Mara Holdings, Inc. · NASDAQ:MARA

Mara — Q1 2026 Earnings Presentation

32 slides from Mara's earnings presentation, May 11, 2026

  1. Slide 1: Investor Presentation

    MARCH 2026 INVESTOR PRESENTATION NASDAQ: MARA MARA.COM

  2. Slide 2: Disclosure

    Disclosure Investor Notice Forward-Looking Statements Investing in our securities involves a high This presentation contains forward-looking statements within the meaning of the federal securities degree of risk. Before making an investment laws. All statements, other than statements of historical f

  3. Slide 3: MARA's Evolution: From Bitcoin Mining to AI Infrastructure

    MARA’s Evolution: From Bitcoin Mining to AI Infrastructure to . . . Compute Capacity… 70 EH/s …to Energy Ownership 60 EH/s 50 EH/s 40 EH/s 30 EH/s 50 MW 139 MW 20 EH/s Flare Gas-to-Power Wind Power 10 EH/s 0 EH/s …to Optionality Bitcoin Mining ● Rented Capacity ● Owned Capacity Our asset-light strat

  4. Slide 4: Digital Energy Arbitrage

    … Allocating Energy to its Highest-Value Use With energy no longer the constraint, scale is measured in megawatts, not hashrate Digital Energy Arbitrage Power AI & Bitcoin Grid Generation Mining Sales 4

  5. Slide 5: From Renting Compute to Owning the AI Value Stack

    From Renting Compute to Owning the AI Value Stack PHASE 4 AI/HPC/Inference PHASE 3 Energy PHASE 2 Owned Infrastructure PHASE 1 Rented Capacity 5

  6. Slide 6: MARA's Deliberate, Early Positioning in AI Inference

    MARA’s Deliberate, Early Positioning in AI Inference October 2024 July 2025 February 2026 “The area we believe is most interesting is AI at “At MARA, we’ve anticipated these shifts and are “The AI industry has already commoditized around the edge, which is all about inference.…this is a building sol

  7. Slide 7: From Bitcoin Mining to AI Infrastructure, Energy → Intelligence

    From Bitcoin Mining to AI Infrastructure, Energy → Intelligence We’re applying what we learned optimizing cost-per-coin to optimize cost-per-token—starting with inference at the edge MARA: Energy Ownership & Site Control Power-secured, strategically located infrastructure with long-duration visibili

  8. Slide 8: Deploying MARA's Energy Advantage Across AI

    Deploying MARA’s Energy Advantage Across AI ENERGY STARWOOD EXAION Power & Control Scale & Stability Precision & Capability • Directing Resources • Hyperscale Conversions • Private Cloud & Edge Inference • Optionality & Flexibility • Infrastructure & Growth • Targeted Deployments 8

  9. Slide 9: Compute is Revenue

    Compute is Revenue In the AI economy, token generation drives enterprise and hyperscale revenue growth… Compute requires power. Power is scarce. MARA controls the power. Our sites were built to mine Bitcoin efficiently, but they were built around power. As AI/HPC demand accelerates, our economics ar

  10. Slide 10: The AI 5-Layer Cake — Where MARA, Exaion, and Starwood Fit

    The AI 5-Layer Cake — Where MARA, Exaion, and Starwood Fit APPLICATIONS Exaion Enterprise Inference AI Products & Services & Use Cases MODELS Exaion Model Partners & Foundation & Training Ecosystem CHIPS Hardware Agnostic GPUs, TPUs, ASICs Access MARA + Starwood Power + INFRASTRUCTURE Sites (Core Ad

  11. Slide 11: MARA's AI Platform: Energy at the Core, Revenue Up the Stack

    MARA’s AI Platform: Energy at the Core, Revenue Up the Stack Three complementary capabilities. One integrated AI infrastructure platform. Hyperscale Scaling Engine Sovereign Enterprise Transforms MARA’s energy- Positions MARA in secure, data- advantaged footprint into residency-sensitive markets hyp

  12. Slide 12: Exaion: The Sovereign & Enterprise Layer of MARA's AI Strategy

    Exaion: The Sovereign & Enterprise Layer of MARA’s AI Strategy Exaion extends MARA into sovereign & regulated AI markets where data residency, compliance, and infrastructure control are critical. Who is Exaion? Sovereign-Grade Architecture Exaion extends MARA into jurisdiction- Exaion delivers integ

  13. Slide 13: The Trusted, Sovereign Cloud Layer of MARA's AI Platform

    The Trusted, Sovereign Cloud Layer of MARA’s AI Platform Stable Long- Alternative Public Term Revenue EU Sovereignty Full Stack AI Platform HPC Cloud Tailwinds & Relationships EUROPEAN COMPLIANCE FOOTPRINT ADVANTAGE Capital Efficient Contracted Multi- Expansion Year Revenue REVENUE COMPUTE DURABILIT

  14. Slide 14: Starwood Joint Venture: Power Certainty into Capacity Certainty

    Starwood Joint Venture: Transformational Partnership Converts Power Certainty into Capacity Certainty Together, MARA and Starwood will accelerate the conversion and expansion of select MARA sites into next generation AI, enterprise, and high-performance computing digital infrastructure POWER FLEXIBI

  15. Slide 15: Joint Platform to Deliver AI Capacity at Scale

    Joint Platform to Deliver AI Capacity at Scale Under Favorable Timelines • Leading energy and compute infrastructure • Data center development platform of Starwood company Capital Group • 16 operating campuses across the U.S. and UAE • Institutional capital base and disciplined • Energy-backed sites

  16. Slide 16: A Capital Efficient Buildout of High-Density Compute

    A Capital Efficient Buildout of High-Density Compute Across MARA’s Portfolio • MARA contributes a dedicated data center while Starwood leads EPC MARA OPTION TO INVEST UP execution, secures customers and operates the asset TO 50% IN JV PROJECTS • Joint ownership of all projects • Shared upside and di

  17. Slide 17: Operations Roadmap

    Operations Roadmap STRONG EARLY LEVERAGING MARA’S INITIAL WORK UNDERWAY CUSTOMER DEMAND POWER-RICH SITES Already engaged in Design, permitting, and Joint platform is expected to discussions with prospective commercial leasing processes support more than 1 gigawatt AI and high-performance are underwa

  18. Slide 18: Capacity Demand in Gigawatts

    The Intersection of Energy and Compute Data center demand is accelerating The partnership directly targets the accelerating demand Data Center Power Demand From AI & for data centers, driven by factors including AI Non-AI Workloads Worldwide 250 • 255% increase in AI power demand expected from Capac

  19. Slide 19: Why MARA's Energy-Backed Compute Wins

    Why MARA’s Energy-Backed Compute Wins MARA’s power-rich sites give customers what they need most: Starwood’s development engine adds the strong • Predictable access to energy at scale execution and operating • Adjacent to generation and high-voltage infrastructure capabilities and deep • Existing op

  20. Slide 20: Partnership Strengthens Our Position As a Leader in Digital Infrastructure

    Partnership Strengthens Our Position As a Leader in Digital Infrastructure Accelerates MARA’s transition into a digital energy and compute infrastructure company Expands vertically integrated model unifying energy, Bitcoin mining and AI compute Brings together experienced operators and institutional

  21. Slide 21: Advancing MARA's Strategy Across Energy-Backed Infrastructure and Diversified Compute

    Advancing MARA’s Strategy Across Energy-Backed Infrastructure and Diversified Compute To Create Durable Value MARA’S FLEXIBLE PLATFORM: TURNING ENERGY INTO VALUE ENTERPRISE, HYPERSCALE, PRIVATE CLOUD AND AND HPC AS A ENERGY OWNERSHIP SOVEREIGN-GRADE VALUE CREATION ENGINE CAPABILITY The Starwood plat

  22. Slide 22: Q4 2025 Financial & Operational Overview

    Q4 2025 Financial & Operational Overview Revenues Energized Hashrate1 BTC Holdings Decreased 6% YoY to $202M Up 25% YoY to 66.4 EH/S Increased 20% to 53,822* from 44,893 at from 53.2 EH/S the end of Q4 2024 Net Income (Loss) Cost per Petahash / Day BTC Production ($1.7B), or ($4.52)/share, a decreas

  23. Slide 23: Trending BTC Production & Holdings

    Trending BTC Production & Holdings* 20% 31% 33% 30% 28% 23% 53,822 52,850 49,951 47,531 5,938 Pledged as Collateral 44,893 9,377 Loaned 38,507 Unrestricted 2,492 2,286 2,358 2,144 2,011 703 666 694 633 595 Q4 2024 Q1 2025 Q2 2025 Q3 2025 Q4 2025 BTC Production BTC Blocks Won Total BTC Holdings (inc.

  24. Slide 24: Appendix

    Appendix

  25. Slide 25: Q4 2025 Financial Performance

    Q4 2025 Financial Performance $ in millions 1 Data as of December 31, 2025 *Includes bitcoin that is loaned, actively managed, or pledged as collateral 1. Adjusted EBITDA is a non-GAAP financial measure. Please refer to the Appendix for the definition of Adjusted EBITDA and a reconciliation to the m

  26. Slide 26: Cost per Petahash per Day Improved 36%

    Over the Past 11 Quarters, Cost per Petahash per Day Improved 36% Daily Cost per Petahash* $47.9 $46.1 $39.7 $38.1 $37.8 $37.0 $31.7 $31.3 $30.5 $28.5 $28.7 Q2 2023 Q3 2023 Q4 2023 Q1 2024 Q2 2024 Q3 2024 Q4 2024 Q1 2025 Q2 2025 Q3 2025 Q4 2025 Data as of December 31, 2025 *Daily cost per petahash q

  27. Slide 27: Building a Scalable Energy Network Through Strategic Site Expansion

    Building a Scalable Energy Network Through Strategic Site Expansion 1.3 GW 1.9 GW Operational Capacity Total Capacity Central Texas 411 MW 550 MW North Texas 78 MW 180 MW Central Ohio 28 MW 150 MW East Ohio 79 MW 225 MW Central Nebraska 127 MW 142 MW International 57 MW 77 MW Central North Dakota 27

  28. Slide 28: Quarterly Financials: Q4 2025 vs. Q4 2024

    Quarterly Financials: Q4 2025 vs. Q4 2024 SUMMARY BALANCE SHEET (UNAUDITED) (in millions) December 31, 2025 December 31, 2024 CASH & CASH EQUIVALENTS $ 547.1 $ 391.8 RESTRICTED CASH 12.0 12.0 DIGITAL ASSETS, CURRENT PORTION 2.4 4.3 OTHER RECEIVABLES 18.5 6.3 DEPOSITS 19.5 18.8 DERIVATIVE INSTRUMENT,

  29. Slide 29: Non-GAAP Reconciliation

    Non-GAAP Reconciliation (in thousands) UNAUDITED Q4 2025 Q4 2024 RECONCILIATION TO ADJUSTED EBITDA NET INCOME (LOSS) ATTRIBUTABLE TO COMMON STOCKHOLDERS $ (1,709,644) $ 528,528 INTEREST EXPENSE (INCOME), NET (3,695) 93 INCOME TAX EXPENSE (BENEFIT) (183,386) 118,262 DEPRECIATION AND AMORTIZATION 289,

  30. Slide 30: Management Team

    Management Team Fred Thiel Salman Khan Zabi Nowaid CHAIRMAN & CEO CHIEF FINANCIAL OFFICER GENERAL COUNSEL • Proven tech entrepreneur with a track record as • Decades in public companies, Big 4 accounting, • Former GC and Corporate Secretary at Open CEO of three publicly traded companies and consulti

  31. Slide 31: Seasoned Board of Directors

    Seasoned Board of Directors Fred Thiel Doug Mellinger CHAIRMAN & CEO LEAD INDEPENDENT DIRECTOR • Proven tech entrepreneur with a track record as CEO of three publicly • 40+ years of experience leading in FinTech traded companies • Focused on strengthening governance and strategy • Drives company val

  32. Slide 32: Thank You

    Thank You CONTACT US AT IR@MARA.COM

MARA Holdings — Q1 2026 Earnings Presentation on 6ix