The chart is a lie. Not the kind of lie that misrepresents data—the kind that prices in a future that hasn't yet materialized. KEEL (formerly Bitfarms) saw its stock surge on the announcement of a pivot from Bitcoin mining to AI/HPC data centers. The market, in its infinite hunger for narratives, treated this as a rebirth. But when you peel back the layers—the power contracts, the site permits, the empty tenant pipeline—what you find is not a transformation. You find a company that has shut down its revenue engine and is now selling a story of electricity, hoping someone will pay for the privilege of using it.
Context: The Mining Exodus
KEEL is a public company that spent years building Bitcoin mining infrastructure across North America. Its assets were physical: land, power substations, cooling systems, and the regulatory approvals to run thousands of ASICs. The mining industry, however, entered a new phase after the 2024 halving. Margins compressed, energy costs rose, and the narrative shifted from "digital gold" to "energy waste." Miners with cheap power and flexible sites began looking for a second act. The playbook was written by Core Scientific, which emerged from bankruptcy by signing a 12-year, multi-billion-dollar deal with CoreWeave to host AI workloads. KEEL, like many others, decided to follow that script.
But the script is not the performance. The market sees KEEL as a data center developer in waiting. The reality is that KEEL is a mining company that has stopped mining, with no confirmed tenants for its repurposed facilities. As of the latest disclosures, the company has shuttered its Moses Lake, Washington site, and its Q2 revenue is expected to drop further. The pivot is a bet on a future that may or may not arrive.
Core: The Asset Repurposing Mechanism
Let's dissect the technical and financial reality. KEEL's transition is not a technology upgrade—it's an asset repurposing. The core assets are power capacity and site permits. The company has 96MW of capacity in Quebec (conditional transfer), and three priority sites in the US that are nearing full permitting. The bottleneck is not compute; it's power. The CEO explicitly stated that "power is the constraint." The three US sites are awaiting interconnection with PJM and Washington State grids, with uncommitted capacity expected by 2027.
Based on my audit experience in analyzing mining balance sheets, I've seen this pattern before. Companies often overestimate the speed at which power can be redirected. The reality is that grid interconnection queues are congested, and AI/HPC tenants are selective. They want not just power, but density—the ability to run high-performance clusters with advanced cooling. KEEL's sites were built for mining, which requires lower density and simpler cooling. Retrofitting to liquid cooling or direct-to-chip cooling is capital-intensive and time-consuming.
Moreover, the company has not signed a single data center tenant. The income gap is real: they shut down the Moses Lake mine in April, sacrificing revenue, while hoping to secure a tenant later. This is a liquidity gamble. The company is burning cash on idle sites, paying for permits and interconnection fees, with no offsetting revenue from the new use case. The market prices the narrative of "AI transition," but the balance sheet still reflects a mining company in decline.
**The narrative arbitrage here is clear: The stock is trading on the promise of a CoreWeave-style deal, but KEEL is not CoreWeave. It's a small player with MW-level assets, competing against GW-level hyperscalers. The market is ignoring the unit economics. The cost to repurpose a mining site for AI/HPC is not trivial. It requires new cooling, higher power redundancy, and often network upgrades. Those costs are not yet capitalized in the market's valuation.
Contrarian: The Blind Spots of the Pivot Narrative
Here is the contrarian angle that the market is missing: The pivot to AI/HPC is not a differentiated strategy. Every mining company with a power contract is now pitching the same story. Marathon Digital, Riot Platforms, and even smaller players like Hut 8 are all repurposing sites or selling power. The glut of former mining capacity is about to hit a market that may not have infinite demand. The AI infrastructure boom is real, but the demand is concentrated in the highest-density, lowest-latency locations. KEEL's sites in Quebec and Washington may not meet the latency requirements for real-time AI inference. They are better suited for batch processing or training, but those segments are already saturated by hyperscalers.
Another blind spot is the regulatory risk. The Quebec capacity is conditional on local approvals, and the US sites face community opposition to data center construction. The narrative of "green AI" is fragile—if local grids impose moratoriums or demand higher interconnection fees, the economic case collapses. The market is pricing in a smooth transition, but the history of infrastructure projects is one of delays and cost overruns.
The liquidity is a mirror, not a foundation. KEEL's stock price reflects the reflection of another company's success (Core Scientific), not its own fundamentals. When the market realizes that the tenant pipeline is empty and the timeline slips, the mirror will shatter. The current valuation is a bet on narrative, not on execution.
Takeaway: The Next Narrative Will Be About Tenants
The next phase of the story will be determined by one thing: a signed lease. Without a tenant, KEEL is a mining company with no mines and a pile of power contracts. The market will eventually demand proof of revenue. The question is whether the company can secure a deal before the cash burn forces a dilutive capital raise. The signs are not promising. The company has not disclosed any serious negotiations, and the timeline for the first US site is 2027. That is a long time to wait for a narrative to materialize.
Every chart is a story waiting to be corrected. The correction will come when the market decodes the narrative before the price reacts. For now, the story is about power. Soon, it will be about whether anyone actually wants to use it.
Illusions break; logic remains. KEEL's transition is a textbook case of narrative economics: the market buys the story, but the underlying business model is still in search of a customer. The arbitrage lies in understanding that human fear of missing out on AI is driving the price, not the fundamentals. The question is: when will the fear turn to disappointment?
Who owns the attention? Follow the capital. Right now, the capital is following the narrative. But capital is also impatient. If KEEL doesn't convert its power assets into signed contracts within the next 12 months, the attention will shift to the next story. And the old story—the one about mining—will be forgotten.
Decoding the narrative before the price reacts is the only way to profit from this transition. The current price is already pricing in a successful pivot. The real alpha is in understanding that the pivot is a long shot, not a sure thing. The sites are real, but the business model is not. The power is there, but the customers are not. The story is compelling, but the execution is uncertain.
In the end, KEEL is a test case for the entire mining-to-AI thesis. If they fail, it will be a signal that the market overestimated the ease of repurposing. If they succeed, it will validate the model. But the odds are not in their favor. The market is pricing in a narrative that has not yet been proven. As a Narrative Hunter, I see the story, but I also see the gaps. And the gaps are wider than the market cares to admit.
