LEADER TRADING
Stock Trader, AI & Data-Center Equities, Host of DDUI
IREN lacks a near-term catalyst strong enough to drive a real move to new highs, yet this is a level to buy, not sell. IREN's Dan Roberts suggested deals should not be expected soon, and anything short of a deal likely brings only a small pop and fade. The risk is subpar November earnings and no deal before February results, since the chart cannot afford a five-month fade. x.com x.com x.com x.comÂ
IREN looks horrendous on the chart, but institutions are not selling, setting up a big pop to $40+ once everyone gives up. The declines have come on a complete lack of volume. Over the first half of 2026 it saw $250B of high-volume activity and twice failed to make new lows in this area, a textbook long-term accumulation campaign; key institutional supports above still need to be reclaimed to confirm a short-term bullish trend. x.com x.comÂ
Open-source models likely wrecked the heavy-capex thesis behind IREN. Debt CDS and falling token prices make the still-bullish view definitely less so. Unlike Roku, a previous long-term hold with a unique, massive user base, low capex and positive free cash flow, IREN lacks a clear path forward; without one, it is better not to hold it. If Warsh remains hawkish, a full exit from IREN is very likely. x.com x.comÂ
IREN's predatory management and markedly weaker execution than Nebius make it hard to take seriously while Mike Alfred remains on its board. Barring an oversold bounce, IREN looks cooked big time with more time. x.com x.comÂ
IREN's deal economics, not the Anthropic name alone, will determine whether locked-in demand earns good returns. The reported deal is good news, but its terms remain unknown, and a buyer that large can negotiate hard on price. IREN should do well: it still has more capacity to sell and is expected to charge a premium for it. x.comÂ
IREN's long-term appeal rests on its standing as one of the best vertically integrated AI neoclouds. In the short term, the stock has not been this oversold since 2022. x.comÂ
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.comÂ
IREN and Hut 8 look like buying opportunities because they just locked in multi-billion-dollar contracts while the market keeps pushing them lower. That selloff is a gift, not a red flag. The AI infrastructure trade rests on a power shortage that has persisted for over a year, and more energy is needed. x.comÂ
IREN's newly signed multiyear deal with Anthropic should drive a big pop in the stock within 15 minutes. x.comÂ
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.comÂ
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.comÂ
