MARA
MARA Holdings, Inc.Thesis. MARA is a levered, 5.4-beta proxy on Bitcoin plus a nascent AI-infrastructure pivot (Exaion, $1.5B acquisition) that adds a call-option leg. But GAAP earnings swing violently with BTC mark-to-market ($808M profit in Q2'25 to $1.26B loss in Q1'26), net debt of $1.95B, and consistent open-market insider selling cap the near-term upside despite a 51% analyst target premium.
Scoreboard
At $12.22, MARA trades 47.9% below its 52-week high of $23.45 but 83.4% above the $6.66 low, sitting in the 33rd percentile of its annual range. Market cap is $4.66B against an enterprise value of $6.99B — the $2.33B gap reflects net debt of $1.95B. Consensus target of $18.54 implies 51.8% upside, but the wide $7-$30 dispersion across 11 analysts underscores the binary, BTC-driven nature of the story.
Crypto Market Context
- 7d ETF outflow $-705M — sustained institutional supply
BTC sits at $63,701 (flat 24h) with mildly positive funding of 0.0077% per 8h (~8.4% APR) and a near-balanced 1.02 long/short ratio — a neutral derivatives regime with OI down 2.3%. But 7d ETF outflows of -$705M and 30d of -$5.84B signal persistent institutional supply, which caps hashprice and directly pressures MARA's mining economics. Until ETF flows turn positive, MARA's beta cuts against it.
QoQ Changes
Q1'26 revenue of $174.6M fell 13.7% QoQ from $202.3M and is down sequentially for two straight quarters from the $252.4M Q3'25 peak. Basic EPS of -$3.31 missed consensus by 50.7%, though the loss is driven almost entirely by BTC mark-to-market swings, not operations.
Gross margin held at a healthy ~74% at the quarter level ($129.9M on $174.6M), but operating margin cratered to -148% on $258M of operating loss. Reported EBITDA of -$1.09B reflects digital-asset revaluation, not cash burn — the trailing EBITDA margin of -85.9% is optically distorted by BTC accounting.
FCF margin is deeply negative at -61.2% and FCF yield sits at -11.4%, consistent with a capital-intensive miner scaling AI compute. Operating cash generation remains hostage to hashprice and BTC spot.
Cash of $513.7M ($1.35/share) against total debt of $2.46B yields net debt of $1.95B and debt/equity of 105.6%. Liquidity is adequate near-term — current ratio 1.84, quick ratio 1.63 — but the balance sheet is leveraged to convertible/BTC-backed financing.
Shares trade at 5.4x P/S, 8.1x EV/revenue, and 2.08x book ($5.86/share). Forward P/E of -18x and EV/EBITDA of -9.4x are meaningless given GAAP volatility; PEG of 0.1 is an artifact.
MARA is aggressively pivoting from pure BTC mining to AI/HPC infrastructure — the ~$1.5B Exaion acquisition (French government-linked) and rebranding from Marathon Digital anchor an EU AI-cloud strategy at sector-leading $0.04/kWh owned-site power. Headcount is lean at 266, generating $3.26M revenue/employee.
Ownership & Insider Activity
Insiders are consistent net sellers: Form 4 data shows $2.87M in net open-market sales and zero purchases over the visible window, with CEO Thiel selling 27,505 shares monthly (Apr-Jun 2026 at $11.68-$14.25) plus large tax-withholding dispositions. Short interest is extreme at 30.8% of float (102.6M shares, up from 99.3M prior month), setting up meaningful squeeze potential on any BTC rally — the June 22 +11% surge was explicitly attributed to short-covering dynamics. BlackRock (15.7%) and Vanguard anchor the 69.6% institutional base.
Recent Insider Transactions
| Date | Insider | Position | Shares | Value |
|---|---|---|---|---|
| 2026-06-22 | MELLINGER DOUGLAS K | Director | 7,000 | $112.0K |
| 2026-06-17 | KHAN SALMAN HASSAN | Chief Financial Officer | 16,000 | $228.0K |
| 2026-06-17 | NOWAID ZABI | General Counsel | 7,000 | $99.8K |
| 2026-06-17 | THIEL FREDERICK G | Chief Executive Officer | 27,505 | $391.9K |
| 2026-05-18 | KHAN SALMAN HASSAN | Chief Financial Officer | 16,000 | $192.0K |
| 2026-05-18 | NOWAID ZABI | General Counsel | 8,250 | $99.0K |
Earnings Quality
| Period | Actual EPS | Estimate | Surprise | Surprise % |
|---|---|---|---|---|
| 2026-03-31 Q1 | $-3.31 | $-2.20 | $-1.11 | -50.71% |
| 2025-12-31 Q4 | $-4.52 | $-0.13 | $-4.39 | -3425.74% |
| 2025-09-30 Q3 | $0.31 | $0.45 | $-0.14 | -31.51% |
| 2025-06-30 Q2 | $1.89 | $0.63 | +$1.26 | +200.29% |
MARA beat only 1 of the last 4 quarters (25% beat rate) with a distorted average surprise of -827% — the -3,426% Q4'25 miss reflects BTC mark-to-market, not operational failure, but it signals analysts cannot reliably model this name.
Surprises are erratic rather than trending — from +200% (Q2'25) to -3,426% (Q4'25) to -50.7% (Q1'26) — confirming that reported EPS is a BTC-price derivative, not a management-controlled variable.
Analyst Action
| Month | Distribution | Strong Buy | Buy | Hold | Sell | Strong Sell |
|---|---|---|---|---|---|---|
| 2026-07 | 4 | 10 | 7 | 1 | 0 | |
| 2026-06 | 4 | 9 | 7 | 1 | 0 | |
| 2026-05 | 4 | 9 | 8 | 1 | 0 | |
| 2026-04 | 4 | 10 | 7 | 1 | 0 |
The rating composition is stable-to-slightly-bullish: Strong Buy held at 4 while Buy ticked from 9 (June) to 10 (July) and Hold eased from 8 (May) to 7, with a lone Sell persisting. Citizens initiated at Market Outperform with a $24 target on June 24.
Momentum is neutral-to-mildly-positive — no upgrades or downgrades in the last 15 days but a modest drift toward Buy over the quarter.
Seven Essential Metrics
ROE -67.3%, ROA -15.6%, EBITDA margin -85.9% — all GAAP figures whipsawed by BTC revaluation.
Revenue -18.4% YoY and down two straight quarters to $174.6M in Q1'26.
FCF margin -61.2% and FCF yield -11.4% amid heavy mining/AI capex.
Debt/equity 105.6% with net debt of $1.95B against $513.7M cash.
Beta of 5.37, 30.8% short float, and earnings fully hostage to BTC spot.
5.4x P/S and 2.08x book — reasonable for the crypto-miner cohort but earnings-based multiples are non-meaningful.
Miner funds growth via convertibles and equity/ATM issuance; no buyback offsets share creep.
All capital reinvested into hash capacity and AI infrastructure.
Competitive Snapshot
| Company | EBITDA Margin | 3Y Rev CAGR | FCF Margin | Leverage | Fwd P/E |
|---|---|---|---|---|---|
RIOT Riot Platforms | ~-20% | ~25% | ~-40% | Net cash | n/m |
CLSK CleanSpark | ~30% | ~40% | ~-30% | <1x | n/m |
CIFR Cipher Mining | ~10% | ~60% | ~-50% | Low | n/m |
MSTR MicroStrategy | n/m | ~-5% | ~-10% | High | n/m |
MARA is the largest listed pure-play BTC miner by hashrate and, like RIOT and CLSK, is racing to reframe itself as AI-infrastructure to escape halving-driven hashprice compression. Its $0.04/kWh owned-site power cost is a genuine competitive edge, but its 105.6% debt/equity is heavier than net-cash peer RIOT, adding NAV-decay risk in a BTC drawdown akin to MSTR's leveraged treasury model.
Business & Strategy
Revenue is dominated by Bitcoin mining (block rewards plus transaction fees), with a growing AI/HPC compute and energy-monetization leg following the Exaion acquisition. Management is deliberately diversifying toward AI inference to monetize excess and underutilized power.
Primarily the Bitcoin network itself (mining rewards) plus emerging AI-compute tenants across North America, the Middle East, Europe, and LatAm.
Block subsidies and fees from mining, hosting/energy monetization, and nascent AI-cloud/inference contracts. The mix is shifting from ~100% mining toward a hybrid energy-and-compute model.
Electricity (targeted $0.04/kWh at owned sites), ASIC/GPU depreciation, and datacenter buildout capex dominate the cost base.
MARA's moat is scale plus low-cost, self-controlled power — the vertical integration into owned energy assets differentiates it from hosting-dependent miners. The AI pivot is early and unproven; execution and tenant quality will determine whether the moat widens beyond a commoditized hashrate business.
Monetary-Policy Sensitivity
- Lower rates boost risk appetite and BTC spot, the primary MARA driver
- Reduced financing cost on $2.46B debt load
- Improved liquidity for high-beta crypto equities
As a 5.37-beta BTC proxy, MARA amplifies macro liquidity shifts far more than fundamentals would suggest. A dovish pivot that lifts BTC would disproportionately re-rate MARA via both spot and short-covering channels.
SWOT Analysis
- Largest listed pure-play BTC miner with scale hashrate
- Sector-leading $0.04/kWh owned-site power cost
- 69.6% institutional ownership anchored by BlackRock/Vanguard
- Optionality from AI-infrastructure pivot (Exaion, EU AI cloud)
- Net debt of $1.95B, debt/equity 105.6%
- Deeply negative FCF (-61% margin) and ROE (-67%)
- Revenue declining -18.4% YoY
- Persistent open-market insider selling ($2.87M net)
- AI/HPC compute monetization of stranded power
- Short squeeze on 30.8% short float during BTC rallies
- BTC recovery to $100K (21Shares 2026 target)
- Government-validated EU AI infrastructure moat
- BTC drawdown (already -50% from October peak) crushing hashprice
- 7d ETF outflows of $705M signaling institutional supply
- Post-halving reward compression
- Dilution from continued equity/convert issuance
Catalysts & Event Risks
- Q3 2026Q2'26 earnings
Next reported quarter will reveal BTC mark-to-market impact and AI-compute ramp progress.
- Q3 2026Exaion / AI-cloud milestones
First tenant contracts or revenue disclosure from the EU AI infrastructure buildout could re-rate the AI optionality.
- 2026-12-31BTC cycle target
21Shares projects a year-end BTC recovery to $100K, which would drive outsized MARA re-rating.
- Q3 2026Short-squeeze dynamics
30.8% short float and 2.62 days-to-cover set up sharp upside on any positive BTC catalyst.
- Q4 2026Fed rate decisions
Dovish policy would boost BTC spot and MARA's high-beta multiple simultaneously.
The dominant catalyst is BTC spot direction — everything else (AI contracts, short squeezes, earnings) is second-order. With BTC down ~50% from October and ETF flows negative, the near-term catalyst skew is asymmetric to a liquidity reversal.
Technical Analysis
MARA sits at $12.22, below both its 50-day ($13.24) and marginally below its 200-day ($12.44) averages, in the 33rd percentile of its 52-week range. The stock recently flashed a 'golden cross' per Benzinga and showed relative strength versus the broader crypto-equity selloff, but volume of 24.8M trails the 43.5M average, signaling waning conviction. Risk-reward is balanced: support near the $11.74 day-low, with a break above the $13.24 50-day needed to confirm any recovery leg.
Verdict
Macro context. The crypto-miner cohort is undergoing a narrative shift from BTC pure-plays to AI-infrastructure proxies, but a BTC price ~50% off its October peak and $5.8B of 30-day ETF outflows keep the fundamental backdrop hostile. High-beta names like MARA remain leveraged bets on a Fed liquidity pivot and BTC cycle recovery.
MARA offers genuine optionality — low-cost power, an AI-infrastructure pivot with government validation, and a 30.8% short float primed for squeezes — but is fundamentally a leveraged BTC derivative with $1.95B net debt, -67% ROE, and GAAP earnings that swing $2B+ per quarter on mark-to-market. With BTC 50% off highs, ETF flows persistently negative, and insiders selling into every rally, we rate shares HOLD; we'd upgrade to ACCUMULATE on a sustained turn in BTC ETF flows or confirmed AI-cloud revenue, and REDUCE on a break below $11 support.
Data source: Yahoo Finance / yfinance · fetched 7/7/2026, 7:06:28 PM