Jefferies initiates Riot Platforms stock with buy on AI potential By Investing.com

Jefferies initiated Riot Platforms with a Buy rating and a $37 price target, implying about 48% upside from $24.92. Riot also reported Q1 fiscal 2026 revenue of $167.2 million, up 9.4% sequentially, including $33.2 million from first-time data center revenue. The stock’s AI/data center optionality, AMD lease expansion at Rockdale, and nuclear-powered data center collaboration with Terrestrial Energy support the bullish thesis.

Analysis

The market is starting to re-rate RIOT less as a pure hash-rate proxy and more as a hybrid power-and-real-estate option on AI load growth. That transition matters because the equity can sustain a higher multiple only if data-center revenue becomes repeatable and underwritten by long-duration tenants; otherwise the stock remains tethered to mining sentiment and crypto beta. The first-order beneficiary is RIOT’s own equity, but the second-order winners are equipment vendors and infrastructure contractors that can monetize a more credible buildout pipeline if anchor leases land.

The real competitive edge here is not the sites alone, but vertical integration in electrical gear and permitting execution. In a market where data-center timelines are bottlenecked by transformer/switchgear lead times, internal supply can compress deployment schedules by quarters and improve lease economics versus pure-play developers that must source everything externally. That advantage is most valuable in metro-adjacent inference workloads, where latency and power availability matter more than scale; if that thesis is right, RIOT’s sites could command better terms than remote hyperscale land banks.

The key risk is timing mismatch: capital markets are likely to price the AI optionality before the cash flow arrives, but tenant commitment can lag by 6-18 months. If AMD-related expansion is seen as a one-off fit-out rather than a template, the market may de-rate the story quickly because the stock already embeds meaningful success. Also, nuclear-powered data center collaboration is a long-dated call option, not near-term earnings support; investors should treat it as upside optionality, not a valuation anchor.

Contrarianly, the move may be partially overdone if the crowd is extrapolating infrastructure scarcity into near-term rent power. The most interesting setup is a “prove-it” window over the next two quarters: if RIOT converts a second anchor lease or shows recurring lease revenue acceleration, the multiple can hold; if not, the equity likely trades back toward mining comps. The asymmetry is better on pullbacks than breakouts, because the fundamental catalyst stack is real but still under-monetized.

Key Decisions for Investors

• Buy RIOT on weakness, preferably on a 5-10% pullback from current levels, targeting a 3-6 month window; upside is driven by lease repeatability, while downside is capped if data-center revenues keep compounding.
• Pair trade: long RIOT / short a basket of pure-play miners over 1-2 quarters; the thesis is that RIOT’s power and infrastructure optionality should decouple from hash-rate-only names as AI monetization becomes more visible.
• Sell downside in RIOT via 2-3 month cash-secured puts or put spreads struck below the recent high; this expresses a view that the market will defend the story unless a lease/capex disappointment appears.
• Add AMD only as a secondary beneficiary on any RIOT-driven weakness in the market, not as the primary expression; RIOT-specific execution is the cleaner catalyst and better risk/reward for the next earnings cycle.
• Avoid chasing IMSRW until there is clearer evidence that the nuclear data-center collaboration moves from narrative to contracted power delivery; this is a 12-24 month option, not a near-term trade.