Bitfarms Ltd. primarily owns and operates data centers that house computers, referred to as miners, designed to validate transactions on the Bitcoin blockchain. The company runs its miners around the clock to generate hashing power that it sells to mining pool operators under a full pay per share arrangement. As of March 26, 2025, Bitfarms operates 15 data centers located in Canada, the United States, Paraguay, and Argentina, with an energized capacity of 461 megawatts and a…
Bitfarms Ltd. primarily owns and operates data centers that house computers, referred to as miners, designed to validate transactions on the Bitcoin blockchain. The company runs its miners around the clock to generate hashing power that it sells to mining pool operators under a full pay per share arrangement. As of March 26, 2025, Bitfarms operates 15 data centers located in Canada, the United States, Paraguay, and Argentina, with an energized capacity of 461 megawatts and a contracted capacity of 908 megawatts. The firm sources electricity from Hydro Quebec, Hydro Sherbrooke, Hydro Magog, the City of Baie Comeau, and predominantly power grid operators in Washington State and Pennsylvania in the United States, as well as from the Paraguay national electricity grid; much of this power is derived from clean hydroelectric generation. In addition to mining, Bitfarms provides electrical contracting services through its subsidiary Volta, which serves commercial and residential customers in Québec, Canada. The company also utilizes its proprietary MGMT System, upgraded to MGMT 2, to monitor miner location, status, computing power, and temperature to maximize uptime. Bitfarms’ strategy includes diversification beyond Bitcoin mining into areas such as energy generation, energy trading, heat recycling, and high performance computing and artificial intelligence.
Bitfarms generates revenue mainly by selling computational capacity to mining pool operators, which pays the company in Bitcoin on a daily basis under the full pay per share model. The Bitcoin received is either retained under the Company’s digital asset management program or exchanged for U. S. dollars through established cryptocurrency trading platforms. Under the digital asset management program, Bitfarms decides how much of the mined Bitcoin to hold in custody with trusted custodians such as Anchorage Digital Bank N. A. The firm also operates a hedging program designed to mitigate the impact of Bitcoin price volatility on its operating cash flows. Additionally, Bitfarms runs a synthetic hodl program that uses financial instruments to create Bitcoin equivalent exposure without holding the underlying asset. Volta contributes incremental revenue by providing electrical installation, maintenance, and repair services to commercial and residential clients in Québec, Canada. No other material revenue streams are described in the filing.
Bitfarms positions itself as a geographically diversified Bitcoin miner with access to low cost hydroelectric and alternative energy sources, which reduces its exposure to volatile electricity prices. The company competes with other publicly traded mining firms such as Riot Platforms, Stronghold Digital Mining, Hut 8 Corp., and Marathon Digital Holdings. Bitfarms differentiates itself through long term power contracts, proprietary MGMT 2 monitoring software, and a strategy to expand into high performance computing, artificial intelligence, energy generation, and heat recycling initiatives. These advantages aim to improve operating efficiency, lower operating costs, and create additional revenue streams beyond pure bitcoin mining. The firm’s ability to secure electricity at competitive rates across multiple jurisdictions provides a resilient foundation for sustained operations.
The company’s primary customers are mining pool operators that purchase hashing power to secure Bitcoin network rewards. Through Volta, Bitfarms also serves commercial and residential electricity customers in Québec, Canada. The filing does not disclose specific customer names, so the base is described by type rather than by individual entities.
Sectors:Technology · IndustrialsSector rationaleThe company's dominant revenue stream comes from operating data centers and selling computational hashing power to mining pool operators, which falls under Crypto Mining in the Technology sector. A secondary sector is justified because the company operates a distinct business line through its subsidiary, Volta, which provides electrical contracting, installation, and maintenance services to commercial and residential customers, fitting the Industrials sector.Industries:Crypto MiningTechnologyPrimaryBitfarms primarily operates data centers housing miners to validate Bitcoin blockchain transactions and generates revenue by selling hashing power to mining pool operators. The company's core economic output is the Bitcoin received from these mining operations.Utility ConstructionIndustrialsSecondaryThrough its subsidiary Volta, the company provides electrical contracting services, including installation, maintenance, and repair for commercial and residential customers in Québec.Classified using BQ-MICSCIK: 0001812477
Investment Thesis
▲ Bull case
The company holds approximately five hundred thirty three million dollars in cash and Bitcoin as of May 8 2026 which management states fully funds Panther Creek Sharon and Moses Lake through lease execution and construction commencement at Moses Lake as well as general and administrative expenses through 2028. This liquidity eliminates the need for additional external financing in the near term reducing dilution risk and preserving balance sheet strength. With a solid cash cushion the firm can continue to invest in development activities and pursue lease negotiations from a position of strength rather than necessity. The ability to fund the entire near term plan internally is a significant advantage that the market may be underestimating when assessing the company’s valuation.
The firm controls secure power capacity in three of the most supply constrained markets in North America including Panther Creek in PJM Sharon in Western PJM and Moses Lake in the Pacific Northwest corridor. These locations are situated near established hyperscaler and neocloud clusters giving Keel a locational advantage that competitors lacking similar proximity cannot replicate. Management emphasizes that power availability is the single biggest bottleneck constraining the growth of the AI economy and that the company’s secured and expandable capacity directly addresses this constraint. By offering ready to use power in high demand areas Keel can negotiate lease terms that reflect the scarcity premium potentially yielding higher rental rates than the market currently anticipates.
Keel is tailoring its data center designs to support next generation GPU deployments such as Nvidia’s Vera Rubin platform which management notes is seeing increasing demand relative to current generation hardware. The engineering requirements for Vera Rubin are described as a magnitude of order more complex and sophisticated than existing solutions indicating that early readiness could be a differentiator. By aligning campus architecture with the anticipated needs of Vera Rubin the company positions itself to attract tenants that require advanced infrastructure for future AI workloads. This forward looking approach may allow Keel to capture lease commitments from firms seeking to future proof their investments well before competitors can adapt their facilities.
The company is advancing permitting land development and environmental work streams in parallel with active lease negotiations meaning that progress on one front does not have to wait for completion on the others. Management explicitly stated that lease negotiations and permitting are a parallel process and that they have already achieved a credible pathway to power that can be shown to prospective tenants. This approach reduces the typical sequential delay that many developers face when waiting for permits before initiating commercial discussions. By maintaining parallel work streams Keel aims to secure leases by the end of the year and commence construction shortly thereafter accelerating the timeline to revenue generation.
At the Scrubgrass site a detailed load study covering seven hundred fifty megawatts is underway with results expected near the end of the year which could more than double the company’s currently secured capacity if the expansion is contracted. Securing this additional power would transform Scrubgrass into a potential giga campus in Pennsylvania a market where such large scale sites are rare outside of Texas. A giga campus outside of Texas is described by management as fiercely contested giving Keel a strategic advantage in a capacity constrained environment. The ability to bring a multi hundred megawatt campus online would create a long term growth driver that extends well beyond the near term lease execution targets.
The company holds approximately five hundred thirty three million dollars in cash and Bitcoin as of May 8 2026 which management states fully funds Panther Creek Sharon and Moses Lake through lease execution and construction commencement at Moses Lake as well as general and administrative expenses through 2028. This liquidity eliminates the need for additional external financing in the near term reducing dilution risk and preserving balance sheet strength. With a solid cash cushion the firm can continue to invest in development activities and pursue lease negotiations from a position of strength rather than necessity. The ability to fund the entire near term plan internally is a significant advantage that the market may be underestimating when assessing the company’s valuation.
The firm controls secure power capacity in three of the most supply constrained markets in North America including Panther Creek in PJM Sharon in Western PJM and Moses Lake in the Pacific Northwest corridor. These locations are situated near established hyperscaler and neocloud clusters giving Keel a locational advantage that competitors lacking similar proximity cannot replicate. Management emphasizes that power availability is the single biggest bottleneck constraining the growth of the AI economy and that the company’s secured and expandable capacity directly addresses this constraint. By offering ready to use power in high demand areas Keel can negotiate lease terms that reflect the scarcity premium potentially yielding higher rental rates than the market currently anticipates.
Keel is tailoring its data center designs to support next generation GPU deployments such as Nvidia’s Vera Rubin platform which management notes is seeing increasing demand relative to current generation hardware. The engineering requirements for Vera Rubin are described as a magnitude of order more complex and sophisticated than existing solutions indicating that early readiness could be a differentiator. By aligning campus architecture with the anticipated needs of Vera Rubin the company positions itself to attract tenants that require advanced infrastructure for future AI workloads. This forward looking approach may allow Keel to capture lease commitments from firms seeking to future proof their investments well before competitors can adapt their facilities.
The company is advancing permitting land development and environmental work streams in parallel with active lease negotiations meaning that progress on one front does not have to wait for completion on the others. Management explicitly stated that lease negotiations and permitting are a parallel process and that they have already achieved a credible pathway to power that can be shown to prospective tenants. This approach reduces the typical sequential delay that many developers face when waiting for permits before initiating commercial discussions. By maintaining parallel work streams Keel aims to secure leases by the end of the year and commence construction shortly thereafter accelerating the timeline to revenue generation.
At the Scrubgrass site a detailed load study covering seven hundred fifty megawatts is underway with results expected near the end of the year which could more than double the company’s currently secured capacity if the expansion is contracted. Securing this additional power would transform Scrubgrass into a potential giga campus in Pennsylvania a market where such large scale sites are rare outside of Texas. A giga campus outside of Texas is described by management as fiercely contested giving Keel a strategic advantage in a capacity constrained environment. The ability to bring a multi hundred megawatt campus online would create a long term growth driver that extends well beyond the near term lease execution targets.
Operating loss widened from thirty five million dollars in the prior year period to ninety eight million dollars in the Q1 FY26 reflecting increased operating expenses and a significant noncash depreciation charge. Adjusted EBITDA moved from positive seven million dollars to negative seventeen million dollars year over year indicating that the core business is currently burning cash before any lease revenue materializes. Management attributes part of the increase to a fifteen million dollar rise in energy and infrastructure expenses and a seven million dollar unfavorable shift in digital asset sale gains or losses. Until the company secures leases that generate recurring cash flow the ongoing cash burn could erode the large liquidity buffer if the timeline to revenue extends beyond current expectations.
The company has set a clear target of signing three leases by the end of the year one each at Panther Creek Sharon and Moses Lake but the Q&A dialogue revealed uncertainty about the exact timing and terms that management is willing to accept. When asked about leaning toward specific tenant types the CEO avoided giving a definitive answer indicating that negotiations are still fluid and that the final deal structure may shift. Lease negotiations in the data center sector often involve lengthy due diligence and custom build specifications which can extend timelines beyond initial projections. If the company fails to secure the anticipated leases by year end the expected inflection point of converting assets into contracted cash flows would be delayed affecting near term valuation.
Management described hyperscaler clients as being tighter on economics but offset by credit quality while neoclouds tend to pay higher rates but come with a higher cost of capital. This dichotomy creates a balancing act that could pressure lease economics depending on which tenant segment the company ultimately selects. If the mix leans heavily toward neoclouds the higher cost of capital may reduce the net present value of lease payments despite higher headline rates. Conversely a focus on hyperscalers could result in lower rental rates even though the tenant credit is strong. The uncertainty around the optimal tenant mix introduces risk to the projected cash flow profile from the planned leases.
Although zoning approvals are complete at all three near term sites land development and environmental permits remain in progress with a target completion window of mid to late summer. The company has expressed high confidence in this timeline but acknowledged that permitting processes can experience unexpected delays due to regulatory stakeholder or environmental considerations. Any slip in the permitting schedule would push back the earliest possible ready for service dates which are currently projected for 2027. A delay in achieving ready for service would postpone the commencement of lease revenue and could lead to a reassessment of the company’s near term growth prospects.
The current liquidity of approximately five hundred thirty three million dollars is said to fully fund Panther Creek Sharon and Moses Lake through lease execution and construction commencement at Moses Lake as well as cover general and administrative expenses through 2028. However this projection assumes that development costs will remain in line with the company’s existing capital plan and that no significant cost overruns will occur. Large scale data center construction is susceptible to increases in material prices labor expenses and unexpected site conditions which could increase the required capital beyond the available cash. If costs rise substantially the company might need to seek external financing potentially diluting existing shareholders or increasing leverage.
Operating loss widened from thirty five million dollars in the prior year period to ninety eight million dollars in the Q1 FY26 reflecting increased operating expenses and a significant noncash depreciation charge. Adjusted EBITDA moved from positive seven million dollars to negative seventeen million dollars year over year indicating that the core business is currently burning cash before any lease revenue materializes. Management attributes part of the increase to a fifteen million dollar rise in energy and infrastructure expenses and a seven million dollar unfavorable shift in digital asset sale gains or losses. Until the company secures leases that generate recurring cash flow the ongoing cash burn could erode the large liquidity buffer if the timeline to revenue extends beyond current expectations.
The company has set a clear target of signing three leases by the end of the year one each at Panther Creek Sharon and Moses Lake but the Q&A dialogue revealed uncertainty about the exact timing and terms that management is willing to accept. When asked about leaning toward specific tenant types the CEO avoided giving a definitive answer indicating that negotiations are still fluid and that the final deal structure may shift. Lease negotiations in the data center sector often involve lengthy due diligence and custom build specifications which can extend timelines beyond initial projections. If the company fails to secure the anticipated leases by year end the expected inflection point of converting assets into contracted cash flows would be delayed affecting near term valuation.
Management described hyperscaler clients as being tighter on economics but offset by credit quality while neoclouds tend to pay higher rates but come with a higher cost of capital. This dichotomy creates a balancing act that could pressure lease economics depending on which tenant segment the company ultimately selects. If the mix leans heavily toward neoclouds the higher cost of capital may reduce the net present value of lease payments despite higher headline rates. Conversely a focus on hyperscalers could result in lower rental rates even though the tenant credit is strong. The uncertainty around the optimal tenant mix introduces risk to the projected cash flow profile from the planned leases.
Although zoning approvals are complete at all three near term sites land development and environmental permits remain in progress with a target completion window of mid to late summer. The company has expressed high confidence in this timeline but acknowledged that permitting processes can experience unexpected delays due to regulatory stakeholder or environmental considerations. Any slip in the permitting schedule would push back the earliest possible ready for service dates which are currently projected for 2027. A delay in achieving ready for service would postpone the commencement of lease revenue and could lead to a reassessment of the company’s near term growth prospects.
The current liquidity of approximately five hundred thirty three million dollars is said to fully fund Panther Creek Sharon and Moses Lake through lease execution and construction commencement at Moses Lake as well as cover general and administrative expenses through 2028. However this projection assumes that development costs will remain in line with the company’s existing capital plan and that no significant cost overruns will occur. Large scale data center construction is susceptible to increases in material prices labor expenses and unexpected site conditions which could increase the required capital beyond the available cash. If costs rise substantially the company might need to seek external financing potentially diluting existing shareholders or increasing leverage.