CleanSpark stock has the kind of catalyst that invites shortcut arithmetic. A $6.6 billion lease, $330 million of expected annual net operating income and a 20-year term look like instant proof that a Bitcoin miner has become an AI-infrastructure landlord. The signed contract matters. The shortcut does not.
The part shareholders still have to solve is construction. CleanSpark estimates landlord project costs of $10 million to $12 million for each megawatt of critical IT load. Across the 175 megawatts covered by the Sandersville, Georgia lease, that points to roughly $1.75 billion to $2.10 billion before the first scheduled deliveries in the fourth quarter of 2027.
That cost is not a footnote to the story. It is the story’s financing hinge. The contract can create a long-duration revenue stream only if CleanSpark raises and deploys the capital, meets the construction schedule and satisfies the other conditions summarized in its filing.
The market has already recognized both sides. TECHi’s market-data stack recorded CLSK at $13.635 at 11:04:07 a.m. Eastern on July 14, up 10.32% from the prior completed-session close of $12.36. The shares had traded as high as $15.10 after opening at $14.665, then surrendered part of that gap. That is a regular-session snapshot, not a final close.
Article Brief
Key Takeaways
5 Points30s Read
CleanSpark’s investor announcement describes a 20-year triple-net lease with two five-year extension options. The unnamed tenant is characterized as a leading global technology company with a high-investment-grade credit profile. CleanSpark says the customer will deploy production-grade infrastructure for a range of computing workloads at the Sandersville campus.
The initial lease covers 175 megawatts of critical IT load. Deliveries are expected to begin in the fourth quarter of 2027. CleanSpark estimates $6.6 billion of contract value during the first 20 years and as much as $11.6 billion if the tenant exercises both extensions.
The July 14 Form 8-K is more useful than the celebratory language around it. CleanSpark says it must satisfy specified financing, construction and delivery milestones, plus other covenants and conditions. Missing an applicable milestone on time may produce rent abatements or termination of the lease.
That disclosure changes how the headline should be read. This is a definitive lease, not a memorandum of understanding. It is also a conditional execution program, not $6.6 billion sitting in a bank account. The public filing does not attach the full lease, identify the tenant, disclose the financing package or lay out the milestone calendar.
This is a more advanced stage than the partnerships that often surround AI infrastructure. TECHi recently found that SLB’s data-center alliance still lacked a disclosed order. CleanSpark has crossed that line. The remaining question is whether the contract terms and the capital structure can meet in the middle.
The company’s own numbers produce a useful underwriting frame. Dividing $6.6 billion by 20 years gives $330 million a year, the same average annual NOI contribution CleanSpark cites in Exhibit 99.1. That equals about $1.89 million per megawatt per year, or roughly $157,000 per megawatt per month.
Against estimated project costs of $1.75 billion to $2.10 billion, $330 million of annual NOI implies a simple unlevered NOI-to-cost ratio of about 18.9% at the low end of cost and 15.7% at the high end. The corresponding simple payback period is roughly 5.3 to 6.4 years.
Those figures are TECHi calculations, not company return guidance. They exclude the time value of money, construction timing, financing expense, taxes, possible overruns, the ramp from initial delivery to full service and any rent abatements. They also assume the project reaches the annual economics described in the release.
The distinction matters because NOI is not free cash flow. A triple-net lease can shift property taxes, insurance, maintenance and other operating obligations to the tenant, supporting a high NOI margin after the facility is operating. It does not make the landlord’s construction bill disappear. CleanSpark itself put the $10 million to $12 million per megawatt estimate in the transaction summary.
The two extensions need their own discount. The headline rises from $6.6 billion to $11.6 billion if both five-year options are exercised. That additional $5.0 billion over ten years averages $500 million a year, presumably reflecting annual escalators and the later years of the contract. Yet those years are tenant options, not guaranteed initial-term revenue.
This is where the current coverage differs from the most optimistic read-through. A large contract value proves demand and improves financing credibility. It does not automatically tell investors the equity return after debt, dilution, construction risk and time.
CleanSpark entered this deal with a larger balance sheet than a small project developer, but the proposed construction cost is still large relative to its liquid resources.
The company’s Form 10-Q for the quarter ended March 31 reported $260.3 million of cash, $813.2 million of current and noncurrent Bitcoin, $1.10 billion of current assets and $133.1 million of current liabilities. It also carried $1.79 billion of long-term debt, net of the current portion, discounts and issuance costs. Gross debt outstanding was about $1.82 billion.
The midpoint of the Sandersville cost estimate is about $1.93 billion. That is more than seven times the March cash balance and roughly 1.75 times total current assets. Bitcoin provides liquidity, but using it for construction would reduce the treasury exposure that many CLSK shareholders originally bought. New debt can preserve Bitcoin holdings, but it adds fixed claims before the data-center rent is fully flowing. Equity financing protects near-term liquidity while diluting existing owners.
The tenant’s credit profile could materially improve the financing package. CleanSpark explicitly says the confidential tenant’s standing supports financing options and a multi-decade term. A bankable long-term lease with an investment-grade customer is precisely the kind of asset lenders can underwrite.
Investors still need the actual package. The latest 10-Q says CleanSpark may require additional capital and could use debt or equity financing to respond to business opportunities and technological change. The filing also shows why clean accounting language matters: the March quarter included a $378.3 million net loss heavily affected by Bitcoin fair-value movements, while cash, debt and construction commitments tell a different part of the risk story.
This balance-sheet test resembles the question TECHi raised around CoreWeave’s enormous backlog and debt load and Dell’s AI server boom becoming a credit business. AI demand can be genuine while cash conversion remains the harder investment problem.
A single high-grade tenant can simplify underwriting. One customer, one long lease and one set of technical requirements can produce better visibility than speculative multi-tenant capacity. The customer also has enough confidence to place CleanSpark’s Texas portfolio under exclusivity.
The same structure creates concentration. The tenant’s identity is confidential, and the public filing does not provide a named rating, guarantor, termination fee or parent-company support. “High investment grade” is a description from CleanSpark, not a substitute for seeing the legal counterparty and guarantee.
Milestone language is especially important because the first deliveries are not expected until late 2027. CleanSpark must fund and build through a period when AI hardware, cooling density and power requirements can change quickly. The 8-K acknowledges construction, power availability, regulatory approvals and changing AI or high-performance-computing needs among the risks.
The signed lease therefore de-risks demand more than execution. It tells investors there is a customer willing to contract for the capacity. It does not yet prove the facility will arrive on time, at the estimated cost or without abatements.
The wider power constraint is real. TECHi has tracked how AI data centers are turning electricity access into a stock-market variable and how Meta’s Louisiana campus is outrunning its power plan. Sandersville gives CleanSpark an asset that hyperscalers want: an existing campus with power history and room for high-density compute. That advantage is valuable precisely because delivery is difficult.
The Georgia lease came with a second headline: an executed letter of intent and exclusivity arrangement covering CleanSpark’s entire Texas portfolio. That portfolio spans 718 acres and up to 885 megawatts of secured and planned power capacity.
CleanSpark breaks it into 271 acres with nearly 300 megawatts at Sealy and 447 acres at Brazoria. The company says Brazoria’s transmission infrastructure can support an initial 300-megawatt demand load and potentially expand to 600 megawatts.
The scale is large enough to transform the company if it becomes definitive business. It should not be combined with the Sandersville lease in a base-case revenue total today. A letter of intent and exclusivity arrangement is not the same legal stage as the signed 175-megawatt lease. The filing provides no Texas rent, construction budget, delivery date or contract value.
Exclusivity can be a positive signal because a sophisticated customer is reserving the pipeline. It can also limit CleanSpark’s ability to market those sites elsewhere while negotiations continue. Without the duration and conditions of the exclusivity period, the safest treatment is option value.
That discipline matters across the sector. Bitcoin miners have been pitching AI pivots as mining economics tightened. CleanSpark now has more commercial proof than most of that group. The Georgia agreement deserves to be valued as a signed lease subject to execution. Texas deserves to be watched as a possible sequel, not counted as if it were already Georgia.
The stock’s double-digit gain shows that a definitive customer contract resolved one major uncertainty. The retreat from the session high shows that investors are not treating the headline value as immediate equity value.
The next disclosures should be judged against a concrete list:
The most important near-term number is not another expansion headline. It is the amount and price of capital CleanSpark must commit before the lease begins producing cash.
CleanSpark’s lease is a meaningful de-risking event. The company moved from talking about an AI and high-performance-computing pivot to signing a long-duration contract for 175 megawatts with a creditworthy technology customer. That is real progress, and it separates CLSK from miners whose AI strategy remains a land-and-power presentation.
The $6.6 billion figure still needs to be valued through the $1.75 billion to $2.10 billion construction requirement, the late-2027 delivery window and the milestone clauses disclosed in the 8-K. On the information available today, the deal validates demand more clearly than it validates the eventual return to common shareholders.
CleanSpark stock now has two businesses with very different clocks. Bitcoin mining produces current revenue and volatility. The AI lease offers longer-duration contracted economics, but only after a capital-intensive build. The investment case becomes stronger when management shows how those clocks will be financed without letting one consume the other.
SLB has spent a century learning how to move complicated industrial systems from drawings to…
The Esports World Cup 2026 has just begun in Paris and is expected to see…
The Esports World Cup 2026 has just begun in Paris and is expected to see…
Article BriefKey Takeaways5 Points30s Read01The setup-Four companies — ASML, Applied Materials, Lam Research and KLA…
Article BriefKey Takeaways4 Points24s Read01The launch-AWS added a guided SageMaker Studio workflow for generative AI…
Article BriefKey takeaways4 Points24s Read01Map exposure-Inventory every management interface, dependency, owner, and reachable source before…