IREN's low-margin bare-metal business is being financed on uncontracted future capacity, so rising opex and added financing will pressure it. A Goldman Sachs report now flags exactly that: extra financing supports a material amount of uncontracted capacity, the Secured Power metric IREN touts. Most AI data center leasing models will be pressured in some way, and IREN is the most obvious example. x.com 

IREN's spending and ongoing financing still weigh on its shares despite a signed Anthropic deal. Its 2027 capex is put at $25-30B and its $6B ATM appears roughly halfway complete. Mirantis has improved fundamentals, but markets likely care more in about six months as revenue approaches and the ATM/capex window passes, as with Nebius after its Meta deal. A risk-off rotation to larger names adds pressure. x.com 

IREN is worth a small position because customer cash funds most of its buildout and its signed contracts are priced like CoreWeave's. Prepayments covered 87% of June-quarter capex, versus 21% at Nebius and 12% at CoreWeave, and ~$14B of cash, financing and prepayments covers its $13.8B in June commitments. EV is 0.88x contract value versus 0.74x for CoreWeave. The main risk is running GPU clouds worse than peers, eased as bulk buyers manage GPUs themselves. $11-16B of capex is unfunded. x.com 

IREN's Microsoft buildout is close to the contract's size even if everything goes well, leaving it essentially trying to break even while executives collect generous pay and perks. Perhaps only four or five customers worldwide can afford or need to prepay a multibillion-dollar bare-metal deal, so talk of diversification rings hollow. This is a classic pump and dump. x.com 

IREN's lack of deal announcements is a red flag when gigawatts of power are worth at least twice what they were recently. Per SemiAnalysis, inference revenue can reach $100B per GW per year at 90%+ gross margins, so a roughly $5B power plant pays for itself in about 20 days. x.com 

IREN's uncommitted future capacity lets it seek bigger customer prepayments if credit tightens. CoreWeave signed many contracts when funding was easier and, per its management, deprioritized prepayments in early 2026, leaving more capex to fund externally. IREN is nearly certain to get the Rubin GPUs Sweetwater needs: it is about a year away and was designed with Nvidia as a flagship Rubin deployment. SW1's use of DSX pods is less clear, though months of joint design tilt the answer toward yes. … Show more

IREN's AI data-center expansion needs enormous capital, making financing needs and execution delays material risks. Bitcoin exposure adds another source of uncertainty despite the size of the AI opportunity; the stock looks tempting but is not one to buy here. x.com 

CoreWeave's risk is buildout speed: each new build needs capital before the last pays for itself. Its analysis shows $35.6B of loans lasting about nine months beyond customer contracts; $14.6B is due in 2031-32 as final contract cash arrives. IREN's Microsoft project loan matches its contract, but $6.3B of convertible bonds mature in 2031-33, mostly after contracted cash ends; low shares mean cash repayment. Slower AI demand growth could let both earn back projects before funding more. x.com 

IREN is a buy and hold because the reaction to its capex plan is overblown. Its projected $25-30B of FY27 capex is mostly funded by customer prepayments and GPU financing, per HyperTechInvest: $14B is already secured and another $8B is expected the same way, leaving $3-8B from cash flow, data center financing and equity. Negative sentiment around data centers is driving hasty selling, making now a good time to position for the future. x.com  x.com 

IREN's worsening fundamentals warrant patience despite its potential. Trailing free cash flow is -$2.45 billion, net income -$703 million and EPS -$1.64 even as revenue has grown; it needs a catalyst change and a positive trend in fundamentals. x.com