What Should I Invest In? I'm Putting My Money in These 2 Stocks for 2027.

Iren reported $1.0B in operating annual recurring revenue and targets $4.0B ARR by end-2026, supported by a $9.7B five-year Microsoft deal and negotiations for $25M per megawatt (up to a potential $145B ARR if realized across its 5.8GW portfolio). Netlist secured a Samsung licensing agreement worth $239M up-front plus up to $32.9M quarterly royalties for five years, with total potential gross license revenue up to $897M, and it still faces a $445M patent-infringement obligation from Micron. The article frames both names as AI infrastructure winners, despite some investor concern over capex for Iren and a prior Micron-related headline for Netlist.

Analysis

The important mechanism in IREN is not “AI demand” but the monetization of scarce power + land + interconnects into a scarce-capacity annuity. If management can keep signing contracts at meaningfully higher $/MW while customers prepay a large chunk of build cost, the equity becomes less about financing risk and more about option value on future pricing. The market’s mistake is assuming ARR growth is linear; in this model, one or two large deals can reprice the entire asset base.

The nearer-term risk is that outsourced compute is still a buyer’s market unless hyperscaler demand stays tight. If GPU supply eases, or if cloud customers internalize more capacity, per-MW pricing can flatten quickly and the multiple can compress before the operating story breaks. That makes the next 1-3 quarters the key window: contract announcements matter more than reported revenue, while 6-18 months depends on whether IREN can prove its backlog is repeatable rather than one-off.

NLST is a different animal: this is a legal monetization trade, not a clean AI operating story. The Samsung outcome creates a template, but the market is probably over-extrapolating that template to Micron before the cash actually lands. The upside is asymmetric if collections and royalties persist; the downside is that settlement headlines often reprice faster than they convert to durable free cash flow, so this is a catalyst-driven name, not something to underwrite on terminal value.

Contrarian view: the consensus may be underestimating how much of the AI infrastructure trade is already in the pricing of “scarcity” assets, while still underpricing the litigation optionality in NLST. But the reverse is also true: if the scarcity narrative stalls, these names can de-rate hard because much of the thesis sits in future contract economics rather than current earnings.

Key Decisions for Investors

• Long IREN on pullbacks over the next 1-3 months; treat as a contract-announcement trade, not a valuation anchor. Upside is a re-rating if $/MW continues to move higher; stop if new bookings or forward ARR guidance decelerate materially.
• Pair trade: long IREN / short NBIS for relative-value exposure to the same neocloud theme. Thesis is that IREN has more room for multiple expansion if it closes the revenue-recognition gap; invalidate if NBIS signs a larger, cleaner hyperscaler contract or IREN stumbles on execution.
• Hold NLST only as a small event-driven position until cash collection and royalty cadence are confirmed in filings. The trade works best on verification, not headlines; if Micron-related recovery slips or appeals drag, cut exposure.
• Avoid shorting MU outright on the NLST settlement story alone. The alleged legal overhang is too small relative to MU’s memory-cycle drivers; the better expression is a watch item on royalty-related margin leakage rather than a standalone short.
• Set an alert on IREN contract pricing / MW disclosures. If pricing stalls for two consecutive updates, assume scarcity premium is peaking and reduce exposure before the market reassesses the whole neocloud basket.