Jefferies initiates Riot Platforms stock with buy on AI potential

Jefferies initiated Riot Platforms with a Buy rating and a $37 price target, implying about 48% upside from $24.92. The firm highlighted Riot’s AI/data center optionality, vertical engineering capabilities, and potential lease wins at Corsicana and Rockdale, while first-quarter fiscal 2026 revenue rose 9.4% sequentially to $167.2 million, including $33.2 million of initial data center revenue. Multiple other analysts also raised or reiterated bullish targets, supporting a constructive outlook despite execution risk.

Analysis

The market is starting to re-rate RIOT less as a pure miner and more as an option on scarce, utility-grade power with embedded industrial capabilities. The underappreciated second-order effect is that Riot’s in-house electrical gear business can compress deployment timelines and bargaining leverage versus pure-play colocation competitors that still depend on third-party procurement; in a supply-constrained AI buildout, schedule certainty is often worth more than headline megawatts.

The bigger winner may be AMD’s ecosystem, not just RIOT. If Riot can keep landing anchor tenants, AMD gets a differentiated path into edge-adjacent inference capacity where latency and site power matter more than brand moat, while also diversifying away from hyperscaler concentration. That said, the current setup still looks like a story stock until repeatable lease conversion is demonstrated across multiple sites; one-off fit-out revenue is not yet evidence of durable normalized cash flow.

Risk is mostly execution timing, not demand. Over the next 1-3 months, any delay in lease signing, interconnect, or capex funding can collapse the multiple fast because the equity is pricing infrastructure optionality before it is monetized. Over 6-12 months, the contrarian bear case is that investors are overestimating how much AI inferencing will remain power-dense near Riot’s sites versus shifting to incumbents with better balance-sheet flexibility, lower financing costs, and less customer concentration.

For NVDA, the policy backdrop is mildly incremental rather than transformative: limited incremental China demand for H200s helps clear inventory and supports mix, but it does not materially change the long-term U.S.-China decoupling risk premium. The main read-through is to supply-chain normalization and near-term sentiment, not a new earnings regime.

Key Decisions for Investors

• Long RIOT on a 1-3 month horizon only on pullbacks, targeting a breakout continuation trade toward prior highs with a tight stop below the post-earnings base; reward is another 15-20% if a second lease catalyst lands, but downside is sharp if tenant conversion stalls.
• Buy RIOT call spreads 2-4 months out to express upside from additional anchor-lease announcements while capping risk to premium; use strikes around current spot to avoid overpaying for momentum.
• Pair trade: long RIOT / short a higher-quality colo or power-infrastructure name if any valuation gap widens on AI optionality; the idea is to own the “power + engineering” scarcity premium while hedging broader AI beta.
• For AMD, stay constructive but use strength to trim; the trade is a slow-burn beneficiary of edge/inference deployment, but near-term upside is capped unless Riot-like deals become replicable across multiple sites.
• Avoid chasing NVDA on this headline alone; treat it as a sentiment-positive but low-conviction catalyst and only add on weakness if broader semis confirm improving China channel-clearing data.