Keel Infrastructure reported a $145M net loss after its Bitfarms rebrand, yet shares rose 8.3%. The loss included $41M in non-cash bitcoin charges and a $22M credit facility extinguishment. Liquidity stands at $533M. The next catalyst is a signed HPC lease.
Alpha Score of 50 reflects moderate overall profile with strong momentum. Based on 1 of 4 signals – score is capped at 50 until remaining data ingests.
Keel Infrastructure Corp. reported a $145 million net loss for the first quarter of 2026, the first earnings release after the company completed its rebrand from Bitfarms Ltd. and redomiciled to the United States. Shares of KEEL rose 8.3% to close at $4.30 on May 11. The stock move ignored the headline loss and focused on the composition of the charges and the liquidity that underpins the AI infrastructure pivot.
The loss includes $41 million in non-cash digital asset fair value changes, a $22 million loss from extinguishing the Macquarie credit facility, and elevated G&A tied to restructuring. For traders, the immediate question is whether the market is correctly pricing the execution risk of a company that has abandoned its legacy revenue stream for a capital-intensive, supply-constrained HPC buildout.
Keel Infrastructure’s $145 million net loss was not a simple operational miss. The company attributed $41 million of the loss to non-cash changes in the fair value of digital assets. This accounting line reflects the mark-to-market of bitcoin holdings during a quarter when BTC price volatility was elevated. A further $22 million loss came from the extinguishment of the Macquarie credit facility, a deliberate step to clean up the balance sheet as part of the strategic reset.
The operating loss reached $98 million, up from $35 million in the same quarter a year earlier. General and administrative expenses climbed to $27 million from $18 million year over year. Keel linked the increase to professional services tied to the U.S. redomiciliation, GAAP conversion, and the Paso Pe sale, which closed on April 21. These are largely one-time costs, yet they highlight the frictional expense of unwinding a multinational bitcoin mining operation.
The Paso Pe transaction closed on April 21, and Keel stated it is exiting operations in Latin America entirely. The restructuring involves shutting down physical mining sites and selling associated infrastructure. The cost of that exit is embedded in the quarter’s G&A spike, and any residual liabilities from those jurisdictions remain a tail risk.
CEO Ben Gagnon framed the rebrand as the completion of a strategic transformation. The company now targets, in his words, “supply-constrained HPC/AI markets in North America” and is advancing primary sites through lease execution. The pivot is a bet that the scarcity of high-performance computing capacity will command premium pricing, and that Keel’s existing power infrastructure and data center footprint can be repurposed faster than greenfield competitors can build.
“supply-constrained HPC/AI markets in North America”
On April 1, Bitfarms became Keel Infrastructure Corp. and redomiciled to the United States. The corporate restructuring made Keel the ultimate parent of Bitfarms Ltd. The practical effect is a shift in regulatory oversight, tax treatment, and investor perception. A U.S.-domiciled AI infrastructure company trades in a different valuation universe than a Canadian bitcoin miner, and the 8.3% post-earnings rally suggests the market is willing to price in that optionality.
The pivot leaves a near-term revenue gap. Bitcoin mining generated the bulk of Bitfarms’ historical top line. Keel has not yet disclosed any material HPC or AI-related revenue. The company sold 269 bitcoin for $20 million in proceeds between January 1 and May 8, reducing its digital asset exposure. Those sales provided operating cash. They also signal that the company is not waiting for a BTC price recovery to fund the transition. The risk is that the AI revenue ramp takes longer than the cash burn allows.
As of May 8, Keel reported total liquidity of approximately $533 million. This included $336 million in unrestricted cash and $197 million in unencumbered bitcoin. The liquidity position is the single most important number for traders assessing whether the company can survive the transition without dilutive capital raises.
The $197 million in unencumbered bitcoin is both a buffer and a risk. A sharp decline in BTC price would reduce the value of that reserve, shrinking the liquidity cushion. Keel has already demonstrated a willingness to sell bitcoin to fund operations. If the AI buildout requires more capital than expected, further BTC sales could pressure the market and signal desperation. Conversely, if bitcoin rallies, the company gains optionality to fund the pivot without equity dilution.
The $336 million in unrestricted cash provides a meaningful runway. The operating loss of $98 million in a single quarter, even with one-time items, implies a quarterly cash consumption that could eat through reserves in two to three years if no AI revenue materializes. The company’s ability to execute lease agreements and secure HPC customers will determine whether that cash is a bridge to profitability or a countdown clock.
The market’s willingness to look through the $145 million loss hinges on tangible progress in the AI infrastructure buildout. Gagnon’s reference to “advancing primary sites through lease execution” is the closest thing to a forward catalyst. Traders should watch for announcements of signed leases with hyperscalers or enterprise AI customers. A single material contract would validate the pivot and likely re-rate the stock toward the valuation multiples of pure-play AI infrastructure companies.
A credible HPC lease would specify power capacity, term length, and a take-or-pay structure that provides revenue visibility. Keel’s existing power infrastructure in North America gives it a potential speed-to-market advantage. The company must prove it can retrofit bitcoin mining sites to the higher power-density and cooling requirements of AI workloads. Any lease announcement that lacks specific capacity commitments or counterparty names should be treated with skepticism.
AlphaScala’s proprietary Alpha Score for KEEL sits at 50/100, a Mixed reading that reflects the binary nature of the pivot. The score incorporates the liquidity buffer as a positive factor, offset by the absence of AI revenue and the execution risk inherent in a full business model transformation. The stock’s post-earnings rally suggests momentum traders are betting on the narrative. The Alpha Score cautions that the fundamentals have not yet shifted. For more on the stock’s metrics, see the KEEL stock page.
Keel’s remaining $197 million in bitcoin ties the stock’s risk profile to crypto markets even as the company tries to distance itself from mining. A sustained BTC selloff would reduce the value of the reserve and could force the company to sell into weakness, compounding the downside. Additionally, any residual liabilities from the Latin America exit or the Paso Pe sale could surface in future quarters.
The $22 million loss from extinguishing the Macquarie credit facility removed a secured debt obligation. It also consumed capital that could have been deployed toward the AI buildout. The decision to pay off the facility rather than restructure it suggests the company wanted a clean balance sheet. If Keel needs to raise debt again for the AI pivot, it will be starting from scratch with lenders who have no history with the new entity.
A delay in lease execution beyond the next two quarters would be the most damaging signal. Without a visible path to AI revenue, the cash burn becomes the dominant narrative. A simultaneous decline in bitcoin prices would shrink the liquidity buffer and could force an equity raise at a depressed valuation. Any regulatory or permitting hurdles at the primary North American sites would also undermine the speed-to-market thesis that underpins the pivot.
The market has given Keel the benefit of the doubt for one quarter. The next earnings report needs to show more than a rebrand and a clean balance sheet. It needs to show a customer.
Drafted by the AlphaScala research model and grounded in primary market data – live prices, fundamentals, SEC filings, hedge-fund holdings, and insider activity. Each story is checked against AlphaScala publishing rules before release. Educational coverage, not personalized advice.