IREN's 7% jump isn't hype — conditional ERCOT status for a 2GW hub means the company can actually secure the electricity that AI data centers desperately need, converting it from a bitcoin miner into a power-and-datacenter landlord. The CNBC report showin

Verdict: the ERCOT story is real, but only half the pair is entry-ready — wait for confirmation

The thesis has a real kernel: IREN's conditional ERCOT status for its 2GW Sweetwater hub (per the September 8, 2026 Yahoo Finance report) secures the scarcest input in AI data-center buildout, and the backtest over 24 months shows the momentum-with-stop rules returned 79.0% on 66 trades despite a 48.5% win rate and a 28.7% maximum drawdown. But the freshest year of that same backtest was essentially flat (-0.13% across 45 trades with an 18.4% drawdown), the 60-month evaluation could not run, and no robust parameter setup was established after the sensitivity analysis failed to complete. On fundamentals, the last quarter was ugly for both legs: IREN's revenue fell 21.6% to $144.8M with a -$247.8M net loss and -$1.28B free cash flow, while APLD's net loss widened to -$99.3M with -$720.2M free cash flow and shares outstanding up another 2.3%. The strongest point for the trade is that the APLD entry condition is live today — close of $28.31 below its 50-day EMA at $29.84 with RSI at 59.3 — with a roughly 4-to-1 reward-to-risk profile (2.4% stop near $27.64, +10% target near $31.14). The strongest point against is that with a 0.5707 pair correlation, both unprofitable, dividend-free names burning cash are one concentrated bet, and a 2.4% stop is routinely whipsawed in this pair. The verdict flips if IREN cools to the stated entry band (RSI at or below 60 and a close below $42.40) while the next quarterly prints show free-cash-flow burn narrowing rather than compounding; ownership filings covering June 30, 2026 are past their deadlines and show only 3 (APLD) and 5 (IREN) reported holders, so there is little institutional ballast to lean on either way.

Trade now: APLD is live, IREN is a level away

**Actionable today: APLD.** Every entry condition is currently met. APLD closed at $28.31, below its 50-day EMA at $29.84 (price sits about $1.53 under that line), with RSI (14) at 59.3 — inside the required band (above 40, below 60). That makes the RSI/EMA long entry on APLD live as of this session. **Position mechanics if you enter APLD here:** the hard stop-loss fires at a 2.4% loss on the position (about $27.64 from a $28.31 entry), the strategy's take-profit level is +10% (about $31.14), and single positions are capped at 25% of the book with sizing tied to a fixed 2.35% risk per trade. That is roughly a 4-to-1 reward-to-risk profile per trade. Note the second-tier support line sits just below the entry price, so a support-based stop would be even tighter than the 2.4% stop. **Wait on IREN.** IREN closed at $46.93 with RSI at 68.7 — two of its four entry conditions are not met: RSI needs to be at or below 60 (it is 8.7 points too high) and price needs to be below the 50-day EMA at $42.40 (it is $4.53, or about 9.7%, above it). Waiting concretely means no IREN position until both RSI cools below 60 and price closes back under $42.40 — or an alternative entry triggers via a tap-and-hold of first support at $46.00. Chasing the ERCOT-driven strength (the idea's thesis, per the CNBC rotation report) at 68.7 RSI is exactly what the entry band is designed to avoid. **Evidence read:** this is a backtested setup — over 24 months the strategy produced a 79.0% return on 66 trades with a 48.5% win rate and a 28.7% maximum drawdown, so the 48% win rate is expected behavior, not a flaw; the payoff skew does the work. Also note the support-based stop-loss fills were approximated on daily bars in that test, so treat realized drawdown as coarse rather than precise.

Grid access plus sector rotation: why the momentum case has real support

The thesis, per the Yahoo Finance report from September 8, 2026, is that IREN's 7% jump on conditional ERCOT status for its 2GW Sweetwater hub is a genuine repricing event: the company now controls the scarcest input in the AI buildout, firm power. Applied Digital climbing 6% on the same day suggests the market is pricing the pair as a theme, not isolated headlines. The same-day CNBC piece on Qualcomm and Corning deal announcements, which lifted the broader AI-infrastructure group including Intel and AMD, corroborates the rotation argument the idea leans on. The fundamentals give the power-landlord story some footing on the IREN side. For fiscal 2025 (year ended June 30, 2025), IREN reported $501.0M of revenue, up 167.7% year over year, with a 68.3% gross margin, positive net income of $86.9M, and $245.9M of operating cash flow. That is a business that has already crossed into profitability while building datacenter capacity, which is more than can be said for most of the peer group. The quant evidence is backtested, not hypothetical. Over a 24-month daily window the rule set traded 66 times on the IREN pair and returned 78.97% with a 48.5% win rate, with a peak drawdown of 28.7%. Equity spiked as high as roughly +139% before giving some back, which is consistent with a momentum-with-stop approach to a high-beta theme. The strategy's structure — entries near support and below the 50-day EMA band, a 2.4% stop and 10% profit-taking with 25% max position size — matches the idea's…

Scores

• Conviction score breakdown: 53
• Thesis support: 70
• Trade readiness: 60
• Risk quality: 45
• Backtest evidence: 55
• Fundamentals trend: 35

Watch items

• APLD — RSI (14)
• APLD — Price vs 50-day EMA
• APLD — Position stop-loss
• APLD — Take-profit level
• IREN — RSI (14)
• IREN — Price vs 50-day EMA
• IREN — First support
• IREN — Second support structural stop
• APLD — Next quarterly filing (free cash flow)
• IREN — ERCOT grid-status milestones for the 2GW hub