X 讨论IREN sold off about 13% after FY26 results even as AI Cloud revenue doubled to $70.5M and overtook Bitcoin mining for the first time. Traders are fighting over whether the $4B contracted ARR and customer-funded GPU buildout justify buying the dip, or whether a $703M net loss, mining impairments, and $25–30B FY27 capex plus ongoing dilution invalidate the AI pivot thesis.
多头The selloff ignores a real business inflection: AI Cloud has already doubled quarter on quarter and passed mining revenue, with $4B of contracted ARR and 2026 capacity largely sold out. Dip buyers are adding size near $35 and treating the loss as mostly a non-cash write-down of old mining gear as Horizon capacity ramps into revenue later this year.
空头The capital plan does not pencil: James Chanos notes roughly $25–30B of FY27 capex against a $4B ARR forecast leaves returns unworkable. Share count already jumped from 283M to 394M, so the same $35 print buys a ~$13.8B company instead of a sub-$10B one, and the $703M FY26 loss plus serial ATM funding keep dilution as the dominant risk.
帖子
多头论据
@midascabal13% drop is an overreaction after AI Cloud doubled to $70.5M and topped mining; added $500k at $35
@growthrapidlyAfter-hours dump makes no sense after AI Cloud more than doubled QoQ and beat mining revenue for the first time
@moninvestorWants a low-$30s or high-$20s entry; thesis hangs on hitting the $4B+ 2026 ARR target as Horizon 2–4 deliver
空头论据
@RealJimChanos$25–30B FY27 capex on a $4B ARR forecast means the capital math still does not work
@babyfolioSame $35 price is not cheap: shares outstanding rose from 283M to 394M, lifting market cap to ~$13.8B from under $10B
@joinautopilot~$639M mining-equipment impairment drove a $702.6M FY26 net loss and roughly $1.85B of market-cap wipeout in a day