
Riot Platforms signed a $9.0B, 20-year compute deal with Anthropic to lease 191MW of power at its Rockdale, Texas campus, positioning Riot to transition from bitcoin mining to AI infrastructure landlord. Revenue is expected to total $9.1B over the initial term, rising to about $16.1B if extended by two additional 5-year periods. Shares spiked more than 20% before giving back gains, while Compass Point highlighted a “two-tenant campus” with $9.8B in contracted data center revenue and reiterated a Buy rating with a $29 price target, despite ERCOT scrutiny that may slow speculative projects.
This is less a bitcoin story than a scarcity-of-grid-power story. If the contract terms are real and financeable, RIOT’s equity should migrate from a high-beta mining proxy to a quasi-utility on a long-duration cash-flow stream, which can justify a materially higher multiple than the market historically applies to miners. The second-order effect is negative for pure-play names like MARA and, to a lesser extent, CLSK: investors will increasingly discriminate between those with monetizable power infrastructure and those still carrying full BTC commodity exposure.
The near-term reaction is likely to stay noisy because the market will discount headline revenue heavily until it sees capex, delivery milestones, and tenant credit protections. The key 1-3 month catalysts are contract filing details, financing structure, and ERCOT/process approvals; the main bear case is that RIOT has merely swapped one capital-intensive business for another with long payback and execution risk. Over 6-18 months, the real upside is only realized if this becomes a repeatable leasing model, not a one-off rebrand.
The contrarian view is that the move may be partially overdone: 20-year nominal revenue is not the same as present value, especially if the project needs substantial power, cooling, and retrofit spending before cash flow turns positive. If bitcoin rallies materially, the relative value of the mining fleet itself rises and the market may reassess whether giving away optionality to become an AI landlord was optimal. The thesis is falsified if RIOT cannot secure timely permits/interconnection, if tenant economics are weaker than implied, or if disclosure shows the lease requires too much upfront spend for too little current FCF.
AI-powered research, real-time alerts, and portfolio analytics for institutional investors.
Request DemoOverall Sentiment
strongly positive
Sentiment Score
0.55
Ticker Sentiment