Hook & thesis
MARA is no longer just a Bitcoin miner. Two recent moves - the acquisition of over 1,200 acres of powered land in Texas (up to $600M) announced 07/10/2026 and the Long Ridge Energy & Power deal announced 04/30/2026 - create a combined platform that takes the company from a mining operator to a provider of rare, powered real estate and utility-grade capacity. When combined with already-announced sites, management now points to roughly 4.8 GW of potential capacity. That is the asset the market is underpricing.
This is a trade on asset reconfiguration and valuation re-rating rather than a pure call on Bitcoin price. If MARA executes on converting land and owned-generation into leased AI/HPC or colocation capacity, the path to substantially higher revenue per MW and recurring contracted cashflow is clear. At a market cap of roughly $4.64 billion and enterprise value near $6.55 billion, the stock is priced like a cyclical miner instead of a hybrid infrastructure play - that disconnect creates the opportunity.
What the company does and why investors should care
MARA Holdings historically mined Bitcoin at scale and sold software and advisory services in the crypto ecosystem. Over the last year the company has pivoted toward developing digital infrastructure campuses for AI and high-performance computing, using a mix of owned generation, leased power, and large acreages of powered land.
Why this matters: powered land with grid access - particularly parcels that can realistically host multi-hundred-megawatt campuses - has become a scarce input for hyperscalers and large AI customers. Owning a combined-cycle gas plant (Long Ridge - 505 MW) plus Texas sites that can deliver 2 GW by 2028 materially short-circuits one of the biggest friction points for AI deployment: guaranteed, dispatchable power at scale. That gives Marathon optionality to lease capacity into higher-margin AI/HPC markets rather than being exposed largely to Bitcoin price swings.
Key facts and recent financial backdrop
- Current price: $12.61 (previous close $12.60).
- Market capitalization: approximately $4.64 billion.
- Enterprise value: approx. $6.55 billion, with EV/sales ~ 7.55x and price/sales ~ 5.36x.
- Cash on the balance sheet: ~$1.55 billion. Debt-to-equity: ~1.08.
- Recent quarter (Q1 2026): revenue of $174.6 million, down 18% year-over-year; a large digital asset write-down produced a net loss of ~$1.3 billion and an adjusted EBITDA loss near $1.0 billion.
- Shares outstanding: ~381.3 million; float ~372.6 million. Short interest as of 06/30/2026: ~106.9 million shares (~28.7% of the float) with ~2.33 days to cover.
Why the market should re-rate Marathon
Think of Marathon’s new footprint as a combination of three valuable ingredients: (1) owned generation (Long Ridge - 505 MW combined-cycle gas plant), (2) powered land in a high-capacity Texas grid node (2 GW potential by April 2028 from the HIF USA site), and (3) operating expertise in energy-intensive compute. Each of these assets has standalone value to AI customers and hyperscalers.
At $12.61 the market is implicitly valuing Marathon as if it will remain a cyclical miner with lumpy commodity-driven revenue and periodic write-downs. But the mix changes if even a fraction of the 4.8 GW is contracted to long-term AI/HPC tenants at materially higher revenue per MW than Bitcoin mining economics. A single 100 MW contracted campus with multi-year pricing and strong uptime obligations would meaningfully lift recurring adjusted EBITDA and reduce earnings volatility.
Valuation framing
At current market cap of ~$4.64B and EV ~$6.55B, MARA trades at EV/sales ~7.55x. That multiple looks rich by commodity-mining standards but cheap relative to the implied replacement value of multi-GW powered campuses plus a combined-cycle plant. Consider the following points:
- Cash of ~$1.55B cushions balance-sheet execution risk for development and regulatory work.
- Owned generation (505 MW) and the Texas sites reduce the time-and-money friction customers face when securing power for AI builds; that premium could justify materially higher valuation multiples if Marathon secures long-term contracts.
- Price/book is ~2.08x, but book value understates the scarcity of deliverable, permitted powered land adjacent to robust transmission.
In short, the headline EV/sales multiple does not capture the replacement cost of controlled, permitted, powered acreage plus generation plant economics and the upside from contracted, recurring revenue.
Catalysts (what to watch)
- Regulatory approvals and closing of the Long Ridge deal - expected to close in H2 2026 pending HSR and FERC approvals. A close would de-risk the transaction and should be a re-rating catalyst.
- Progress updates on the Texas powered land development timeline and firming of interconnection rights toward the stated 2 GW by 04/2028.
- Announcements of long-term AI/HPC or hyperscaler leases or power-purchase/hosted agreements - these are direct evidence of structural revenue and margin change.
- Quarterly operating cadence showing less volatile adjusted EBITDA and fewer digital-asset mark-to-market losses, signaling the business mix is changing.
Trade plan (actionable)
Trade direction: Long
Entry price: $12.50
Stop loss: $10.25
Target price: $20.00
Horizon: long term (180 trading days). Rationale: the thesis depends on regulatory clearances (Long Ridge) and site development that unfold over multiple quarters. Allowing up to ~180 trading days gives time for approvals, initial permitting, and at least one quarter of operational updates or lease announcements that would materially rerate expectations.
Position sizing note: treat this as a thematic, event-driven position. Limit allocation so a failed pivot or another large crypto write-down does not disproportionately damage the portfolio.
Why this is compelling
The entry around $12.50 buys a company with substantial cash (~$1.55B), an EV that includes a 505 MW gas plant and acres of powered land, and visible optionality to pivot into AI/HPC. If Marathon secures even modest contracted ARR (annual recurring revenue) on a portion of the 4.8 GW footprint, the company’s earnings profile will look less like a cyclical miner and more like a capital-efficient infrastructure operator - a structural re-rating is plausible.
Risks and counterarguments
- Execution and permitting risk: converting powered land to ready-to-build campuses is capital intensive and regulatory-heavy. Delays or higher-than-expected capex would push timelines beyond the trade horizon.
- Commodity and mark-to-market volatility: Marathon still has exposure to digital assets. Further Bitcoin weakness could force additional write-downs and earnings volatility, compressing the stock regardless of the infrastructure story.
- Financing and dilution risk: development of multiple GW will likely require capital. Even with ~$1.55B cash, the company could raise equity or debt and dilute current shareholders if project economics don’t support non-dilutive funding.
- Short-seller pressure and sentiment: short interest remains high (~106.9M shares as of 06/30/2026, roughly 28.7% of float). That can amplify downside on bad news and increase volatility around earnings or regulatory updates.
- Counterargument: The market could be right - the company’s historic cash generation is poor (free cash flow ~- $1.28B), and the pivot may be a capital-intensive chase into a crowded space where Marathon has less operating leverage than established data-center operators. If Marathon fails to secure anchor tenants, the acreage and generation may not be enough to justify higher multiples.
What would change my mind
I would reduce conviction if any of the following occur: (1) the Long Ridge transaction fails to close or faces unexpected regulatory conditions that curtail plant availability; (2) management is forced to raise significant equity at distressed prices; (3) Q3/Q4 operational updates show no progress in contracting or interconnection for the Texas site; or (4) another large-scale digital asset impairment drives the balance sheet into a cash-constrained state despite the current ~$1.55B cash balance.
Conclusion
Buying MARA at ~ $12.50 is a way to back a tangible real-asset pivot into AI/HPC capacity at what looks like a discount to replacement value. This trade is not a pure call on Bitcoin - it is a bet that Marathon can successfully monetize powered land and owned generation into higher-margin, contracted recurring revenue over the next several quarters. The path is noisy and execution risk is material, so position sizing and a firm stop at $10.25 are essential. A successful close of Long Ridge and early lease announcements would be the clearing events that likely push the stock toward the $20 target within a 180 trading-day window.
Trade idea summary: Long MARA at $12.50, stop $10.25, target $20.00, horizon long term (180 trading days). Catalyst-driven, balance-sheet-backed, but execution-sensitive.