1 Artificial Intelligence (AI) Stock Flying Under the Radar That I Think Every Investor Should Consider

Cipher Digital is positioning for substantial AI data center growth, with 2.5 gigawatts expected to be energized in 2028-2029 and 270 megawatts ready in 2027. The company has already signed a 15-year, $5.5 billion Amazon deal for 300 megawatts, implying about $367 million in annual revenue once operational. While Q1 revenue fell 29% year over year because it is still mostly a crypto-mining business, the shift toward long-term AI infrastructure contracts is a clear positive for future recurring revenue.

Analysis

The market is still underpricing the optionality in CIFR’s business mix shift because it is valuing the company on current revenue, not on contracted capacity conversion. The key second-order effect is that the landlord model should compress execution risk versus vertically integrated peers: fewer sunk dollars per MW, faster site turn-up, and a higher probability that pre-leased capacity converts into visible revenue ramp without the balance sheet strain typically associated with AI infrastructure buildouts.

That said, the competitive moat is narrower than the headline suggests. IREN and NBIS may look more capital intensive today, but that intensity also gives them more control over the full stack and potentially higher eventual revenue per MW if chip supply loosens or customer demand shifts toward managed solutions. CIFR’s lower per-unit economics mean it can scale faster, but also that it is more exposed to contract pricing pressure if hyperscalers regain leverage or if power availability becomes commoditized across the sector.

The biggest catalyst is not the signed leases; it is the 2027-2029 energization schedule. This creates a classic “long-duration backlog” setup where the stock can rerate well before cash flow appears, but only if the market believes the company can keep hitting milestone dates without dilution or permitting slippage. The main tail risk is a multi-quarter delay in energization combined with equity issuance to fund capex, which would turn the current operating leverage story into a financing story.

Consensus is too focused on near-term revenue optics and too dismissive of the fact that this is becoming a capacity-options business. If CIFR can repeatedly pre-sell future MW, the equity should trade less like a miner and more like an infrastructure developer with embedded scarcity value in power access. The contrarian miss: the real upside may come not from the lease economics themselves, but from a sector-wide scarcity premium on energized power at a time when AI demand is forcing customers to pay up for speed rather than perfection.

Key Decisions for Investors

• Long CIFRW/CIFR on pullbacks over the next 1-3 months: the asymmetry is strongest before energization milestones re-rate the name; target 30-50% upside if execution stays on track, but cut if site timing slips materially.
• Pair trade: long CIFR vs short IREN or NBIS into the next 6-12 months; CIFR should benefit more from a capital-light leasing model if investors continue rewarding faster conversion of power into contracted revenue.
• Add AMZN as the highest-quality beneficiary of cheaper, faster AI capacity buildout over 12-24 months; if CIFR proves delivery, Amazon’s external capacity strategy looks more scalable, while model risk sits with suppliers, not the buyer.
• Avoid chasing NVDA or SNDK here as a direct trade on this article; the read-through is second-order and already partially embedded in the AI infrastructure complex, so the cleaner risk/reward sits in infrastructure landlords and hyperscale customers.
• If long CIFR, hedge with a short-dated call spread or cash-secured put structure around known capital-markets or execution windows; this protects against dilution/permit-delay downside while preserving upside into 2027-2029 milestone re-pricing.