Keel Infrastructure Corp. (Nasdaq/TSX: $KEEL): From Bitcoin Miner To AI Power Infrastructure — A High-Beta Data Center Pivot To Watch

Executive Summary

Keel Infrastructure Corp. is one of the more interesting speculative infrastructure names sitting at the intersection of three market narratives: AI power scarcity, data center capacity shortage, and the migration of former Bitcoin-mining assets toward high-performance computing and artificial intelligence infrastructure. The company is not a clean, mature data center REIT. It is not a traditional utility-backed infrastructure platform. It is not yet a proven hyperscale data center landlord. It is a former Bitcoin miner, previously known as Bitfarms, that has redomiciled to the United States, rebranded as Keel Infrastructure, and repositioned itself as a North American developer and owner of data centers and energy infrastructure for HPC and AI workloads.

The bull case is easy to understand. The market is no longer looking only for Nvidia chips, AI models, cloud software or hyperscaler capex beneficiaries. Increasingly, the market is searching for the physical infrastructure that makes AI possible: power access, grid interconnections, land, substations, transmission relationships, cooling, data center shells, construction capacity and contracted megawatts. KEEL is trying to enter that part of the value chain. Its own investor materials present a total development pipeline of 2.2 GW, including 341 MW of energized capacity, 430 MW of secured capacity and 1.5 GW of expansion capacity.

The bear case is just as important. Keel is still in the middle of a difficult corporate transformation. Its most recently reported quarter showed approximately $37 million of revenue, a large operating loss, and a very significant net loss. The company still primarily derives revenue from Bitcoin mining, while the HPC/AI strategy is still being built, financed and commercialized. That means investors are not buying a proven AI data center cash-flow machine today. They are buying a transition story where the market is attempting to price future lease execution before the economics have been fully proven.

The most important recent capital event is the June 2026 closing of $458 million aggregate principal amount of 1.250% convertible senior notes due 2032. The financing strengthens liquidity and gives the company flexibility for project development, equipment deposits and letters of credit. But it also creates a clear capital-structure reference point: the initial conversion price is approximately $7.41 per share and the capped-call cap price is $11.86 per share. Those levels matter for traders because they frame dilution psychology, upside mechanics and the market’s interpretation of future stock strength.

Snapshot

Why KEEL Matters Now

KEEL matters because the AI trade has moved beyond the question of who sells the GPU. The next question is who can power, house and operate the compute. The first phase of the AI equity trade was dominated by semiconductors, hyperscalers, cloud software and networking. The next phase has increasingly moved into electrical equipment, grid infrastructure, power generation, nuclear exposure, cooling, data center real estate, energy interconnection and former crypto-mining sites that may be repurposed for HPC or AI workloads.

This is the opening KEEL is trying to exploit. The company inherited and assembled infrastructure assets that were originally tied to Bitcoin mining economics but may have strategic relevance in an AI/HPC environment. Bitcoin mining and AI data centers are not the same business. The hardware is different. The reliability standards are different. Customer concentration risk is different. Cooling requirements can be different. Contract structures are different. Enterprise service-level expectations are higher. Still, both businesses share one essential foundation: large-scale access to electricity.

That shared foundation is why the market is willing to look at former miners with new eyes. If a miner owns, leases or controls sites with meaningful power access, the market may begin to treat those assets less like commodity Bitcoin-mining boxes and more like scarce digital infrastructure options. This is not automatic, but it is the reason KEEL has become a serious watchlist name for traders following the AI power bottleneck.

Keel’s own filings acknowledge that HPC and Bitcoin mining data centers consume large amounts of energy, and that energy costs can represent a significant portion of operating expenses. The company also notes that reliable, low-cost energy has become increasingly important because of global data center growth and wider electrification. That language is not just boilerplate. It describes the heart of the trade.

The Corporate Transformation: Bitfarms Becomes Keel

The formal transformation became visible in 2026, but the strategic shift started earlier. Keel’s Q1 2026 communication described the rebrand as the completion of a nearly two-year strategic transformation. CEO Benjamin Gagnon said the company had redomiciled to the United States, built out the team, exited Latin American megawatts and focused the development pipeline on high-demand and supply-constrained HPC/AI markets in North America.

The legal structure is important. Effective April 1, 2026, Keel Infrastructure Corp., a Delaware corporation, became the ultimate parent company of Bitfarms and its subsidiaries through a statutory plan of arrangement. Former Bitfarms shareholders received one share of Keel common stock for each Bitfarms Canada common share, and Keel became the successor issuer. The common stock began trading under the symbol KEEL on Nasdaq and TSX.

This matters for several reasons. First, U.S. redomiciliation can make the company easier to understand for U.S. investors, index providers, institutions and counterparties. Second, the shift from foreign private issuer to U.S. domestic issuer increases the relevance of U.S. reporting expectations and may improve comparability with domestic AI infrastructure peers. Third, the rebrand is designed to shift market perception away from Bitcoin mining and toward digital and energy infrastructure.

However, rebrands are not execution. A ticker change does not create revenue. A U.S. domicile does not automatically create data center leases. A new name does not remove the economic weight of the legacy business. Keel has taken major structural steps, but the commercial proof still needs to arrive.

The Asset Base: 2.2 GW Is The Headline, But Quality Matters

Keel’s investor overview presents the company as having a 2.2 GW total pipeline, broken into 341 MW energized, 430 MW secured and 1.5 GW expansion capacity. The definitions are crucial. Energized capacity is power capacity provided by utilities and currently being used on site. Secured capacity refers to executed agreements with utilities for future capacity delivery. Expansion capacity includes capacity under application, under utility study or being evaluated for behind-the-meter generation.

That distinction matters because traders often treat “GW pipeline” as if every megawatt has the same quality. It does not. Energized megawatts are different from secured future megawatts, and both are very different from expansion capacity that still requires utility work, studies, permitting, development or behind-the-meter generation. The market will likely reward KEEL more for concrete milestones than for pipeline size alone.

The milestones that matter are clear: utility agreements, zoning, permits, customer leases, long-lead equipment, construction starts, energization dates and customer revenue. Keel has said zoning has been secured and site development is on track across near-term sites including Panther Creek, Sharon and Moses Lake, while land development and environmental permits are in progress. That is useful progress, but the next step is commercial validation.

Panther Creek

Panther Creek is the asset most likely to dominate the near-term KEEL narrative. The company has described Panther Creek as a strategic Pennsylvania campus with significant power potential and a central role in the AI/HPC pivot. In its Q1 2026 filing, the company said it entered into an agreement with a utility provider in March 2026 to supply high-voltage electricity to a large-scale, energy-intensive facility at Panther Creek. Under that arrangement, the utility provider is responsible for power delivery and, if needed, for the design and construction of transmission links and substations.

Panther Creek also has a meaningful capital-structure history. Keel’s Q1 filing says the company had converted a credit facility into a $300 million project debt facility for Panther Creek development in October 2025, then fully repaid that facility in February 2026 and recorded a $21.6 million loss related to termination. That creates two possible readings. The bullish reading is that Keel improved flexibility and removed restrictive project debt. The cautious reading is that large infrastructure financing decisions are expensive and complex.

Sharon

Sharon is another important Pennsylvania property. The company’s Q1 filing says Keel entered into an agreement for the sale and delivery of electricity to its Sharon property over a four-year term, supported by a letter of credit to the utility provider. Sharon has received less public attention than Panther Creek, but it may become very important if Keel can demonstrate that the AI/HPC pivot is not a single-site story.

For traders, the next step is clear. Any update that links Sharon to a credible customer, a defined MW commitment, permitting progress, construction timing or project financing would increase the market’s confidence in KEEL as a repeatable platform rather than a one-asset option.

Moses Lake

Moses Lake is also part of the near-term development set. Management has repeatedly grouped Panther Creek, Sharon and Moses Lake together as the sites being advanced toward leasing and construction milestones. The company has said its liquidity was expected to fund these sites through lease execution and the start of construction at Moses Lake.

Moses Lake is relevant because Washington State power, data center demand and regional infrastructure can carry strategic value. But the same execution questions apply: customer identity, lease economics, capex, construction schedule, uptime standards and return profile.

Scrubgrass And Stronghold

Keel’s Stronghold acquisition is a core part of the infrastructure story. In March 2025, the company acquired Stronghold Digital Mining in a stock-for-stock merger. Keel’s filing describes Stronghold as a vertically integrated power generation and data center company focused on environmental remediation and reclamation services in Pennsylvania. The transaction aligned with the company’s strategic objective to diversify operations and expand its U.S. presence through vertical integration of power generation and energy arbitrage capabilities.

Strategically, this is meaningful. It gives Keel a more differentiated energy-infrastructure angle than a miner with simple leased facilities. But it also brings cost complexity. The Q1 filing shows infrastructure and energy expenses rose partly because of Panther Creek and Scrubgrass power plant operations following the Stronghold acquisition. That is the double-edged nature of the asset base: control of power assets may create upside, but operating and converting them into AI/HPC infrastructure is capital-intensive.

Business Model: The Intended Shift From Bitcoin To AI/HPC Infrastructure

Keel’s legacy revenue is still tied primarily to Bitcoin mining. The company’s Q1 filing explicitly says it primarily derives revenues from Bitcoin mining and earns Bitcoin in exchange for computational power sold to mining pool operators. This matters because the market narrative has already moved toward AI infrastructure, while the reported income statement has not fully moved there yet.

The intended business model is to leverage existing power, development and infrastructure capabilities into HPC and AI data centers that can support specialized workloads for enterprise and hyperscale customers. That transition is logical, but it is not automatic. A Bitcoin mining facility can tolerate economic and operational dynamics that an enterprise AI/HPC customer may not accept. AI infrastructure customers typically require better reliability, different cooling profiles, stronger network and security requirements, clearer service-level commitments and more complex commercial arrangements.

The attractive part of the story is that Keel may have access to scarce ingredients: power, land, grid interconnection and operating experience with high-energy compute. The difficult part is that AI/HPC customers are demanding, well-informed and able to choose between competing infrastructure providers. Power access can open the door. It does not by itself close the deal.

For KEEL, the first major lease will matter far beyond its headline value. It will tell the market what kind of customer is willing to sign, what kind of economics may be available, whether the company can move from energy access to data center execution, and whether the former miner identity is becoming a true infrastructure platform.

Financial Review: Liquidity Is Real, But Losses Are Heavy

The latest reported quarter is difficult if read as a normal operating business, but understandable if read as a transition quarter. For Q1 2026, Keel reported approximately $37 million in revenue, down from the prior-year quarter. The company reported a gross loss, a large operating loss, a loss from continuing operations and a net loss of approximately $145 million.

The balance sheet at March 31, 2026 showed hundreds of millions of dollars in cash, a meaningful digital asset position, significant property, plant and equipment, more than $1 billion in total assets, substantial total liabilities and positive stockholders’ equity. The company later said that as of May 8, 2026 it had approximately $533 million of liquidity, consisting of unrestricted cash and unencumbered Bitcoin. It also said it had sold Bitcoin as part of the wind-down of its Bitcoin position.

There are two core takeaways. First, Keel has more liquidity than many speculative infrastructure-transition companies. That matters because site development, letters of credit, long-lead equipment, engineering, deposits and construction are expensive. A weak balance sheet would make the whole story much harder to believe.

Second, the company is still burning capital and reporting large losses. This is not an asset-light software story. It is a capital-intensive infrastructure story. The path to value creation depends on returns from assets that still need to be commercialized.

Convertible Notes: Balance Sheet Strength Or Dilution Overhang?

The June 2026 convertible note offering is one of the most important pieces of the KEEL story. The company closed $458 million aggregate principal amount of 1.250% convertible senior notes due 2032, including the full exercise of the initial purchaser option. The notes pay interest semi-annually beginning in January 2027. The initial conversion price is approximately $7.41 per share, representing a premium to the reference share price used in the offering, while the capped-call cap price is $11.86 per share.

Management’s explanation is straightforward. Existing liquidity was expected to be sufficient to develop Panther Creek, Sharon and Moses Lake through leasing, while the new proceeds improve flexibility for value-add investments, long-lead equipment deposits, letters of credit and project acceleration. In other words, the financing is designed to give KEEL more strategic room during a period when speed and credibility may matter.

From a trader’s perspective, the convertible has three layers. The bullish layer is that Keel now has more capital flexibility. If customer demand is real, capital flexibility can matter because AI data center development requires deposits, engineering, equipment, utility coordination, and potentially rapid deployment. The neutral layer is that a low-coupon convertible is a common financing tool for high-growth, high-volatility companies. The bearish layer is that conversion price and capped-call levels become structural reference points for dilution analysis.

The financing is not automatically negative. It may be a smart move if it enables lease-backed development and asset value creation. But it makes the story more sensitive to execution. A company can carry convertibles comfortably if growth arrives. If growth disappoints, convertibles become a heavier overhang.

CEO Review: Benjamin Gagnon

Benjamin Gagnon is central to the KEEL story because this is not just a passive asset conversion. This is a strategic rebuild of a public company identity, capital structure, asset base and investor narrative. Gagnon was appointed CEO of Bitfarms in July 2024 after previously serving as Chief Mining Officer. Before that, he had joined Bitfarms as Director of Business Development in 2019, became Director of Mining Operations in 2020, and became Chief Mining Officer in 2021.

In those roles, he was responsible for mining strategy and operations, including growth strategy, mining economics, miner portfolio management and the evaluation and integration of new technologies. Before joining Bitfarms, he founded and operated two Bitcoin mining companies and held CEO, Chief Mining Officer and CTO roles. He holds an M.Sc. in Internet Computing from the University of Hong Kong and a B.Sc. in Economic Consulting and International Business from Indiana University’s Kelley School of Business.

That background is highly relevant for a company pivoting from Bitcoin mining infrastructure. Gagnon clearly understands mining economics, power strategy, operations and the technical-commercial logic of large-scale compute infrastructure. His experience is a strength for the legacy-to-HPC transition.

But there is a fair question: does Bitcoin mining leadership experience fully translate to enterprise-grade AI data center execution? The answer is partly yes and partly unproven. It helps with power procurement, site operations, energy economics and large-scale infrastructure management. It does not automatically prove hyperscale customer acquisition, AI data center design, colocation contracting, enterprise uptime requirements or long-term data center operating excellence.

On the positive side, Gagnon appears deeply aligned with the transformation and has been directly involved in the company’s evolution. He has overseen a major redomiciliation, rebrand, exit from Latin American megawatts, shift toward North American infrastructure, Stronghold integration path and large capital raises. Those are real corporate milestones.

The biggest success metric for the new KEEL, however, has not yet been fully delivered publicly: signed, credible, economically attractive HPC/AI customer leases. Until that happens, Gagnon’s record should be viewed as strong in mining-infrastructure transition and corporate repositioning, but still unproven in full AI data center commercialization.

Management Depth: Does The Team Have Relevant Experience?

The executive team disclosed in company filings includes Benjamin Gagnon as CEO and director, Jonathan Mir as CFO, Liam Wilson as COO, and Rachel Silverstein as Executive Vice President, General Counsel and Corporate Secretary. This mix matters because KEEL is not only a technology or mining story. It is a power, legal, construction, capital markets and infrastructure execution story.

Liam Wilson is particularly important because KEEL’s edge, if it exists, is deeply tied to power markets and site operations. The company has described Wilson as an operating executive with more than 20 years of leadership experience. Prior company communications said he had transacted in excess of 900 MW of power across the United States, including facilities in Pennsylvania, Ohio, Georgia, Texas and Nebraska, and described him as an expert in the deregulated U.S. energy industry who led a large-scale Bitcoin mining demand-response program in the PJM market.

That is exactly the kind of experience KEEL needs if its advantage is energy access and power-market execution. The PJM angle is especially relevant for Pennsylvania assets like Panther Creek and Sharon. However, the same caveat applies: power expertise is necessary but not sufficient. AI/HPC customers will evaluate full-stack data center credibility, including reliability, cooling, redundancy, operations, security and construction execution.

CFO Jonathan Mir’s role is also critical. KEEL is now a capital-markets story as much as an operating story. The company must manage cash, Bitcoin wind-down, convertible notes, development spend, equipment deposits, letters of credit, potential future project finance and investor communication while the income statement remains loss-making. That is a difficult job. The more KEEL moves toward AI/HPC leasing, the more investors will judge the CFO function on financing discipline and capital allocation, not just reported accounting.

Rachel Silverstein’s legal and corporate secretary role matters because redomiciliation, U.S. reporting, utility agreements, data center leases, customer contracts, environmental and land development matters, permits and capital markets compliance all become more important as the company moves from miner to infrastructure developer. In a business like this, legal execution is not a back-office detail. It is part of the business model.

Governance: Better Structured, But Still Worth Watching

Keel’s governance has changed materially since the old Bitfarms period. In August 2024, Bitfarms announced that CEO Ben Gagnon would join the board, Brian Howlett would become independent chairman, and co-founder Nicolas Bonta would step down. The company said the board would continue with five directors, four of whom were independent at that time.

The April 2026 post-redomiciliation filing shows a broader board and committee structure. Directors listed include Benjamin Gagnon, Edith M. Hofmeister, Brian Howlett, Fanny Philip, Amy L. Freedman, Andrew J. Chang and Wayne Duso. Committee assignments are disclosed for audit, compensation, and governance/nomination/safety/sustainability/technical functions.

This is not a governance-free microcap setup. There is a disclosed board, committee structure and U.S. domestic issuer reporting framework. That is positive. At the same time, governance remains a watch item because the company has gone through a major identity shift, acquisition integration, redomiciliation and capital-structure expansion in a short period.

One historical point matters: the leadership changes followed a period of shareholder pressure and board adjustments at Bitfarms. That does not automatically make current governance weak, but it explains why governance should remain part of the KEEL diligence process. A company in transition needs disciplined capital allocation, strong independent oversight, transparent communication and conservative execution claims.

The April 2026 filing also shows authorized capital of 1.5 billion common shares and 120 million preferred shares. That does not mean dilution will happen, but it confirms that capital structure flexibility exists. In a capital-intensive development story, governance quality and financing discipline are not secondary details. They are core risk factors.

Insider Ownership And Insider Activity

Insider alignment appears meaningful but should be interpreted carefully. Public filing summaries around the transition showed CEO Benjamin Gagnon with direct common share ownership, stock options and restricted stock units. Additional equity awards were also disclosed around the redomiciliation and rebrand period.

This matters because the CEO is not purely cash-compensated. He has equity exposure and incentive awards. However, investors should distinguish between open-market insider buying and compensation-linked equity grants. Equity grants can align incentives, but they are not the same signal as an insider using personal cash to buy shares in the open market.

The company’s Q1 filing also shows outstanding RSUs and PSUs. PSUs vest based on corporate objectives and can convert into shares based on performance outcomes. That means equity compensation is a real component of the capital structure. It can align management with execution, but it also contributes to potential share issuance over time.

Institutional Ownership: One Clean 13G Signal, But Still Not Full Thesis Validation

Institutional ownership is not yet perfectly clean to interpret because KEEL recently changed identity, ticker and domicile. Some ownership databases may still show incomplete or inconsistent data. That is normal during a transition from Bitfarms to Keel, but it means investors should be cautious when relying on aggregated ownership screens.

A clean primary-source institutional data point is Jane Street. On May 18, 2026, Jane Street Group, LLC and related entities filed a Schedule 13G reporting beneficial ownership of 30,542,386 shares of Keel Infrastructure common stock, representing 5.1% of the class. The filing attributes shared voting and dispositive power over these shares to Jane Street entities.

This is a meaningful disclosure, but it should not be interpreted as strategic validation of the AI/HPC thesis by itself. Jane Street is a large trading and market-making organization, and the filing does not explain the economic intent behind the position. It is a real institutional ownership signal, but not the same thing as an infrastructure specialist publicly endorsing the long-term data center thesis.

Third-party ownership aggregators may also show other institutional holders, including multi-strategy funds or AI-infrastructure-linked investors, but those numbers should be treated as secondary until checked against the underlying SEC filings. For KEEL, the most important institutional catalyst would not simply be a higher institutional ownership percentage. It would be higher-quality validation tied to concrete milestones: long-only ownership, infrastructure-specialist participation, customer-backed financing, project-level debt, credit-market validation or inclusion in relevant passive/index flows.

Passive Flow / Index Inclusion Watch

Because KEEL now has a U.S. domicile, Nasdaq listing, multi-billion-dollar market capitalization and meaningful trading attention, it may become more relevant for index and passive-flow screens than it was as a Canadian-domiciled Bitcoin miner. This is not a claim that index inclusion is imminent. It is a monitoring point.

The company’s market cap, float, trading volume, domicile, listing history and classification will matter. If the stock continues to trade with high liquidity and the company’s market value holds, passive and index-related attention could become part of the technical setup. This is especially relevant because KEEL trades inside a theme that ETFs and thematic funds may want exposure to: AI infrastructure, power, data centers, digital infrastructure and former crypto-mining transition plays.

Analyst Coverage

Keel’s investor relations page lists a broad analyst-coverage roster, including Alliance Global Partners, ATB Capital Markets, B. Riley Securities, Cantor Fitzgerald, Chardan, Clear Street, Compass Point, H.C. Wainwright, KBW and Northland Capital Markets. The company properly notes that analyst opinions, estimates and forecasts are those of the analysts and do not represent Keel’s own views.

The number of firms following the company is meaningful. It means KEEL is not completely undiscovered. It also means there is a sell-side framework forming around the pivot. However, for a stock like this, analyst price targets can move quickly and can lag the narrative. The real issue is not whether analysts like the story. It is whether their models begin to include credible lease economics.

The most useful analyst work will likely focus on lease timing by site, expected MW leased, pricing assumptions, capex per MW, customer quality, financing needs after the 2032 convertible, Bitcoin wind-down assumptions, power costs, PJM exposure and valuation comparisons with other miner-to-AI infrastructure pivots.

Retail Sentiment: Extremely Bullish, But Speculative

Retail sentiment around KEEL has become very active. Stocktwits-related coverage has described sentiment as extremely bullish during periods of high message volume, with traders focused on the possibility of a first data center lease announcement and on the broader AI infrastructure narrative.

This is important, but it must be framed correctly. Retail sentiment is not a fact source. It is not confirmation of a lease, hyperscaler customer, strategic deal or financing milestone. It is a measure of trader attention and narrative intensity. In KEEL’s case, retail traders appear to be focused on the same catalyst institutions are likely watching: a credible AI/HPC lease announcement.

The bullish retail narrative is typically built around four ideas. First, Keel has scarce power-backed assets at a time when AI data center capacity is constrained. Second, management appears to be positioning for a major strategic phase. Third, the convertible financing may be interpreted as evidence that the company can attract institutional capital. Fourth, if a hyperscaler or AI infrastructure customer signs a lease, the market could rapidly revalue KEEL from former miner to AI infrastructure developer.

The cautious retail counter-narrative is also clear. The stock has already moved strongly and may be pricing in good news before it arrives. Retail traders may over-read management language and conference appearances. AI infrastructure pivots across former miners have become crowded, and not every miner will become a credible data center platform. A lease announcement without attractive economics may not justify a much higher valuation.

Competitive Landscape

KEEL is competing in an increasingly crowded field. The company itself acknowledges competition from Bitcoin mining companies and HPC data center operators, including large and well-funded companies with significant resources and expertise. That competitive warning is important because the AI infrastructure trade is attractive precisely because demand is strong — and strong demand attracts capital.

Former Bitcoin Miners Pivoting To AI/HPC

The first competitive group includes former or current Bitcoin miners trying to repurpose power and data center assets for AI workloads. These companies often have experience with high-energy compute, power procurement and operational flexibility. The market is trying to separate real AI infrastructure platforms from companies that simply add AI/HPC language to a mining story.

KEEL’s differentiation is the combination of U.S. redomiciliation, the Stronghold/Pennsylvania energy angle and the claimed 2.2 GW pipeline. Its weakness is that it still needs public proof of customer conversion.

Traditional Data Center Operators

The second competitive group includes mature data center operators with more operating credibility, customer relationships, scale, uptime history and balance-sheet access. Keel is not competing on brand maturity. It is competing on access to power and potentially faster development in emerging markets.

Utilities, Power Developers And Private Infrastructure Platforms

The third group may be the most difficult to see from the public markets. Some of the best AI data center opportunities may never become public small-cap trades because private developers, utilities, infrastructure funds and hyperscalers can lock them up directly. Keel’s public-market opportunity is attractive, but the company is competing against sophisticated capital with deep project-finance expertise.

Catalysts To Watch

1. First Major AI/HPC Lease

This is the central catalyst. The market wants proof that Keel can sign a serious customer. A generic memorandum, vague customer language or non-binding indication would be less powerful than a named customer, binding lease, meaningful MW commitment, defined economics and clear construction or energization timeline.

2. Panther Creek Development Update

Panther Creek is the headline asset. Any update on utility work, substations, transmission, permits, construction timeline, customer discussions or financing could move sentiment. The utility agreement gives the site a factual backbone, but the market still needs commercial validation.

3. Sharon And Moses Lake Progress

If Keel can show that the pivot is not a one-site story, the market may assign a higher platform valuation. Updates on Sharon’s electricity delivery agreement and Moses Lake construction plans are important.

4. Use Of Convertible Proceeds

The June 2026 financing gives Keel flexibility. Investors now need to see whether that flexibility is used for accretive long-lead equipment, letters of credit, deposits or project acceleration — not just general overhead.

5. Bitcoin Wind-Down

The company has been selling Bitcoin as part of its wind-down. Continued reduction of Bitcoin exposure may make the story cleaner for AI infrastructure investors, but it also reduces optional upside from Bitcoin holdings.

6. Analyst Revisions And Conference Commentary

Investor events, analyst notes and management commentary can matter for a stock like KEEL because the story is still being modeled by the market. The most important commentary will be specific, not promotional: lease timing, customer pipeline quality, development schedule, financing discipline and expected economics.

Bull Case

In the bullish scenario, Panther Creek, Sharon and Moses Lake convert into credible HPC/AI leases. The market begins valuing Keel on future contracted megawatts rather than current Bitcoin-mining revenue. The 2.2 GW pipeline becomes valuable because investors believe management can repeat the conversion process across multiple sites.

The Stronghold acquisition becomes a strategic energy-infrastructure asset rather than a costly mining legacy. The June convertible becomes smart growth capital rather than dilution pressure. CEO Benjamin Gagnon is viewed as the right operator to lead the transformation, while Liam Wilson’s power-market background becomes a differentiator in PJM and other U.S. markets.

The strongest version of the bull case includes a named customer, investment-grade counterparty, attractive lease economics and financing that is either customer-backed or supported by clear project-level returns. In that case, the stock could trade less like a miner and more like an AI infrastructure development platform.

Bear Case

In the bearish scenario, lease execution is delayed, customer economics are weaker than expected, capex per MW is higher than expected, or the company must raise more capital before reaching durable cash flow. The convertible notes become an overhang rather than a confidence signal. Retail enthusiasm fades if no lease announcement arrives quickly.

The harshest bear case is not that AI data center demand disappears. It is that demand goes to better-capitalized, more experienced operators, while KEEL’s assets require more capital, more time and more operational upgrades than the market currently assumes.

The current reported financials also matter. Heavy losses, negative gross profit in the latest quarter, development spending and remaining legacy mining exposure all create risk if the AI infrastructure pivot takes longer than expected.

Base Case

The base case is a slower, uneven but still potentially valuable transition. Keel probably does not convert the entire 2.2 GW pipeline quickly. Some assets may move faster than others. Panther Creek may become the first major proof point, while Sharon and Moses Lake develop in stages.

In this scenario, the company uses its strengthened liquidity to keep projects moving, but expenses remain high and near-term earnings stay negative. The stock remains volatile and event-driven. It trades on lease rumors, conference language, analyst notes, financing interpretation, Bitcoin exposure and AI infrastructure momentum.

Long-term value creation depends on contract quality, not only megawatt headlines. A smaller but high-quality contracted platform would be more valuable than a larger pipeline with unclear timing and weak economics.

Red Flags

The first red flag is the size of current losses. A quarterly net loss of roughly $145 million is not a minor issue, even in a transition story. The company must eventually show that the infrastructure pivot can create durable, high-quality revenue.

The second red flag is execution complexity. Keel must manage power agreements, permits, construction, customer negotiations, equipment procurement, financing and operational transformation at the same time.

The third red flag is capital intensity. HPC data center development requires major investment before cash flow arrives. Long-lead equipment, utility work, land development and construction costs can be substantial.

The fourth red flag is competitive pressure. KEEL is competing with other miners, dedicated data center operators, private infrastructure platforms, utilities and hyperscalers with deeper resources.

The fifth red flag is dilution and capital structure. The new convertible financing strengthens liquidity, but the conversion price and capped-call ceiling create clear levels for dilution analysis.

The sixth red flag is retail overexcitement. Extremely bullish sentiment can support momentum, but it can also create fragile positioning if expected news does not arrive.

Merlintrader Bottom Line

KEEL is a serious watchlist name in the AI infrastructure theme, but it is not a clean low-risk infrastructure compounder. It is a high-beta transition story with meaningful assets, real liquidity, a credible power-oriented management angle and a powerful market narrative — but also heavy losses, capital intensity, dilution risk and unproven commercial conversion.

The core question is simple: can Keel turn megawatts into signed AI/HPC leases with credible customers and attractive economics?

If yes, the market may increasingly view KEEL as one of the more interesting public AI power infrastructure pivots. If no, the stock risks being exposed as a former Bitcoin miner whose valuation ran ahead of proof.

For now, KEEL is reportable, tradable and strategically important — but it must be monitored with discipline. The first major customer lease, especially around Panther Creek, Sharon or Moses Lake, is the catalyst that could separate a real infrastructure platform from a strong narrative.

Related Merlintrader Research

The KEEL thesis belongs inside the broader AI infrastructure and power scarcity trade. For readers following catalyst-driven research and market structure opportunities, Merlintrader also maintains a dedicated catalyst calendar and a broader research archive.

Visit the Merlintrader Free Catalyst Calendar

Read more Merlintrader market research