). 50+ sites shown last month. Others have power but can't build. Multiple revenue streams = less CoreWeave risk.
⚡
Supply Chain = Moat
Locked supply 2 years ago. Bought out transformer/generator capacity (long lead stuff). 50% fewer parts. Competitors facing delays and price increases, APLD unaffected. 700MW building NOW proves it works. Can do 1-2+ more Dakota sites parallel. Expanding other states. Early-mover edge can't be copied fast.
👑
$350B AI Spending Wave
Big tech spending $350B on AI THIS YEAR - more than Apollo or Interstate Highway but in one year. Wes: bottleneck moved from GPUs to "not enough data centers." DOE: 40-50GW power short. Eric Schmidt: 90+ GW. Demand now: 200MW sites to 1GW+. Some want 10+ GW. Industry just starting. 20-30 year cycle.
🚀
$20-25B Capital, No Dilution
$5B Macquarie unlocks $20-25B total with project finance. No more stock sales = zero dilution. Already $1.6B built. First 100MW online end 2025 = lease money starts. Adj EBITDA $0.5M (breakeven). Path to $1B profit in 5 years clear. Math: PF1 = $500M profit, PF2 similar, more sites = $1B+ doable. Stock rerates hard when profit ramps.
🐻 Bear Case
📉
CoreWeave Still Only Customer
All 400MW at PF1 = one customer CoreWeave. PF2 not signed yet (just "very soon"). If CoreWeave demand shifts, major hit. Even with new deals coming, risk signing them. "Advanced talks" isn't signed contracts. Need real diversification, not promises. First mover can also be first to fail.
🏥
700MW Build = Huge Risk
Building 700MW at multiple sites with 205 people. Wes admits 2027-2028 sees "shakeout" where "new entrants learn lessons" and projects delay. If APLD wrong about their lessons, they're the ones delaying. Weather (North Dakota!), labor shortages, permits, tech problems. Any 6-month delay on first 100MW = story breaks, stock crashes 50%+.
💸
Still Burning Cash
Lost $27.8M Q1 ($0.11/share). Adj EBITDA barely positive $0.5M. Won't profit for quarters. $687M debt, need ongoing financing. $5B Macquarie = dilution at subsidiary (Macquarie gets equity). Already raised $270M+ selling stock. If project financing delayed or worse terms, more dilution coming. Stock up 900%+ YTD on hopes, not profits. After hours pop to $36.25 = valuation stretched more.
⚠️
Market Could Turn Fast
What if $350B AI spending slows? Big tech building own centers. Industry sees "big grab for power" but Wes admits some grabbed power and "don't know what to do." Oversupply risk building. 50+ sites shown to APLD in 4 weeks = lots of competition. If AI hype fades 20%, APLD valuation ($9.75B on $216M FY26 revenue) implodes. 6.59 beta = extreme swings both ways. Stock +23.76% after hours = pure momentum.

This is for info only, not investment advice. Talk to a financial advisor before investing.