$APLD EXECUTIVE CALL SUMMARY: Applied Digital Corp (01/07/26)
Fiscal Q2 2026 was positioned by management as an operational and financial inflection, primarily due to Polaris Forge 1 reaching “ready for service” and the initial recognition of recurring hyperscale lease revenue, shifting the narrative from “construction and contracting” toward “operations and monetization” of the AI factory portfolio.
Reported revenue of $126.6 million increased 250% year-over-year and increased 97% sequentially, driven predominantly by $73.0 million of tenant fit-out services (explicitly characterized previously as 1-time and low-margin) and $12.0 million of partial-quarter lease revenue related to the CoreWeave commencement at Polaris Forge 1, supplemented by $41.6 million of legacy Data Center Hosting revenue.
Earnings quality was mixed: adjusted EBITDA of $20.2 million contrasted with a net loss attributable to common stockholders of $31.2 million, with material noise from accelerated stock-based compensation, step-function higher interest expense, and fair value marks tied to the Babcock & Wilcox position/warrant (items not directly linked to core site-level NOI).
Lease accounting under ASC 842 was highlighted as a near-term modeling complication, with management explicitly stating that lease revenue will be straight-lined over 15 years, causing recognized revenue to differ from cash receipts in the early ramp period; disclosed cash lease revenue was approximately $8 million versus $12 million recognized. This reduces the information content of GAAP revenue in early quarters and increases the importance of cash collections, site-level NOI, and liquidity runway.
The liquidity profile was positioned as a competitive advantage during construction: quarter-end cash, cash equivalents, and restricted cash totaled approximately $2.3 billion against approximately $2.6 billion of debt (majority not maturing until 2030), with additional proceeds subsequent to quarter end. The call emphasized “repeatable” financing with Macquarie facilities and the recently completed $2.35 billion 9.25% senior secured notes due 2030.
Forward-looking commentary was meaningfully more aggressive than prior-quarter framing: management stated an expectation to exceed (not merely reach) the $1 billion NOI target within 5 years, anchored by 600 MW already leased across 2 North Dakota campuses and incremental leases assumed from “advanced discussions.”
Commercial tone was constructive: pricing was described as stable to slightly improving, with greater improvement cited in contractual protections (cancellation economics and transfer restrictions), implying a higher-quality cash flow stream and improved underwriting versus earlier-cycle deals.
Pipeline disclosure increased in specificity: “advanced discussions” were quantified as 3 sites totaling 900 MW, with management repeatedly emphasizing that demand is not the constraint; execution capacity, supply chain, and power availability were framed as the binding constraints.
Power strategy and differentiation were elevated as core to competitive positioning, including the Babcock & Wilcox steam-based natural gas generation pathway intended to accelerate time-to-power versus conventional turbine lead times, with B&W publicly indicating a targeted start of operations in 2028 and an expected full contract release in Q1 2026.
Strategically, the proposed spin-out/combination of Applied Digital Cloud with EKSO to form ChronoScale was positioned as a capital allocation and focus unlock, with Applied Digital expected to own over 80% at close and with management asserting that access to Applied Digital’s data center capacity is a key advantage for the compute platform.
KEY DEVELOPMENTS AND DELTA VERSUS PRIOR GUIDANCE
Several milestones discussed on the call represented execution against explicitly stated prior-quarter operational expectations. In fiscal Q1 2026 disclosures, the 100 MW facility at Polaris Forge 1 was stated to be on track to be operational in calendar Q4 2025, with tenant fit-out revenue expected to “ramp significantly next quarter” and lease revenues expected to “ramp later this year” as installation completed. Fiscal Q2 2026 delivered the anticipated step-up in fit-out activity ($73.0 million versus $26.3 million in fiscal Q1 2026) and included the first partial-quarter lease revenue recognition ($12.0 million) as the site reached ready-for-service.
The prior-quarter disclosure also described Polaris Forge 2 as being in “advanced discussions with an investment-grade hyperscaler” with aligned key terms, with the “when signed” framing explicitly tied to bringing leased capacity to 600 MW across 2 campuses. Fiscal Q2 2026 reflected a transition from “advanced discussions” to signed commitment, with the company reiterating that 600 MW are now under long-term lease (400 MW CoreWeave at Polaris Forge 1 plus 200 MW investment-grade hyperscaler at Polaris Forge 2).
Long-term NOI guidance also shifted at the margin but in an investment-relevant way. Prior-quarter disclosures framed a projected annualized NOI run rate of approximately $500 million once Polaris Forge 1 is fully operational and stated that the tenant signing at the 2nd campus should put the company “firmly on the path” toward the $1 billion NOI target within 5 years. The fiscal Q2 2026 release and call tightened this to an expectation to “exceed” the $1 billion NOI target within 5 years, implicitly assuming additional lease wins beyond the already executed 600 MW.
Incremental new information on the call that was not previously quantified included: 1) an explicit count and size of sites in advanced discussions (3 sites; 900 MW), 2) stronger commentary on improving non-price contract terms (make-whole economics, transfer restrictions), and 3) an operational framing that the business is now supply/execution constrained rather than demand constrained, with internal focus on annual build capacity, MEP supply, and power availability.
FINANCIAL PERFORMANCE AND QUALITY OF EARNINGS
Top-line growth was exceptional on headline metrics but heavily mix-driven, with a meaningful portion attributable to activities that are non-recurring and/or accounting-driven timing items. Revenue of $126.6 million increased 250% year-over-year versus $36.2 million and increased 97% sequentially versus $64.2 million in fiscal Q1 2026. The primary incremental driver was tenant fit-out services tied to hyperscale deployments, recognized as revenue as work is performed, with a corresponding pass-through cost profile.
The quarter’s revenue composition implied 3 distinct economic buckets with very different margin structures and durability: 1) tenant fit-out services ($73.0 million), 2) recurring lease revenue (recognized $12.0 million; cash approximately $8.0 million), and 3) legacy Data Center Hosting revenue ($41.6 million). Management explicitly stated that the difference between recognized and cash lease revenue was driven by straight-line rent recognition under ASC 842 over a 15-year term, indicating that early quarters may show GAAP lease revenue above cash receipts if lease payments are structured with escalators. This dynamic increases the risk of mis-modeling near-term cash generation if GAAP revenue is used as a proxy for cash NOI.
Gross profitability at the consolidated level appeared muted due to the high share of low-margin fit-out work. Total cost of revenues was $100.6 million on $126.6 million of revenue, implying gross margin of approximately 20.5%. However, the company disclosed approximately $69.5 million of cost directly associated with fit-out services, suggesting fit-out gross margin of roughly 4.8% (approximately $3.5 million of implied gross profit on $73.0 million revenue). Ex-fit-out, implied gross margin was materially higher, but that calculation remains approximate because the remainder of cost of revenue includes energy and other direct costs across businesses. The key implication is that quarter-to-quarter gross margin volatility is likely to remain high as fit-out revenue ramps up and down based on tenant deployment schedules rather than steady-state lease economics.
The legacy Data Center Hosting segment remained a meaningful cash and profitability contributor. Revenue of $41.6 million increased 15% year-over-year (and increased approximately 10% sequentially versus fiscal Q1 2026’s $37.9 million), with management reporting approximately $16.0 million of segment operating profit on an approximately $130.8 million reported asset base. That implies segment operating margin of approximately 38% in the quarter and a very high implied annualized return on segment assets, albeit in a business whose economics can remain sensitive to customer utilization and broader crypto-mining conditions.
Operating expense trends were a critical offset to the improving gross profit trajectory. SG&A increased to $57.0 million from $26.0 million in the prior-year period, driven primarily by $23.8 million of stock-based compensation tied to accelerated vesting and elevated legal/professional services associated with financing and growth initiatives. While the accelerated vesting component is plausibly non-recurring, the underlying SG&A run-rate remains elevated relative to the current recurring lease revenue base, implying that corporate overhead absorption is dependent on rapid scaling of lease revenue over the next several quarters.
Interest expense increased to $11.5 million from $2.9 million year-over-year, consistent with the step-up in leverage and project financing. Importantly, interest expense in early ramp periods can remain partially capitalized during construction; as assets move into service, the capitalized portion can decline and P&L interest expense can rise even if absolute debt stays flat. The call’s emphasis on refinancing to lower rates once buildings are operational highlights a material sensitivity of equity value to the terminal cost of debt and to capital markets receptivity at the refinancing window.
Adjusted EBITDA of $20.2 million represented a sharp improvement versus fiscal Q1 2026 ($0.5 million) and likely reflects a combination of stronger Data Center Hosting performance, incremental gross profit from fit-out services, and the initial lease revenue recognition. The quality of this EBITDA remains partially transitional because recurring lease revenue remains in early ramp and fit-out activity is inherently episodic.
LEASE REVENUE, ACCOUNTING, AND CASH FLOW TRANSLATION
The quarter introduced a critical new modeling regime: the company moved from “pre-revenue” in its hyperscale leases to an initial lease commencement, which materially changes how performance should be measured. A key management quote was: “On a cash basis for the leases, revenues were approximately $8 million. The difference between cash received and the revenue recognized reflects ASC 842 lease accounting, which requires lease revenue to be recognized on a straight-line basis over 15 years.” This is a direct warning that GAAP revenue is not a clean indicator of lease cash yield in early years.
Investment implications of this accounting dynamic are significant. In the early lease years, straight-line rent recognition can pull forward revenue relative to cash, which can inflate EBITDA margins and understate the time required for lease cash collections to cover corporate overhead and interest. Conversely, later in the term, cash receipts can exceed recognized revenue. For underwriting, recurring cash NOI (and ultimately cash available for refinancing and deleveraging) is the more relevant metric than GAAP lease revenue.
Cash flow data further reinforced the capital intensity of the model. For the 6 months ended November 30, 2025, net cash used in investing activities was $801.5 million, driven by continued construction, while financing activities provided $3.1 billion, resulting in a net increase in cash. This confirms that near-term enterprise value is still primarily determined by the ability to fund construction efficiently and convert build progress into contracted cash flows rather than by current operating cash generation.
COMMERCIAL ENVIRONMENT, PRICING, AND CONTRACT STRUCTURE
Management’s commentary suggested that the lease environment remains favorable and potentially improving on a risk-adjusted basis. While “headline pricing” was stated to be “stable to slightly better” over the last several months, the more differentiated claim was improved contract quality, including cancellation and transferability protections. A key quote was: “Our contracts are really non-cancellable for 15 years.” Management further stated: “The customer can cancel for convenience. However, they owe us the 15 years of payments if they do.” This “make-whole” structure, if enforceable and properly secured, meaningfully improves downside protection versus more cancellable colocation structures, but still leaves residual exposure to counterparty solvency and to any negotiated remedies/limitations embedded in the definitive documentation.
The call also highlighted an underwriting distinction between nominal lease pricing and “calculated yield,” with management acknowledging that the effective economics depend on build cost assumptions and contractual features beyond rent. The practical investment takeaway is that assessing the durability of the 15-year contracted revenue stream requires focus on: 1) the precise scope of what is included in “prospective lease revenue” (rent versus power pass-through), 2) escalation structures that drive the ASC 842 cash-versus-revenue divergence, and 3) remedies and collateral package supporting the make-whole.
Demand was framed as abundant and broadening. Management stated inbound demand increased “meaningfully” after signing 2 hyperscale leases in North Dakota and disclosed “advanced discussions with another investment-grade hyperscaler across multiple regions,” including additional Dakotas locations and select Southern U.S. markets. The disclosure that the company is “qualified” with 5 of the 6 targeted hyperscaler counterparties (the 5 investment-grade hyperscalers plus CoreWeave) implies a shortened contracting cycle for incremental sites because the onboarding/master agreement process has largely been completed, which management characterized as taking 3 months to as long as 1 year for new relationships.
BUILD-OUT EXECUTION, TIMELINE, AND SCALING CONSTRAINTS
Operational execution risk was a central theme, with the call attempting to de-risk the build narrative. The CEO stated: “Polaris Forge 1 reached ready for service, energizing 100 megawatts on schedule.” This milestone is fundamental because it transitions the CoreWeave relationship from development to operations, reduces the probability of construction delays on the base-case schedule, and supports credibility for simultaneous multi-campus builds.
The contracted North Dakota footprint discussed on the call can be summarized as follows, based on management statements: 1) Polaris Forge 1 is contracted for 400 MW for CoreWeave with campus completion expected by end of 2027, and 2) Polaris Forge 2 is contracted for 200 MW with initial capacity expected in 2026 and full build-out in 2027. In aggregate, management cited 600 MW of leased capacity representing approximately $16 billion in prospective lease revenue across the 2 campuses.
The PF2 lease terms disclosed outside the call, but directly relevant to expansion commentary on the call, included an ROFR for an additional 800 MW of critical IT load, representing the full expansion potential of a 1 GW campus, with projected PUE of 1.18 and near-zero water consumption. This aligns with the call’s repeated emphasis that each campus can scale to at least 1 GW and supports the strategic rationale of landing an anchor tenant with a contractual pathway for multi-phased expansion.
Execution scalability was framed as the primary limiter, not demand. The CEO stated: “I don’t think demand is going to be the limiter for us,” and described internal focus on how many sites can be built simultaneously, with attention to supply chain, personnel, and weather risk. The call emphasized modular design and prefabrication: multiple concrete plants, prefabricated components delivered via trucking, and flexible building footprints capable of supporting different GPU/ASIC architectures and networking. This is consistent with an attempt to industrialize deployment and reduce the cost and cycle-time variance that typically undermines first-time hyperscale builders.
The most explicit new pipeline disclosure was: advanced discussions on 3 sites totaling 900 MW. Management cautioned on certainty (“Nothing is done until it’s done”), implying a non-trivial probability of slippage or fall-through, but the disclosure is still investment-relevant because it indicates a potential second wave of contracting beyond the already committed 600 MW.
If the 3-site/900 MW pipeline is converted, management asserted that each of the additional campuses could scale to 2 GW, and that a path exists to approximately 5 GW of capacity over the next 5 years without adding more campuses, implying significant embedded optionality. This optionality is highly valuable in a power-constrained environment but introduces execution risk because the limiting factor becomes power procurement and interconnection timelines rather than land or customer demand.
CAPITAL STRUCTURE, COST OF CAPITAL, AND FINANCING FRAMEWORK
The call placed heavy emphasis on a “repeatable” and “capital-efficient” financing framework designed to minimize reliance on public equity issuance while retaining majority ownership. The CFO stated: “We have agreements in place with top-tier financial institutions that allow us to execute this repeatable and capital-efficient framework.” The described sequencing was: 1) a development loan facility with Macquarie Equipment Capital to fund pre-lease work and early construction, and 2) a preferred equity facility with Macquarie Asset Management (MAM) accessed after a mutually agreed executed lease with an investment-grade hyperscaler, supplemented by project-level debt.
The senior secured notes were a major quarter event. The company completed a $2.35 billion private offering of 9.25% senior secured notes due 2030, issued at 97% of par, with proceeds allocated to construction of the 100 MW and 150 MW facilities at Polaris Forge 1, repayment of the SMBC loan, and required debt service reserves.
The MAM preferred equity facility is economically material to equity value and was a key enabling factor for scaling. Publicly disclosed facility terms include: a 12.75% annual dividend (payable in kind or cash at the company’s election), step-ups if still outstanding after the 5th and 6th anniversaries, a minimum 1.80x multiple of invested capital liquidation preference (inclusive of the value of the common equity issued to MAM), and common equity representing 15% of the HPC subsidiary’s fully diluted common equity at issuance. The facility structure also includes a right of first refusal for up to an additional $4.1 billion across the future HPC pipeline and sets minimum equity contribution requirements of $1.0 million per MW for the rest of the Ellendale campus and $0.75 million per MW for the remainder of the pipeline.
These terms imply that while the framework can reduce the need for public equity, it embeds a high implied cost of capital at the project level and introduces a meaningful preferred claim and non-controlling interest that will sit structurally senior to common equity in the cash flow waterfall. The call’s stated intention to refinance to lower rates after stabilization is therefore a critical equity lever, because the spread between a construction-phase cost of capital and a stabilized-phase cost of capital will determine the terminal free cash flow available to common equity and the viability of a REIT-like multiple over time.
Management explicitly addressed this dynamic: “Project-level debt typically carries higher interest rates initially as it finances the riskier portion of development. But once the buildings are operational, our goal is to refinance at lower rates.” This statement should be interpreted as both a plan and a risk factor, given dependence on capital markets conditions and on demonstrated operating performance at the time of refinancing.
Liquidity at quarter end was described as a strategic advantage. The company ended the quarter with $2.3 billion in cash, cash equivalents, and restricted cash versus $2.6 billion in debt, with additional proceeds raised after quarter end. The investment committee implication is that near-term dilution risk may be reduced by the current cash balance and committed facilities, but not eliminated, because scaling beyond the contracted 600 MW plus the 900 MW pipeline could still require substantial incremental equity, preferred, and/or project debt depending on build cost per MW and the cadence of customer deployments.
POWER AVAILABILITY, TIME-TO-POWER, AND TECHNICAL DIFFERENTIATION
The CEO positioned the Dakotas as a durable structural advantage due to low-cost energy, cool climate, and ample land. The call framed global AI infrastructure as constrained by power availability and by the long lead times for new generation, including commentary that many buyers are being pushed toward 30-year power plant developments. This framing supports the strategic rationale for siting and for the pursuit of alternative generation pathways.
The Babcock & Wilcox partnership was discussed as an attempt to shorten time-to-power relative to traditional turbine supply chains. Management stated that ordering a traditional natural gas turbine “today” could imply deliveries in 2031-2032 and emphasized the need for earlier power. B&W has publicly stated it signed a limited notice to proceed with Applied Digital for a project valued at over $1.5 billion to deliver 1 GW of power, expects a full contract release in Q1 2026, plans to design and install 4 300 MW natural gas-fired plants, and is targeting initial operations in 2028. The mismatch between “earlier power” narrative and a 2028 target date is still directionally favorable versus 2031-2032 but remains a medium-term timeline; therefore, it should not be underwritten as a near-term catalyst for 2026-2027 lease ramp but rather as an enabler for 2028+ campus expansions and for multi-GW optionality.
Cooling technology and future-proofing were also elevated. The company disclosed a $15 million investment in Corintis and provided a technically detailed description of Corintis’s microchannel cold plate approach, emphasizing chip-level heat mapping and the potential to maintain cooling effectiveness as power density increases across chip generations. The investment case relevance is 2-fold: 1) improved power density support can raise revenue per square foot and per MW by enabling higher-density deployments, and 2) superior cooling and PUE can improve customer TCO and increase competitiveness in a market where power is increasingly the marginal constraint.
STRATEGIC: APPLIED DIGITAL CLOUD SPIN-OUT AND CHRONOSCALE
The call provided incremental clarity on the timeline and process for the proposed ChronoScale transaction. Management stated an expectation for a definitive agreement later in January or early February 2026, followed by a shareholder vote, and suggested an April-May 2026 closing window (with March discussed as an early-case possibility). The transaction was described as a non-binding LOI and remains subject to customary approvals and closing conditions.
Strategically, the call framed the compute market as having “a big opportunity,” citing recent deal activity and indicating that Applied Digital has been involved in discussions with counterparties. Management described ChronoScale’s potential advantage as its relationship with Applied Digital and access to large-scale data center facilities for accelerated compute deployments (GPUs/TPUs/LPUs). This is directionally supportive of a “stack” strategy, but management also avoided specificity on how future GPU purchases would be allocated between Applied Digital and ChronoScale, implying uncertainty on intra-company commercial arrangements and on potential related-party dynamics.
The call cited the cloud business as generating over $60 million in trailing 12-month revenue with $313 million in assets and stated Applied Digital is expected to own over 80% of ChronoScale upon closing. The magnitude of ownership and the future capital needs of ChronoScale will determine whether this becomes a value-unlock and focused growth vehicle or a future dilution vector at the subsidiary level.
