Applied Digital Corporation

Overview

This is a Keith Gill / Roaring Kitty-style deep value and narrative analysis of Applied Digital Corporation (APLD) as of August 16, 2026. APLD is a battleground AI infrastructure company trading at $31.20 with a market cap near $9 billion against roughly $36 billion in contracted base-term lease revenue. The report asks whether the market is too focused on GAAP losses, stock-based compensation, and construction execution risk, while under-pricing a business transitioning from speculative developer to contracted AI data center landlord.

The Bear Case

The consensus bear story is easy to tell. APLD reported a fiscal 2026 net loss of $249.2 million and a Q4 net loss of $110.6 million. Stock-based compensation was enormous: $220.1 million for fiscal 2026, including $116.8 million in Q4 alone. GAAP EPS for fiscal 2026 was negative $0.91. The balance sheet carries approximately $5 billion in debt against $1.7 billion of reported equity, and the stock trades at 5.23 times book value. Bears also point out that revenue quality is messy: Q4 services revenue included $152.4 million of tenant fit-out services, which is lower-margin and less recurring than core lease rent. Customer concentration is severe: roughly $20 billion of the $36 billion contracted base-term lease value comes from one high-investment-grade hyperscaler, and the company has only three announced tenants across five campuses. The stock is down 38% from its 52-week high of $50.73, trading below its 50-day moving average of $34.11 and its 200-day moving average of $32.43. Zacks has flashed a Very Strong Sell style signal, and Seeking Alpha has published a downgrade arguing the bull case requires near-perfection. The bear conclusion is that at nearly $9 billion in market cap, APLD already prices in flawless multi-gigawatt execution.

The Bull Case

The contrarian thesis is that the market is valuing APLD as a cash-burning developer when management has restructured it into a contracted infrastructure landlord. The company ended fiscal 2026 with 1.41 gigawatts of contracted critical IT load across five campuses, representing approximately $36 billion in base-term contracted lease revenue, or about $86 billion including renewal options. That base-term contracted revenue is roughly four times the current market cap. Management guided to a $1 billion net operating income run rate in about a year, which would be three years ahead of its original target. Q4 net operating income was already $39.9 million at a 91% margin on just 100 megawatts of operating capacity, and total live capacity at Polaris Forge 1 has since reached 175 megawatts. Financing has improved dramatically: project-level notes were placed at 6.75% and 7.00%, compared with the initial 9.25% placement, and CoreWeave has since obtained investment-grade A3 credit enhancements for its leases. The stock's pullback from $50.73 to $31.20 comes even as the contracted backlog and project financing have improved. The bull view is that the market is pricing a punitive cap rate and elevated execution risk, while the actual model is shifting toward long-duration, take-or-pay, investment-grade-backed lease cash flows.

Fundamental Deep Dive

Sentiment & Technical Setup

Catalyst Analysis

The most important catalyst is execution on the existing buildout: delivering Polaris Forge 1 and Polaris Forge 2 on schedule and bringing contracted megawatts to ready-for-service. Management guided to reaching a $1 billion NOI run rate within about a year, which would be a major re-rating event. Additional lease signings are also catalysts: APLD is in advanced negotiations on roughly 250 megawatts of expansion leases that could add more than $6 billion of contracted revenue. New campus announcements are possible, especially given that APLD appears to be evaluating a Brookwood, Alabama site. Financing milestones matter too: the company has already lowered its cost of project debt from 9.25% to 6.75%-7.00%, and further credit enhancement or refinancing would improve returns. The ChronoScale public stub could surface value if the market begins assigning APLD credit for its 96% ownership. Finally, if short interest is elevated, any of these catalysts could trigger a sharp squeeze.

Key Risks

Conclusion

The market is treating APLD as an overleveraged cash-burning developer, but the company has already secured long-duration, take-or-pay contracts that are more characteristic of a data center REIT. The current $31.20 price appears to be pricing a stressed scenario even while the contracted backlog and financing structure have improved. If management achieves its $1 billion NOI run-rate target in about a year, the stock could re-rate substantially, especially with average analyst targets near $74. However, this is not a classic deep value stock with a margin of safety based on liquid assets; it is a leveraged infrastructure build with real execution risk. The setup favors patient bulls, but position sizing and respect for downside volatility are essential.

Overview

A Peter Lynch-style fundamental review of Applied Digital Corporation (APLD) using its FY2026 results, balance sheet, contracted lease backlog, and recent news to classify the stock, assess valuation and insider signals, and give a HOLD/BUY/SELL view.

The Two-Minute Story

Applied Digital is best understood as a landlord and developer for AI data centers. It buys land, secures power, builds massive buildings, and signs long-term take-or-pay leases with hyperscalers such as CoreWeave and an investment-grade U.S. hyperscaler. Once a building is finished and energized, rent starts. The company says it already has about $36 billion in contracted base-term lease revenue across 1.41 gigawatts, which is roughly four times its market value. The catch is that only 175 megawatts is live today and the rest must still be built, financed, and connected to power. If management delivers on time and finances the construction without heavy dilution, earnings could inflect sharply. If construction slips or credit gets expensive, the equity is the shock absorber.

Stock Category

Do You Understand This Business?

The business is understandable for an average person: Applied Digital is a landlord for power-hungry AI computing facilities. It does not own the GPUs; it owns the real estate, power access, and cooling. The edge for an investor is following power agreements, lease announcements, and construction milestones, because those translate directly into future rent. The complexity is in the balance sheet and project financing, not in the product.

PEG Ratio Analysis

Lynch's Checklist

Tenbagger Potential

A tenbagger from $31 would require the stock to reach about $310 and market value near $90 billion. That is possible only if the company successfully builds out most of its 5 GW vision, converts tens of billions in contracted rent into stable NOI, avoids major dilution, and gets valued like a premium data center landlord. That is a multi-year, execution-heavy outcome, not a near-term base case. The current sell-side bull case of around $75 is not a tenbagger, and the bear case around $16 shows the other side.

Key Risks

Conclusion

Peter Lynch would respect the simple landlord story and the gigantic contracted backlog, but he would be wary of a fast grower that is not yet GAAP profitable, carries $5 billion in debt, and has insiders selling rather than buying. The valuation cannot be supported by a normal PEG ratio because earnings are negative. This is a show-me story: if management starts delivering buildings on time, refinances project debt successfully, and reaches the $1 billion NOI run rate, the stock could work. Until those items appear on the income statement, the risk of dilution and execution failure is too high for a full buy, but the contracted demand and large runway are too strong to call it a sell.

Overview

A Joel Greenblatt Magic Formula-style review of Applied Digital Corporation (APLD) as of August 16, 2026. The analysis focuses on buying good businesses cheaply using EBIT/Enterprise Value for earnings yield and EBIT/(Net Working Capital + Net Fixed Assets) for return on capital, while normalizing temporary distortions and requiring patience.

Business Quality Assessment

Applied Digital is not currently a high-return-on-capital business in the Magic Formula sense. GAAP operating income is deeply negative: FY2026 operating loss was approximately -$236.5 million. On a more generous adjusted basis, management reports adjusted operating income of about $57.8 million for FY2026, and total segment operating profit for Data Center Hosting plus HPC Hosting was roughly $87.5 million. Against an estimated invested capital base of roughly $4.7-$8.8 billion, the adjusted ROIC is only about 0.6%-1.9%, and GAAP ROIC is negative. There is no history of high returns: FY2024 and FY2025 also produced GAAP operating losses. The low returns reflect enormous capital intensity, construction-in-progress not yet earning rent, heavy stock-based compensation, and one-time tenant fit-out pass-through revenue. The company may eventually earn real-estate-like returns on stabilized AI data centers, but it has not demonstrated that yet.

Valuation Analysis

Using market cap of about $8.98 billion, total debt including finance leases of about $5.03 billion, and cash plus restricted cash of about $4.15 billion, enterprise value is approximately $9.86 billion. Against adjusted operating income of about $57.8 million, the current earnings yield is only about 0.6%. Even using total segment profit of about $87.5 million, the earnings yield is still under 1%. Adjusted EBITDA of $107.2 million implies EV/Adjusted EBITDA of roughly 92x. This is not cheap relative to current earnings power and is far below the threshold that would make a Magic Formula screen attractive compared to bonds or the broader market. The stock is priced for substantial future contracted lease cash flows, not current EBIT.

Magic Formula Ranking

Normalized Earnings Analysis

Current reported earnings are noisy and not representative of normalized owner earnings. FY2026 GAAP net loss was about -$249.2 million attributable to common shareholders, but this includes approximately $220 million of stock-based compensation, a $59.7 million loss on held-for-sale classification, large non-cash fair-value gains on derivatives and investments, and one-time tenant fit-out services with low margin. The fit-out services revenue is largely pass-through construction-type revenue, not durable rental income. Management's non-GAAP adjusted net income was only $36.1 million for the year. A more realistic proxy for current recurring segment operating profit is roughly $87 million, but even that is small relative to the enterprise value and the debt load. The bullish normalized earnings case rests on management's guidance to reach a $1 billion net operating income run rate within about a year; however, Greenblatt's approach emphasizes current earnings power, not future promises, so that future NOI should not be capitalized as if it already exists.

Why The Market Is Wrong

The market is not clearly wrong in the way the Magic Formula requires. The bearish concerns are valid: massive capital spending before rent arrives, project financing risk, customer concentration, power and interconnection constraints, dilution from stock-based compensation and warrants, and GAAP losses. The contrarian bull case is that APLD has $36 billion of contracted base-term lease revenue, with about $20 billion from a high investment-grade hyperscaler, and that as campuses stabilize the company could look like a data-center landlord earning high NOI margins. Applying a 5%-8% cap rate to a future $1 billion NOI run rate supports an enterprise value well above today's price. But that is a venture-capital-like optionality argument, not a current cheap-and-good Magic Formula argument. The market may be overdiscounting execution risk, but APLD is still not a stock where current EBIT/EV and current ROIC justify purchase.

Key Risks

Conclusion

Under the Magic Formula, APLD fails both screens: it is not cheap on current EBIT relative to enterprise value, and it is not a high-return business on current invested capital. The stock is an expensive bet on future AI data-center execution, supported by large contracted leases but burdened by heavy capital intensity, debt, stock-based compensation, and customer concentration. Greenblatt would not buy a company with negative GAAP earnings, sub-1% adjusted earnings yield, sub-1% adjusted ROIC, and a valuation that relies on a future $1 billion NOI run rate. A Magic Formula investor should wait for the company to prove it can earn high returns on its asset base and for the price to reflect current normal earnings, rather than pay upfront for construction success that has not yet happened.

Overview

A Druckenmiller-style top-down macro and reflexivity analysis of Applied Digital Corporation (APLD), assessing whether this AI data center infrastructure buildout is an opportunistic long at current prices or an over-levered story stock vulnerable to a negative feedback loop.

Macro Context

The economy is in a late-cycle AI-driven capital expenditure boom. Central bank policy remains the swing factor: recent inflation data supports a hold rather than aggressive easing, keeping the cost of capital restrictive for capital-intensive developers. The secular trend is powerful: hyperscaler capex has surged from roughly $400B toward $700B annually, with AI infrastructure demand exceeding available power and timely data center supply. Geopolitically, AI dominance and energy security are national priorities, amplifying domestic data center, power generation, and grid investment. In this environment, capital races toward 'picks and shovels' AI infrastructure, but the reflexivity is clearest in financing markets: contract wins lower funding costs, which enables more construction, which supports more contract wins. The same loop runs in reverse when equity prices fall.

Company Position in Macro Landscape

APLD is a direct beneficiary of the AI infrastructure shortage. It has transitioned from crypto hosting toward building and leasing large-scale, purpose-built AI data centers. The company reports 1.41GW of contracted critical IT load across five campuses, approximately $36B of base-term contracted lease value, and roughly $86B including renewal options. About 70% of contracted revenue is backed by investment-grade hyperscalers. However, APLD is also a hostage to macro and financing conditions: it is highly capital-intensive, currently generates negative GAAP earnings, and must continually access debt and equity capital to fund simultaneous multi-gigawatt construction. Its current financial profile—$4.2B cash against $5.0B debt—sounds strong, but much of the cash is restricted and project-linked, and the business is still running negative free cash flow.

Reflexivity Analysis

The positive feedback loop is real and visible: lease signings with investment-grade hyperscalers have allowed APLD to refinance project debt from roughly 9.25% down to 6.75-7.00%, lowering funding costs and strengthening its ability to bid for new contracts. Execution on Polaris Forge 1—delivering 175MW on schedule—supports credibility and access to further financing. The negative reflexivity is also present: APLD shares have fallen 33% from recent highs and now trade below the 50-day and 200-day moving averages, insider selling has occurred, and leveraged short products have been launched against APLD and peers. A further equity decline would raise the cost of equity-linked financing, potentially forcing more dilution, delaying projects, and worsening the very execution risk the market is discounting. Sentiment has shifted from narrative-driven momentum to prove-it mode.

Competitive Position & Disruptive Threats

APLD has a real but contested competitive position. Its advantages include speed of delivery, access to power in less crowded markets such as North Dakota, proprietary waterless cooling, modular design, and qualification with five of six major hyperscalers. These create a 2-3 year temporal moat. However, the moat is replicable by larger, better-capitalized incumbents, and customer concentration is high: three recent leases totaling roughly $20B are with a single hyperscaler, while CoreWeave anchors Polaris Forge 1. Disruptive threats include sustained higher power costs, interconnection and transmission delays, and the possibility that major cloud providers bring more data center design and construction in-house. The company is an execution story as much as a technology story.

Asymmetric Risk/Reward

At $31.20 and an ~$9B market cap, APLD is not a classic deeply asymmetric long. The bull path assumes management reaches a guided $1B NOI run-rate within about a year and the market capitalizes it as a stable infrastructure landlord. At a 6-8% cap rate, that implies enterprise value of roughly $12.5-16.7B, supporting a bull case around $68 and analyst targets near $75. That is over 100% upside. The bear case is a financing and execution failure path: a $16 bear scenario implies roughly 49% downside. Thus the arithmetic skew is roughly 2:1 upside-to-downside, but the probability of the bear path is not negligible given negative GAAP earnings, heavy stock-based compensation, and multi-site construction risk. Optionality exists in the additional 1.7GW of marketed capacity, the ChronoScale stake, and the Base Electron power initiative, but these are path-dependent on balance-sheet strength.

Key Risks

Investment Details

Overview

This report applies William J. O'Neil's CAN SLIM methodology to Applied Digital Corporation (APLD) as of August 16, 2026, using the provided quantitative data plus recent earnings, 10-K, and news sources to evaluate whether APLD qualifies as an institutional-quality growth stock.

Financial and Business Overview

Applied Digital designs, builds, and operates AI data center capacity and also has a legacy bitcoin-mining hosting segment. Fiscal Q4 2026 revenue rose 407% y/y to $258.7M, adjusted EBITDA rose to $42.4M from $1.0M, and adjusted net income was $12.9M, but GAAP net loss attributable to common stockholders was $111.6M, or $0.39 per share. Full-year FY2026 revenue was $611.3M, up 167%, but GAAP net loss was $249.2M. The company has contracted 1.41 GW of AI data center capacity representing about $36B of base-term lease revenue, and management targets a $1B net operating income run rate in about a year. The balance sheet remains aggressive: $4.2B cash including restricted, $5.0B debt, and heavy capex of roughly $2.9B in FY2026.

Market Position & Competitive Advantages

APLD has a strong AI-infrastructure narrative: large contracted lease backlog, investment-grade hyperscaler tenants, purpose-built AI factories, proprietary waterless cooling, and speed advantages from repeatable campus designs. The recent signing of multiple 15-year leases worth roughly $20B is a major positive. However, the company remains early-stage operationally: only 175 MW was live at Polaris Forge 1 as of June 30, 2026, versus 1.41 GW contracted. The business is also heavily capital-intensive, reliant on debt and equity-linked financing, and exposed to customer concentration and execution risk.

Stock Performance

APLD closed near $31.20 on August 16, 2026, up 1.89% on the day. The 52-week range is $13.16 to $50.73, so the stock is up about 90.9% over the past year but trades roughly 38.5% below its high. It is below the 50-day moving average of $34.11 and below the 200-day moving average of $32.43. Average 3-month volume is about 20.5M shares, while the 10-day average is lighter at about 16.3M, suggesting the stock is not under powerful accumulation. Recent price action has been weak: one source noted the stock fell 33% in a month while the finance sector gained 2.3%.

CAN SLIM Analysis

Key Risks

Conclusion

Under CAN SLIM, APLD is not currently a buy. It fails C because quarterly GAAP earnings are negative and worsening; fails A because there is no multi-year record of increasing earnings; and fails S and L because the stock is under distribution and below key moving averages. The N criteria are positive, but new products and leases alone are not enough under O'Neil's system. The stock is not near a proper new-high buy point and is showing lagging relative strength. Existing holders should treat this as a sell or avoid candidate until GAAP profitability, technical repair, and institutional accumulation emerge.

Overview

A Warren Buffett-style intrinsic value analysis of Applied Digital Corporation (APLD) as of August 16, 2026, emphasizing business understandability, economic moat, management quality, financial strength, intrinsic value, and margin of safety.

Business Understanding

Applied Digital is an early-stage developer, builder, and operator of large-scale AI and high-performance computing data centers, with a legacy bitcoin-mining hosting segment and a 96% stake in the separated cloud business ChronoScale. The concept is simple in principle: build power-intensive data centers, lease them under long-term take-or-pay contracts to hyperscalers, and collect rent. But execution is not simple. It requires multi-gigawatt construction, complex project financing, power procurement, and tenant fit-out services that currently dominate reported revenue. This sits at the edge of my circle of competence: the real estate lease model is understandable, but the normalized earnings power, free cash flow, and terminal economics are not yet predictable enough to value with confidence.

Economic Moat Analysis

APLD has early, plausible advantages: proprietary waterless or closed-loop cooling and modular design that management says cuts build times from 24 months to 12-14 months; access to low-cost power and cold-climate sites in the Dakotas; completed onboarding with five of six targeted hyperscalers; and roughly $36 billion of contracted lease value across 1.41 gigawatts. However, these advantages are not yet a wide or durable moat. Most are replicable by larger, better-capitalized data center REITs and hyperscalers. Customer concentration is extreme, and the competitive position depends on flawless execution. The moat is narrow and unproven, with at best a few years of first-mover lead rather than a sustainably protected franchise.

Management Quality

Management has shown aggressive salesmanship and operational wins: delivering the first 100 MW on schedule, signing approximately $36 billion in contracted leases, and lowering project debt costs from 9.25% to 6.75-7%. But shareholder orientation is questionable. GAAP losses are large, stock-based compensation was roughly $220 million in FY2026, share count has grown, no dividend or buyback exists, and complex structures involving preferred units, warrants, SPVs, and non-GAAP exclusions obscure common shareholder economics. Insiders have sold shares, and directors have been net sellers. I would grade management as capable but promotional, not yet a proven owner-oriented allocator.

Financial Strength

Financially, this is a leveraged development story, not a fortress balance sheet. Book value per share is only $5.97 against a $31.20 price, implying a price-to-book ratio of 5.23. TTM EPS is negative $0.91 and forward EPS is negative $0.25. Debt is approximately $5 billion against roughly $1.7 billion in stockholders' equity. Cash is $4.2 billion, but much of it is restricted. FY2026 operating cash flow was about $90 million while capital expenditures were about $2.87 billion, so owner earnings are deeply negative. GAAP net loss attributable to common stockholders was about $250 million. Adjusted EBITDA of $107 million and adjusted net income of $36 million exclude large stock-based compensation and other items. Core data center base-rent NOI margins are high at about 91%, but fit-out services are low margin. This is not a high-return, low-debt business.

Intrinsic Value Assessment

At a current market capitalization of roughly $9 billion, investors are paying for a future that has not arrived. If management hits its roughly $1 billion NOI run-rate target in about a year and that NOI were capitalized at an aggressive 7-8% cap rate, enterprise value would be about $12.5-14.3 billion. After subtracting roughly $5 billion of debt and other claims, the residual equity value is close to today's price, leaving little or no margin of safety. At a more conservative 8-10% cap rate or with execution slippage, value falls materially. There is no GAAP earnings stream to anchor a DCF, and owner earnings are negative. I cannot find a suitable margin of safety at $31.20.

Key Risks

Investment Details