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

Iren says it ended fiscal 2026 Q4 with $1.0B in operating annual recurring revenue and expects $4.0B ARR by end-2026, with value per megawatt rising as it waits to sign larger deals (a noted Microsoft five-year deal at $9.7B implies $9.7M per megawatt-year; Iren is negotiating $25M per megawatt). Netlist secured a Samsung licensing deal after a patent dispute: $239M upfront plus up to $32.9M quarterly royalties for five years, potentially reaching $897M gross license revenue over the contract, plus rights to buy up to $300M of Samsung memory products annually. The article frames recent Netlist share weakness on a separate Micron appeal loss as misplaced, citing a separate $445M Micron catalyst.

Analysis

The cleaner read-through is not that these names are "AI winners," but that capital intensity is becoming the moat. For IREN, the market should focus less on reported ARR and more on whether the firm can keep converting scarce power into contracted economics faster than its financing cost rises. If prepayments and GPU financing really reduce dilution, that is a meaningful underwriting advantage; if not, the equity will trade like a levered infrastructure story, not a software multiple, and peer valuation expansion across neoclouds could stall.

Second-order, the upside in this theme accrues to the firms with the best access to chips and power, not necessarily the operators with the biggest valuation narratives. That means the best risk-adjusted exposure may still be the pick-and-shovel layer, with NVDA benefiting from every incremental dollar of deployed compute even if operator margins compress. Near term, IREN/NBIS will likely trade on financing and contract cadence; over 6-18 months, the key question is whether megawatt pricing becomes durable enough to justify infrastructure-style multiples or whether competition forces returns back toward cost of capital.

NLST looks more like a litigation monetization trade than an operating compounding story. The market often extrapolates one settlement into a repeatable royalty stream, but the real variable is enforcement credibility and collection timing; if that slips, the present value collapses quickly. The contrarian risk is that investors may be underpricing the precedent value for future counterparties, but overpricing the speed and certainty of cash conversion, especially if Micron-related milestones drag out on a months-long rather than days-long path.

Key Decisions for Investors

• Long IREN / short NBIS in equal dollars for the next 1-3 months as a relative-value trade; thesis is that IREN’s valuation gap can close if contract monetization accelerates faster than financing fears. Falsify if NBIS keeps converting revenue more cleanly or IREN fails to show incremental MW pricing power.
• Use NVDA as the higher-quality expression of the same compute-buildout theme over a 6-12 month horizon; if neocloud capex remains elevated, the margin pool ultimately accrues to the chip supplier rather than the balance-sheet-heavy operator. Prefer buying NVDA on weakness versus chasing operator breakouts.
• Do not chase NLST as a core long here; treat it as event optionality only. Wait for confirmed cash collection and the next Micron/legal catalyst before sizing a position, since the market may be overestimating how quickly a settlement translates into durable earnings.
• If taking any NLST exposure, keep it small and tactical: buy only after pullbacks and use the next legal milestone as the risk check. The thesis breaks if subsequent payments are delayed, reduced, or the Micron path starts to look materially less replicable than the Samsung resolution.