Hut 8 vs. Riot Platforms: Which Bitcoin Miner Turning Data Center Developer Is the Better Stock Now?

The article contrasts Hut 8 vs. Riot Platforms as both pivot from Bitcoin mining to AI/data-center compute, highlighting Hut 8’s 15-year River Bend AI data center lease valued at ~$7B and AMD-linked expansion at Riot. Despite strong top-line growth (Hut 8 FY2025 revenue ~$235.1M, +45%; Riot FY2025 revenue ~$647.4M, +72%), both posted very large net losses and negative free cash flow (Hut 8 FCF -$132.6M; Riot FCF -$774.3M), driven largely by Bitcoin mark-to-market impacts. Valuation is cited as a differentiator, with Riot trading at a much lower forward P/E (20.9x vs. 84.8x for Hut 8), leading to a conclusion that Riot is the more attractive 2026 entry, but with execution and power-regulatory risks.

Analysis

The market is still pricing these as leverage-to-Bitcoin stories with an AI veneer, but the real variable is cash conversion over the next 2-4 quarters. In this segment, the scarce asset is not “AI narrative”; it is contracted megawatts that can be financed without serial dilution. That makes balance-sheet flexibility and pre-committed revenue far more important than a low sales multiple, because the capex cycle is front-loaded while monetization is delayed.

Second-order, the winners from this buildout may be the chip and power ecosystem, not the operators themselves. The miners-turned-data-center plays are effectively swapping commodity exposure for project-finance risk, which tends to compress returns if utilization ramps slower than hoped. If BTC weakens while buildout costs stay sticky, the market will punish the equity twice: lower token value and a higher cost of capital.

Contrarian view: consensus is overvaluing “cheap” and undervaluing execution quality. A lower P/S can be a value trap when free cash flow is deeply negative and the business still depends on a volatile treasury asset for funding. The key falsifier is whether new capacity actually comes online on schedule with signed customers; if not, these names revert to mining proxies within one reporting cycle.

Key Decisions for Investors

• Long HUT / short RIOT as a 3-6 month relative-value trade: HUT has better revenue visibility per dollar of capex and less cash burn, while RIOT’s valuation discount is justified unless it proves faster conversion of leases into utilization. Use the spread as the expression, not outright beta.
• Keep MARA as the cleaner short in any BTC drawdown over the next 1-3 months: it has the least offsetting contractual revenue and the highest sensitivity to a weaker coin tape. Cover if Bitcoin stabilizes and miner difficulty stops rising.
• Avoid chasing either name on AI headline strength until there is evidence of commencement dates, deposits, or utilization metrics. The risk/reward is poor for buying the rally before the balance sheet turns from funding tool to FCF engine.
• If you want upside to digital-infra spend without project risk, prefer AMD over the miners on pullbacks; it participates in compute demand without carrying grid, construction, and financing execution risk. Treat it as the cleaner 6-18 month exposure.