IREN (IREN.US) Q1 FY2027 Earnings Call: CEO Responds to Microsoft Corporation Delivery Delays; Over US$4 Billion ARR Target Postponed to January Next Year

After the U.S. market close on November 5 (Eastern Time), IREN Limited (IREN.US) reported results for the first quarter of fiscal 2027 (ended September 30, 2026) and held a conference call. The most closely watched item in the call was the delivery dynamics of the Microsoft Corporation contract: the company pushed back its target for reaching more than US$4 billion in contracted annualized run-rate revenue (ARR) from the end of 2026 to January 2027, with a possible delay to the end of February 2027 at the latest, because Microsoft Corporation proactively requested changes and upgrades to the network design. Co-CEO Daniel Roberts responded on the call that this was a customer requirement to serve the latest generation of frontier models, that the associated incremental capital expenditure was "completely immaterial," and that the change would not trigger any penalty payments, emphasizing that demand was "stronger than ever." CFO Anthony Lewis added that the AAA-rated cash flows from the Microsoft Corporation contract are supporting the company's data center financing, and that roughly two-thirds of the US$20 billion revenue backlog comes from investment-grade counterparties. On results: according to LSEG data, the company generated revenue of US$228 million in the quarter, up 66% quarter-over-quarter and above the consensus expectation of approximately US$221 million; AI cloud revenue was US$200 million, up 182% quarter-over-quarter, with segment gross margin (before depreciation and amortization) of approximately 86%; bitcoin mining revenue fell to US$28 million. Net loss was US$263 million, sharply narrower than US$684 million in the prior quarter, with a loss per share of 66 cents, wider than the market expectation of a 48-cent loss per share (market expectation data as cited by GMTEight); adjusted EBITDA was US$66 million, up sharply quarter-over-quarter, at a margin of approximately 29%. Operating ARR reached US$1 billion, and revenue backlog reached US$20.3 billion, up 22% quarter-over-quarter. The company disclosed planned capital expenditure of approximately US$15 billion for the remainder of the fiscal year, of which approximately US$6 billion still requires financing. The results were released after the close; the stock fell about 17% during the regular session that day to close at US$18.03 (amid a broad selloff in the AI infrastructure sector and reports that Microsoft Corporation was tightening its review of "neocloud" suppliers), then rebounded by as much as about 10% after hours (sources cited figures ranging from 4% to 11%, with significant after-hours volatility). Focus and response: Microsoft Corporation delivery delay and postponed ARR target The core focus of the call was the delivery timeline for the Microsoft Corporation Horizon deployment and the postponement of the more than US$4 billion ARR milestone. Market sentiment that day was already fragile: Reuters reported that Microsoft Corporation was tightening its review of projects with "neocloud" suppliers including IREN, Nebius and Nscale, and IREN shares fell about 17% during the regular session. Roberts responded that Horizon 1 had been delivered to Microsoft Corporation in August, after which Microsoft Corporation requested reconfiguration and upgrades to part of the network infrastructure so that the 200 MW of capacity could serve the latest generation of frontier models; delivery would be delayed to January and "could absolutely slip to the end of February," depending on supplier pacing. Horizon 2 has been built and is in commissioning, and the GPUs for Horizon 3 and 4 have arrived or are about to arrive. He stressed that the incremental capital expenditure for the network upgrade was "completely immaterial" "we're talking about new cabling, labor and a different implementation approach." The stock rebounded by as much as about 10% after hours, suggesting the market partly accepted the "delayed, not lost" interpretation, but analysts' follow-up questions also highlighted the company's heavy near-term dependence on the delivery pacing of a small number of large customers. AI cloud scales up, mining exits faster AI cloud was the undisputed star of the quarter: revenue of US$200 million, up 182% quarter-over-quarter (approximately US$71 million in the prior quarter), comprising US$135 million of AI cloud services revenue and US$65 million of AI cloud leasing revenue, with segment gross margin of approximately 86%, flat versus the prior quarter. The core driver of the revenue jump was the delivery of Horizon 1 operating IT capacity rose 127% quarter-over-quarter, from 41 MW to 93 MW. Overall gross margin before depreciation and amortization was 80% (76% in the prior quarter), and adjusted EBITDA was US$66 million at a 29% margin (14% in the prior quarter). Bitcoin mining revenue fell to US$28 million (approximately US$67 million in the prior quarter); CFO Lewis said mining hardware is being actively retired, with the goal of substantially completing the transition to AI cloud by the end of this quarter. Land, power and speed of execution: 14 months versus the industry's 18 months Roberts emphasized the company's vertically integrated model: "We don't lease someone else's data center and put the liability on our own balance sheet. We secure power, we develop land, we build infrastructure, we deploy compute." The company has secured more than 30,000 acres of land and more than 5 GW of grid interconnection, plus a development pipeline of more than 35 GW (he cautioned that "not all of it will convert into interconnection agreements"). On speed of execution, Horizon 1 took only about 14 months from civil construction to completing an NVIDIA Corporation Exemplar-certified GB300 cluster, versus the JLL industry benchmark of 18 months (to build a 50 MW data center alone, excluding GPU deployment) "we delivered the building and the cluster four months faster than that benchmark." Pricing on the rise and contract quality: H100s re-contracted at double the price Pricing continues to rise: three-year contracts signed during the quarter carried annualized revenue of US$19 millionUS$23 million per MW (IT load), and five-year contracts approximately US$17 million; five-year contracts signed after quarter-end reached US$22 million/MW, and short-term contracts exceeded US$40 million/MW. Signed multi-year contracts include customer prepayments covering approximately 90% of GPU capital expenditure. Older H100 GPUs were re-contracted on multi-year terms at twice their original price, which Lewis said confirms that "true long-term contract pricing is still doubling." Of the US$20.3 billion backlog, Microsoft Corporation accounts for 47% (US$9.6 billion) and NVIDIA Corporation for 17% (US$3.4 billion), with roughly two-thirds (US$13 billion) from investment-grade counterparties. Capex cut and financing gap Capital expenditure in the quarter was approximately US$5 billion, and planned capital expenditure for the remainder of the fiscal year is approximately US$15 billion, down from prior guidance. Roberts explained: "This is purely a function of the current state of the financial markets; in the current environment capital has to be respected, and we will not grow at any cost." He stressed that the company is not a "neocloud" model: "We don't have US$100 billion total contract value (TCV) contracts with AI labs." On financing, cash and restricted cash at quarter-end was US$5.8 billion (including approximately US$2.3 billion of restricted cash, mainly for GPU deployment financing), total debt was US$10.2 billion (mainly low-interest convertible notes), and net debt was US$4.43 billion; of the US$15 billion of remaining capital expenditure, approximately US$9 billion is already covered, leaving a US$6 billion gap to be financed, with management listing potential funding sources (excluding convertibles and equity) of approximately US$12 billionUS$15 billion. Guidance for the December quarter: cash operating expenses to increase by US$50 millionUS$60 million quarter-over-quarter; approximately one-quarter of the US$1 billion of operating ARR at the end of September will be recognized as revenue in the second quarter. Community relations and site pipeline On concerns about data center community opposition, Roberts said the community concerns the company has encountered are "entirely reasonable" "data centers are the physical manifestation of AI" and that the company responds through local hiring, local procurement and closed-loop cooling systems that draw water from off-site sources; the recently announced 900 MW Redbank site in Australia is an example of community integration. Chief Commercial Officer Kent Draper added that the company considers these issues from day one of site selection, and that sites are located in regional and rural settings rather than cities. On Sweetwater, Sweetwater 1 (1.54 GW) is already interconnected, and both Sweetwater 1 and 2 (2 GW combined) have been selected into the baseload process, with capacity secured. Full conference call transcript Management opening remarks Daniel Roberts (Co-Chief Executive Officer): This quarter was an important quarter of transformation. AI cloud revenue reached US$200 million, up 182% quarter-over-quarter; total revenue was US$228 million, up 66% quarter-over-quarter. Since day one, the company has insisted on securing land and power before demand arrives, which is why the platform today spans three continents. "Most companies do only one of those things; we do all three." Land and power reserves: more than 30,000 acres of land secured, more than 5 GW of grid interconnection secured, and another 35 GW in development not all of which will convert into interconnection agreements. "We don't lease someone else's data center and put the liability on our own balance sheet. We secure power, we develop land, we build infrastructure, we deploy compute." Speed of execution: Horizon 1 took about 14 months from civil construction to completing an NVIDIA Corporation Exemplar-certified GB300 cluster; JLL estimates the industry average to build a 50 MW data center is 18 months (excluding GPU deployment). "We delivered the building and the cluster four months faster than that benchmark." Site progress: Mackenzie has secured 100 MW of power, large-scale site grading has begun, and it will build liquid-cooled data centers of the same design as those already delivered. The recently announced 900 MW Redbank site in Australia is an example of community integration, using closed-loop cooling with off-site water supply. The company is working with NVIDIA Corporation, Mirantis and CoreWeave's AI teams to improve deployment and operations. Anthony Lewis (Chief Financial Officer): On the financial details, AI cloud revenue was US$200 million (+182% quarter-over-quarter), comprising US$135 million of AI cloud services revenue and US$65 million of AI cloud leasing revenue; bitcoin mining revenue was US$28 million and continues to decline, with mining hardware being actively retired "our goal is to substantially complete this transition by the end of this quarter." Overall gross margin before depreciation and amortization was 80% (76% in the prior quarter), and AI cloud segment gross margin was 86% (flat versus the prior quarter). Adjusted EBITDA was US$66 million at a margin of approximately 29% (14% in the prior quarter). Operating expenses (SG&A and R&D combined) were US$224 million, up US$96 million quarter-over-quarter, including US$63 million of stock-based compensation expense; cash operating expenses rose US$33 million quarter-over-quarter (including consolidation of the Mirantis business and headcount growth). Balance sheet (as of September 30): total assets of US$20.6 billion, including US$12.3 billion of property, plant and equipment; US$6 billion of unencumbered data center assets (expected to reach US$10 billion by the end of FY27); cash of US$5.8 billion (including approximately US$2.3 billion of restricted cash, mainly for GPU deployment financing); total debt of US$10.2 billion, mainly low-interest convertible notes maturing in 2032 and beyond; approximately US$19 billion of total financing over the past 12 months (including US$6.5 billion of GPU financing since May). Cash plus undrawn commitments was approximately US$9 billion. Contract pricing: three-year contracts signed during the quarter at US$19 millionUS$23 million/MW (annualized), and five-year contracts at approximately US$17 million/MW; after quarter-end, five-year contracts at US$22 million/MW and short-term contracts above US$40 million/MW. New five-year contracts include customer prepayments covering approximately 90% of GPU capital expenditure. Older H100s were re-contracted on multi-year terms at twice their original price. The AAA rating of the Microsoft Corporation contract cash flows supports data center financing; roughly two-thirds of the US$20.3 billion backlog comes from investment-grade counterparties such as Microsoft Corporation and NVIDIA Corporation. Funding sources: US$10 billion of unencumbered data center assets by the end of FY27 + US$6 billion of unencumbered GPUs + multi-year contract cash flows, indicating a scale of US$12 billionUS$15 billion. Near-term financing priorities: financing for the Horizons 14 data centers serving the Microsoft Corporation contract, and financing for the Childress air-cooled GPUs backed by NVIDIA Corporation and Leading Lab contracts. Q&A session Paul Golding (Macquarie Research): Could you talk about the delivery dynamics of the Microsoft Corporation capacity, and the relationship between supplier delays and change orders? Daniel Roberts: Horizon 1 was delivered to Microsoft Corporation in August, after which Microsoft Corporation requested "reconfiguration and upgrades to part of the network infrastructure" so that the 200 MW of capacity could serve the latest generation of frontier models. Delivery will be delayed to January "could absolutely slip to the end of February," depending on suppliers. Horizon 2 has been built, racks are powered, GPUs are installed, fiber is laid, and Dell Technologies, Inc. Class C is commissioning; for Horizon 3 all GPUs have arrived and all three data halls are powered; the GPUs for Horizon 4 will arrive within weeks. "This is a very good partnership... we're saying 'yes' to the customer here, not 'no.'" Follow-up (impact on incremental capital expenditure): the incremental capital expenditure for the network upgrade is "completely immaterial" "we're talking about new cabling, labor and a different implementation approach." The relevant experience will be reused at other sites under construction. Brett Knobloch (Cantor Fitzgerald): Are you considering modular data centers (such as NVIDIA Corporation DSX)? Contract terms and the preferences of investment-grade customers? Daniel Roberts: The company has a dedicated internal team (established about nine months ago) studying construction methods, focused on "modularity and the ability to plug more components in on site," and is discussing with NVIDIA Corporation closely how to implement the DSX modular design. On the customer mix: "The answer is both; they are not mutually exclusive." Two-thirds of the US$20.3 billion backlog is investment-grade, backed by "US$4 trillion-class credit support"; the lower-credit portion typically comes with higher prepayments. Joseph Vafi (Canaccord Genuity): What factors determine whether development accelerates or slows? Daniel Roberts: The company is eight years old, focused on a long-term vision, and builds assets layer by layer land and power are fully funded with cash (US$6 billion, rising to US$10 billion by June 30), and with unencumbered debt; "we are not a WeWork-style business model." Demand is "stronger than ever." This week's signings show: customers agreeing to prepay 90% of GPU capital expenditure, short-term contracts at US$40 million/MW, frontier AI lab expansion, and H100s re-contracted at double the price. "The dislocation in the current financial markets is very significant... we can afford to be patient." Greg Lewis (BTIG): FY27 capital expenditure has been cut consecutively why? Daniel Roberts: "This is purely a function of the current state of the financial markets; in the current environment capital has to be respected, and we will not grow at any cost." The company stressed it is not a neocloud model: "We don't have US$100 billion TCV contracts with AI labs." Anthony Lewis (added): The H100s re-contracted at double the price are multi-year long-term contracts, "true long-term contracting... and the result is still a doubling of price." Michael Donovan (Compass Point): With financing constrained, how do you prioritize sites such as Oklahoma and Spain? How much does NVIDIA Corporation software improve efficiency? Daniel Roberts: The "cheap option" philosophy from land options to the capital required to start construction, the requirement is inversely proportional to time-to-compute (the time to reach compute output); six sites are currently under construction; capital sources include private markets, real assets and infrastructure funds. "This dislocation is unlikely to last long"; even if the stock market does not recover, the company can continue to monetize 5 GW. Anthony Lewis: NVIDIA Corporation now treats the entire data center as a single computing unit; the company uses Max LPS software to oversubscribe GPUs, automatically reducing clocks in extreme cases so as not to exceed rated power, thereby deploying more GPUs per MW; NVIDIA Corporation is also researching higher operating temperatures to reduce cooling power draw and free up more power for IT compute. Daniel Roberts (added): Owning grid interconnection and its own data centers means the company can optimize beyond pure software limits; he reiterated Horizon 1's 14 months versus the industry benchmark of 18 months. Mike Colonise (HC Wainwright): Of the 500 MW of incremental data center capacity in 2027, how much has been contracted? Daniel Roberts: He did not give a specific breakdown, but "NVIDIA Corporation is part of that capacity," along with frontier AI labs; Childress air-cooled will advance in 2027, and Vera Rubin (VR) chips will arrive at year-end "VR supply will be extremely scarce next year, and we are seeing a lot of interest." Most of the recent signing activity has been on Childress air-cooled capacity. Michael Ning (Goldman Sachs Group, Inc. Research): Why cut capital expenditure but keep the end-2027 capacity target unchanged? If community opposition causes delays, can customers terminate leases or reclaim GPU prepayments? Daniel Roberts: The capital expenditure cut reflects the pace of investment across the 35 GW development portfolio; "we didn't come here for this 200 MW of Horizon 1, or even for this 5 GW of secured power"; to find large-scale power from scratch now, "you would very easily be into the 2030s," so the company keeps advancing the 35 GW pipeline. The company has not provided 2028 guidance, but capacity will be available by then. On community opposition: the community concerns the company has encountered are "entirely reasonable" "data centers are the physical manifestation of AI"; the company hires locally, uses local suppliers, and uses closed-loop cooling with off-site water supply. Kent Draper (Chief Commercial Officer) (added): These issues are considered from day one of site selection sites are in regional and rural settings rather than cities, bringing local jobs, community funding programs, community consultation and education. Jonathan Peterson (Jefferies Financial Group Inc. Research): What is the interconnection progress at Sweetwater? What are the trends in GPU procurement costs? Kent Draper: Sweetwater 1 (1.54 GW) is already interconnected; both Sweetwater 1 and 2 (2 GW combined) have been selected into the Batch 0/baseload process, with capacity secured; other projects will only be announced once binding interconnection agreements are in place; both projects are under construction, with permits obtained as construction progresses. Anthony Lewis: Industry supply remains tight (especially for SSDs and memory), and GPU prices are expected to keep rising; price increases are passed through immediately to contract hourly pricing, and "over the past six months we have actually seen payback periods shorten." Daniel Roberts (added): The company deliberately does not buy all of next year's GPUs in advance "we view this as right-way risk": if the market weakens and the industry struggles to execute, GPU prices may fall; if the market recovers, the company pays at most a small premium or waits two or three months. The core is time-to-compute, prioritizing physical infrastructure and preserving flexibility. Closing remarks Daniel Roberts: "Horizon 1 has been delivered to Microsoft Corporation with NVIDIA Corporation Exemplar certification status 14 months from civil construction, versus the industry benchmark of 18 months and Horizons 2 through 4 are now nearing completion. More than US$4 billion of contracted ARR is funded, and the target is now set for January, perhaps slipping to the end of February in this quarter... AI cloud revenue nearly tripled, and our operating ARR reached US$1 billion. Our backlog exceeds US$20 billion, two-thirds of it from investment-grade counterparties. Demand... frankly has never been stronger... we own the land, the power and all the data centers beneath them, which is exactly why we can control our own pace. Our next milestone is the 0.3 gigawatts we are targeting for January."