KEEL Bet on Transforming Mining Sites into AI Compute Campuses

The original poster discusses KEEL, formerly Bitfarms/BITF. The core thesis is not a traditional Bitcoin mining valuation, but that the market may re-rate the company as an AI/HPC data center power and campus developer. The post argues that KEEL resembles the “miner-to-compute-infrastructure” stories of IREN, CIFR, APLD, WULF, and others: the key bottlenecks for AI data centers are available power, grid interconnection, and suitable locations, and KEEL already owns several sites with power infrastructure. Therefore, even though it has not yet signed a large AI lease, the market has begun to price its power assets in advance. The evidence chain cited by the author includes: the company has an approximately 2.2GW pipeline; management’s goal is to drive three lease signings by the end of 2026 at near-term sites such as Panther Creek, Sharon, and Moses Lake; Sharon, Pennsylvania is a key project, described as having 110MW secured, of which 30MW is operating and 80MW is under development, located on an 18-acre former industrial-site redevelopment project with zoning progress already achieved; Panther Creek has 350MW secured and could potentially expand to more than 500MW over the long term, making it a larger-scale validation point. The author also emphasizes that Leopold Aschenbrenner’s Situational Awareness sharply increased its position according to its 13F, arguing that this shows AI infrastructure capital has noticed the company; an options put/call ratio of about 0.07 is used to indicate extremely bullish market sentiment. On valuation, the poster says the company’s market cap is about $3.4 billion and argues that if Sharon or Panther Creek signs a “real AI lease,” the stock could be rapidly re-rated; if management partially delivers, it could rise 2-3x, and if all three leases plus the Vera Rubin demand narrative materialize, even 5-10x is imagined. The poster’s position is 2,297 shares in a regular account at a cost basis of $5.22 and 1,795 shares in a ROTH account at a cost basis of $5.24; the main post did not include images, but the poster shared a position screenshot in the comments as proof. The main catalysts are lease signings in 2026, permitting and construction progress at Sharon/Panther Creek, continued momentum in the AI data center theme, potential Russell 2000 inclusion, and so-called “large fund discovery.” The implicit assumptions are that hyperscale customers will be willing to lease these high-power campuses, KEEL’s power assets can be converted into high-quality long-term leases, construction and permitting will not be significantly delayed, and capital markets will value it as an AI data center company rather than as an unprofitable miner. The biggest weaknesses are also clear: the thesis is highly dependent on leases that have not yet been signed and on management execution; the existing business remains exposed to Bitcoin and transformation costs, and Q1 still showed meaningful operating losses. Although projects such as Sharon have zoning approval, they still require further local approvals, and resident opposition over noise, water, and electricity costs could also increase delays. In addition, the stock has already risen significantly, and the 5-10x scenario in the post is more of an emotional upside case than a rigorous valuation based on contracted cash flows.

The comments section is generally tilted toward momentum-chasing and position sharing, with few substantive rebuttals. Multiple users said they already owned shares or long-dated options, arguing that other miner-to-AI companies that signed contracts saw market caps rise from around $3 billion to $10 billion as a reference point, while some called for the stock to exceed $10 this year. A few highly upvoted comments admitted they “don’t really trust the company” and were only participating because they had made money on a similar NBIS trade. After being asked to show positions, the OP posted a screenshot and replied that they were bullish several months out. The discussion sample is sentiment-heavy and lacks in-depth questioning of lease, financing, and approval risks.