Back to Portfolio
CLSK

CleanSpark Inc

$9.3 USD 2.38B market cap 2026-03-27
CleanSpark Inc CLSK BUFFETT / MUNGER / KLARMAN SUMMARY
1 SNAPSHOT
Price$9.3
Market CapUSD 2.38B
EVUSD 3.16B
Net DebtUSD 781M
Shares255.8M
2 BUSINESS

CleanSpark is the largest publicly traded pure-play Bitcoin miner in the United States, operating 50+ EH/s of hash rate across 32+ data center sites in Georgia, Tennessee, Wyoming, and Mississippi with over 1 GW of contracted power. Revenue comes entirely from mining Bitcoin, which the company either HODLs on its balance sheet or sells to fund operations. The company is now pivoting to add AI/HPC data center development as a second revenue stream, leveraging its power and land portfolio.

Revenue: USD 766M Organic Growth: 102%
3 MOAT NONE

Bitcoin mining is a commodity business with no durable competitive advantage. CleanSpark has temporary operational advantages (efficient fleet at ~16 J/TH, low ASIC procurement costs at $21.50/TH, vertically integrated operations) but these are replicable by any well-capitalized competitor. No pricing power, no switching costs, no network effects. The Bitcoin protocol's difficulty adjustment ensures no single miner gains a compounding advantage from scale.

4 MANAGEMENT
CEO: Matt Shultz (since Aug 2025, co-founder, returned from Exec Chairman role)

Mixed track record. Funded explosive growth through massive equity dilution (shares grew from 29M to 318M in 4 years). Recently improved: no ATM since Nov 2024, $460M share buyback via convertible proceeds. ASIC procurement at below-market prices shows discipline. BTC HODL strategy generated significant gains in FY2025 bull market. New $1.15B convertible at 0% coupon shows capital markets access. Insider ownership is low at 3.5%.

5 ECONOMICS
41.6% Op Margin
~15% ROIC
USD -1.02B FCF
2.6x Debt/EBITDA
6 VALUATION
FCF/ShareUSD -4.00
FCF YieldNegative
DCF RangeNot applicable - BTC miner has no predictable cash flows

DCF is not meaningful for a Bitcoin miner. Revenue depends entirely on BTC price (unpredictable), CapEx is driven by halving cycles and competition, and the business has never generated positive FCF. NAV-based valuation: BTC treasury (~$1.2B) + mining infrastructure (~$1-1.5B replacement) + land/power optionality - debt ($824M) = ~$1.4-1.9B or $5.47-$7.42/share. Current price $9.30 implies premium for growth and AI optionality.

7 MUNGER INVERSION -86.0%
Kill Event Severity P() E[Loss]
Bitcoin price decline >50% (crypto bear market) -70% 25% -17.5%
April 2028 halving doubles cost per BTC mined -40% 90% -36.0%
AI/HPC pivot fails (no tenants secured) -30% 40% -12.0%
Continued equity dilution destroys per-share value -25% 50% -12.5%
Global hash rate competition erodes mining economics -20% 40% -8.0%

Tail Risk: A prolonged crypto winter (BTC below $40K) combined with rising energy costs would simultaneously crush revenue and inflate costs, potentially making the business unprofitable at the operating level. The $824M debt load becomes dangerous if BTC treasury value declines below collateral thresholds. In a worst case, the company could be forced to sell BTC at distressed prices to service debt, creating a death spiral. The AI pivot requires significant additional capital that may only be available through dilutive equity issuance.

8 KLARMAN LENS
Downside Case

In a bear case, Bitcoin drops to $40-50K, mining becomes marginally profitable at best. The 13,000 BTC treasury declines from ~$1.2B to ~$600M, pushing debt/equity toward 1.0x. The AI/HPC pivot stalls as the company lacks capital for $10M/MW data center buildouts. Shares diluted further through forced capital raises. Stock could revisit $3-5 range (seen in 2022-2023 bear market).

Why Market Wrong

The market may be undervaluing CleanSpark's power and land portfolio. With 1+ GW contracted and a multi-GW pipeline, the scarcity value of US power infrastructure for AI workloads could be worth significantly more than the current EV implies. If AI data center demand materializes as expected, these power assets could be worth $3-5M+ per MW, implying $3-5B in infrastructure value alone. The BTC treasury provides a floor at ~$4.70/share.

Why Market Right

The market may be correctly pricing this as a commodity miner with no moat, massive dilution history, negative FCF, and an unproven AI pivot. The stock has declined 60% from its 52-week high despite Bitcoin staying above $90K, suggesting investors are not willing to pay a premium for mining exposure when BTC ETFs offer direct exposure without operational risk. The AI pivot is 2+ years from generating meaningful revenue and may never materialize.

Catalysts

Positive: Signed AI/HPC tenant agreement (Sandersville or Sealy), sustained BTC price above $120K, successful deployment of 13.5 J/TH fleet upgrade, positive FCF quarter, S&P 500-class index inclusion. Negative: Crypto bear market, regulatory headwinds, tariff impact on ASIC imports, rising energy costs, competitive pressure from hyperscaler self-builds.

9 VERDICT REJECT
D Rejected
Strong Buy$4.5
Buy$6
Sell$24

CleanSpark fails core Buffett/Graham quality screens: no moat, no consistent profitability (4 of 5 years were losses), massive capital intensity (never generated positive FCF), extreme cyclicality (beta 3.56), and chronic dilution (shares grew 10x in 4 years). The AI/HPC pivot adds interesting optionality but is unproven and 2+ years from revenue contribution. This is a leveraged bet on Bitcoin price and AI infrastructure demand, not a quality value investment. At $4.50-6.00 (below book value, near BTC treasury backing per share), it would become interesting as a deep value / special situation play.

🧠 ULTRATHINK Deep Philosophical Analysis

CLSK - Ultrathink Analysis

The Real Question

The real question here is not whether CleanSpark is a good Bitcoin miner. It clearly is -- probably the best pure-play miner in America by operational metrics. The real question is whether "best Bitcoin miner" is a category worth investing in at all.

Bitcoin mining is the curious case of a business that produces something potentially extraordinary (a scarce, decentralized digital asset) through a process that is inherently commodity-like and returns-destructive. Every dollar CleanSpark invests in faster ASICs, cheaper power, and more efficient cooling is matched by competitors doing exactly the same thing. The Bitcoin protocol's difficulty adjustment is the great equalizer -- it is an algorithm specifically designed to prevent any miner from ever gaining a lasting edge. In what other industry does the product itself actively resist the accumulation of competitive advantage?

This is not like oil exploration, where a company that discovers a rich field earns outsized returns for years. In Bitcoin mining, your "rich field" gets mathematically harder to extract from the moment everyone else notices it. CleanSpark's 50 EH/s and 16 J/TH efficiency are impressive accomplishments that will be rendered obsolete by the next generation of ASICs, available to anyone who can write the check.

Hidden Assumptions

The market is making several hidden assumptions about CleanSpark that deserve scrutiny:

Assumption 1: Bitcoin will remain above $80K long enough for the business to matter. CleanSpark's economics only work at high BTC prices. At $43K marginal cost per coin, the margin of safety is ~50% at $90K BTC but collapses entirely at $50K. The entire investment case rests on the belief that Bitcoin is in a secular bull market. If you believe that, you should just buy Bitcoin (via ETF) -- why take on all the operational risk, dilution, and capital intensity of a miner?

Assumption 2: The AI pivot is real and executable. In Q1-Q2 FY2025, CEO Zach Bradford was emphatic that CleanSpark would NOT pivot to HPC/AI. He called it "reckless" for anyone to build AI data centers without signed customers. Three months after Matt Shultz returned as CEO, the entire strategy reversed. The company is now positioning as a "digital infrastructure platform." This 180-degree pivot should give investors pause. Is this conviction or opportunism? The fact that zero AI tenants have been signed, while capital is being deployed for land and power, suggests the company is betting it can build infrastructure and tenants will come. That is precisely the strategy Bradford warned against.

Assumption 3: Share dilution is over. Management touts the stock buyback and lack of ATM offerings. But the convertible notes ($1.8B total) represent massive potential dilution. And the AI pivot requires enormous capital -- $10M/MW vs $1M/MW for mining. Where will this capital come from if not equity issuance? The history suggests dilution is a feature of the business model, not a bug that has been fixed.

Assumption 4: Power assets are worth more than mining revenue. The Aschenbrenner thesis likely values CleanSpark's 1+ GW power portfolio at a premium for AI/HPC use. But power contracts are not unique assets -- they are agreements with utilities that can be replicated by any developer willing to invest in the same rural markets. The competitive advantage, if any, is temporal (CleanSpark has the contracts NOW while AI demand is surging). But temporal advantages in power markets are exactly that -- temporal.

The Contrarian View

For the bears to be right, one or more of the following would need to be true:

  1. Bitcoin is a cyclical asset, not a secular one. If BTC reverts to $30-50K in the next crypto winter (as it has done multiple times historically, with 70-85% drawdowns), CleanSpark's business model breaks. Mining at $43K cost per coin is unprofitable below ~$50K.

  2. AI data center demand is oversupplied. If the current frenzy to build AI infrastructure results in oversupply (as happened with fiber optics in the late 1990s), power assets will not command premium valuations. CleanSpark would be stuck with expensive land and power contracts without high-value tenants.

  3. BTC ETFs make miners obsolete as investment vehicles. Why own a Bitcoin miner at 1.36x book (with operational risk, dilution risk, and management risk) when you can own BTC directly through a spot ETF at 0.25% expense ratio? The proliferation of BTC ETFs structurally devalues miners as investment vehicles. The only argument for miners over ETFs is leverage (miners amplify BTC moves) and optionality (AI pivot). But leverage cuts both ways, and optionality has a cost.

The bears have a strong case. In a rational market, CleanSpark should trade at or below its net asset value (BTC treasury + infrastructure replacement cost - debt), which is roughly $5.50-$7.50/share. The current $9.30 price implies the market is paying something for AI optionality and growth -- but not much.

Simplest Thesis

CleanSpark is a leveraged bet on Bitcoin price with an unproven call option on AI infrastructure demand, and it fails every quality screen that distinguishes investing from speculation.

Why This Opportunity Exists

Leopold Aschenbrenner sees something different from a value investor. He sees a company sitting on increasingly scarce US power infrastructure at a time when AI demand is creating genuine physical scarcity. His framework is not Buffett's -- it is a technologist's assessment of which physical assets will matter in an AGI world.

The opportunity (if it exists) is that the market prices CleanSpark as a Bitcoin miner (volatile, commodity, no moat) while the underlying assets (power, land, infrastructure, operational capability) may be worth more as AI data center assets. The gap between "Bitcoin miner valuation" and "AI infrastructure platform valuation" is the potential mispricing.

But this is a thesis that requires the passage of time and execution to validate. Today, CleanSpark has zero AI revenue, zero signed tenants, and a management team that reversed its strategy six months ago. The mispricing, if real, will take 2-3 years to close. And in those 2-3 years, a lot can go wrong -- BTC bear market, dilution, execution failure, competition.

What Would Change My Mind

I would reconsider this analysis if:

  1. CleanSpark signs a binding, multi-year AI/HPC tenant agreement with a creditworthy counterparty (hyperscaler or Fortune 500) at either Sandersville or Sealy. This would transform the company from "miner with optionality" to "infrastructure platform with contracted cash flows."

  2. The company generates two consecutive quarters of positive free cash flow without relying on BTC price appreciation. This would prove the business model is self-sustaining.

  3. Shares outstanding remain stable or decline for 4+ consecutive quarters, proving that the dilution cycle is genuinely broken.

  4. Bitcoin sustains above $120K through a full halving cycle (through April 2028), demonstrating that mining economics remain attractive even after reward halving.

None of these conditions are currently met.

The Soul of This Business

At its core, CleanSpark is an energy arbitrage business. It buys electricity cheap in rural America and converts it into Bitcoin. The margin between the cost of electricity and the value of Bitcoin mined is the entire business. When that margin is fat (high BTC price, low energy costs), CleanSpark prints money. When it is thin (low BTC price, high energy costs, high difficulty), the business bleeds cash.

This is not inherently bad -- there are great commodity businesses. But great commodity businesses have cost advantages that persist for decades (think Saudi Aramco's $2/barrel lifting costs or Nucor's minimill cost structure). CleanSpark's cost advantages are measured in months, not decades, because ASIC technology evolves rapidly and power contracts are finite.

The AI pivot is an attempt to transcend the commodity trap -- to use the same power and infrastructure assets for higher-value, longer-duration workloads with contracted revenue. If successful, it would transform CleanSpark from a cyclical commodity miner into something closer to a digital infrastructure REIT. That is a genuinely interesting metamorphosis.

But metamorphosis is rare in business. Most caterpillars that announce plans to become butterflies remain earthbound. The few that succeed (Amazon from books to cloud, Netflix from DVDs to streaming) had deep competitive advantages in their new domains. CleanSpark's advantage in AI data centers over specialized developers like Vantage, QTS, or DataBank is unclear.

The honest assessment: CleanSpark is an operationally excellent company in a structurally challenging industry, attempting a strategic pivot that, if successful, would make it a completely different business. Betting on that transformation at current prices is not value investing. It is venture-style speculation on management's ability to execute a pivot while maintaining a capital-intensive legacy business. For Aschenbrenner, with his AGI-centric worldview and his portfolio's appetite for infrastructure optionality, a 0.4% position is a rational hedged bet. For a value investor seeking durable competitive advantages and predictable cash flows, it is a pass.

Executive Summary

3-Sentence Thesis

CleanSpark is the largest publicly traded pure-play Bitcoin miner in the United States, operating 50+ EH/s across 32+ data center sites in four states with 1 GW of contracted power. The company has achieved "escape velocity" with 55% gross margins, $766M FY2025 revenue, 13,000+ BTC on its balance sheet ($1.2B), and is now pivoting to add AI/HPC data center development as a second growth engine. However, this is a highly cyclical, capital-intensive business with zero durable moat, massive Bitcoin price dependency (beta 3.56), persistent negative free cash flow, chronic equity dilution (shares outstanding grew from 29M to 318M in 4 years), and the AI pivot is unproven -- making it a speculative infrastructure play rather than a quality value investment.

Key Metrics Dashboard

Metric Value Assessment
Price $9.30 Near 52-week low ($6.45)
Market Cap $2.38B
Book Value/Share $6.85 (Sep 2025) P/B = 1.36x
Revenue (FY2025) $766M +102% YoY
Gross Margin 55% (FY2025) Strong but BTC-price dependent
Net Income (FY2025) $364M Includes BTC mark-to-market gains
EBITDA (FY2025) $763M Inflated by BTC fair value changes
Normalized EBITDA (FY2025) ~$305M ~40% margin
Free Cash Flow (FY2025) -$1.02B Massively negative
Total Debt $824M $650M convert + $200M credit line
BTC Treasury 13,000+ (~$1.2B) Largest self-mined public holding
Hash Rate 50 EH/s ~5% global network share
Beta 3.56 Extreme volatility
Shares Outstanding 256M (current) vs 318M (Sep 25) Recent buyback via convert
ROE (FY2025) 16.8% First positive year, driven by BTC gains
ROE (5-yr avg) Deeply negative Consistent losses FY2021-FY2024
Dividend None No dividends

Verdict: REJECT

CleanSpark fails multiple Buffett/Graham quality screens. It is not a business with predictable, recurring earnings or durable competitive advantages. It is a commodity mining operation whose profitability is entirely dependent on Bitcoin price (which it cannot control) and whose growth has been funded by massive equity dilution. The AI/HPC pivot adds optionality but is unproven and introduces execution risk. This is a leveraged bet on Bitcoin and AI infrastructure demand, not a quality value investment.

Phase 0: Why This Opportunity Might Exist

Leopold Aschenbrenner's Situational Awareness LP took a small position ($17M, 0.4% of portfolio) in Q4 2025. His thesis likely centers on:

  1. AGI infrastructure demand: CleanSpark controls 1+ GW of contracted power across the US, which is increasingly scarce and valuable for AI/HPC workloads
  2. Optionality in the AI pivot: The Sandersville, GA (250 MW) and Sealy, TX (285 MW) sites could command premium rents as AI data centers
  3. Bitcoin as AGI hedge: If AGI requires massive compute/energy, Bitcoin mining infrastructure represents a call option on energy-compute scarcity
  4. Cheap power assets: CleanSpark's land+power portfolio may be worth more than its current market cap if AI demand materializes

This is a speculative infrastructure bet, not a value investment. The 0.4% allocation confirms it as a small, optionality-driven position.

Phase 1: Risk Analysis (Inversion - "What Could Kill This?")

Risk Register

# Risk Event Severity Likelihood Expected Loss
1 Bitcoin price decline >50% (bear market) -70% 25% -17.5%
2 Bitcoin halving economics (Apr 2028) halves revenue per BTC -40% 90% -36.0%
3 AI/HPC pivot fails (no tenants, execution delays) -30% 40% -12.0%
4 Continued equity dilution destroys per-share value -25% 50% -12.5%
5 Rising energy costs compress margins -20% 30% -6.0%
6 Regulatory crackdown on crypto mining (energy/environment) -35% 10% -3.5%
7 Network hash rate competition erodes mining economics -20% 40% -8.0%
8 Convertible note conversion dilutes shares at $24.66 -15% 30% -4.5%
9 BTC treasury mark-to-market losses trigger debt covenants -25% 15% -3.8%
10 Management strategy shift/execution risk (new CEO) -15% 25% -3.8%

Total Expected Downside: -107.6% (non-additive; multiple risks correlated)

Critical Risk Deep-Dive

1. Bitcoin Price Dependency (Existential) CleanSpark's entire revenue is Bitcoin price x BTC mined. At $100K BTC, they earn ~$98K/BTC with ~$43K marginal cost = ~56% margin. At $50K BTC, the margin compresses to ~14% -- barely covering overhead. At $40K, the business loses money on every coin mined. The company has zero ability to influence Bitcoin price. This is the defining risk.

2. Halving Cycle (Structural) The April 2024 halving already cut block rewards from 6.25 to 3.125 BTC. The next halving (~April 2028) will cut to 1.5625 BTC. Each halving approximately doubles the cost of mining each Bitcoin (holding hash rate constant). CleanSpark must continuously increase hash rate and efficiency just to maintain production levels. This is an arms race with no finish line.

3. Capital Intensity and Dilution (Chronic) Shares outstanding: 29M (2021) -> 43M (2022) -> 103M (2023) -> 217M (2024) -> 318M (2025, pre-buyback) -> 256M (current after buybacks). That is an 8.8x dilution in 4 years. The recent $1.15B convertible note (0% coupon, 27.5% premium) represents another potential ~40M+ share dilution if converted. Management claims they are done with ATMs, but the business model requires constant capital to grow hash rate.

4. AI/HPC Pivot (Unproven) The Q4 FY2025 earnings call (Nov 2025) marked a significant strategic pivot. New/returning CEO Matt Shultz is repositioning CleanSpark as a "digital infrastructure platform" serving AI, HPC, and grid balancing. Key sites: Sandersville, GA (250 MW) and Sealy, TX (285 MW, energizing 2027). However:

  • No signed AI/HPC tenant agreements yet
  • Data center build-out costs ~$10M/MW vs $1M/MW for mining
  • Capital requirements are enormous
  • Competition from Core Scientific, Applied Digital, Cipher Mining, plus hyperscalers
  • Execution timeline 2-5 years
  • Previous CEO Zach Bradford explicitly rejected HPC pivot; new CEO reversed course

Phase 2: Financial Analysis

Revenue and Profitability (5-Year)

Metric FY2021 FY2022 FY2023 FY2024 FY2025
Revenue $39M $132M $168M $379M $766M
Gross Margin 32.7% 37.5% 17.3% 36.8% 41.6%
Net Income -$22M -$57M -$138M -$146M +$365M
EPS (diluted) -$0.74 -$1.34 -$1.34 -$0.67 +$1.15

Revenue growth is impressive (19.5x in 4 years) but driven entirely by (a) Bitcoin price appreciation and (b) hash rate scale-up funded by massive dilution. FY2025 was the first profitable year, but net income was inflated by ~$460M in BTC mark-to-market fair value gains. Normalized (operations-only) EBITDA was ~$305M on ~$766M revenue = 40% margin.

Most recent quarter (Q1 FY2026, Oct-Dec 2025): Revenue $181M, net loss $379M. The loss was driven by Bitcoin price decline in the quarter requiring mark-to-market write-downs. This illustrates the extreme earnings volatility.

ROE Decomposition (DuPont)

Component FY2025 FY2024 FY2023
Net Margin 47.6% -38.5% -82.0%
Asset Turnover 0.24x 0.19x 0.22x
Equity Multiplier 1.46x 1.11x 1.13x
ROE 16.8% -8.3% -20.4%

FY2025 ROE is misleading -- driven by one-time BTC fair value gains. The business does not generate consistent returns on equity. Asset turnover is extremely low (0.24x), typical of capital-intensive mining.

Owner Earnings / Free Cash Flow

Metric FY2021 FY2022 FY2023 FY2024 FY2025
Operating CF -$35M $71M -$30M -$234M -$461M
CapEx $229M $191M $302M $806M $563M
FCF -$264M -$119M -$333M -$1,040M -$1,024M

Free cash flow has been deeply negative every single year. Cumulative FCF over 5 years: approximately -$2.8 billion. This is not a cash-generating business -- it is a capital-consuming machine that requires continuous external funding (equity issuance, convertible debt, BTC-collateralized loans).

Important caveat on OCF: CleanSpark's operating cash flow is negative because they HODL mined Bitcoin rather than selling it immediately. The BTC sits on the balance sheet as an investment, so the "cost" of mining flows through OpEx but the "revenue" (BTC retention) is classified as investing activity under the new FASB fair value accounting standard. If they sold all mined BTC immediately, OCF would be positive. However, the CapEx remains enormous, so FCF would still be deeply negative.

Balance Sheet Analysis

Metric FY2025 Assessment
Total Assets $3.18B ~40% is Bitcoin at fair value
Total Equity $2.18B Book value inflated by BTC holdings
Total Debt $824M $650M convert + lines of credit
Net Debt ~$781M Cash $43M, debt $824M
Debt/Equity 0.38x Moderate, but increased 10x in one year
Current Ratio 4.18x Strong, BTC classified as current asset
Book Value/Share $6.85 Stock trading at 1.36x book
BTC Treasury 13,000+ BTC (~$1.2B) 3rd largest among public miners

The balance sheet looks strong on the surface but is extremely fragile because ~40% of total assets are Bitcoin at fair value. A 50% BTC decline would wipe ~$600M from assets and push Debt/Equity to dangerous levels. The $650M convertible is at 0% coupon (no cash interest cost), which is favorable, but conversion at $24.66 would add ~26M shares (10%+ dilution).

Valuation

Metric Value Comment
P/B 1.36x Below book, but book is BTC-inflated
P/S (TTM) 3.0x $785M TTM revenue
EV/Revenue 3.7x
EV/EBITDA 20.85x EBITDA includes BTC fair value gains
EV/Normalized EBITDA ~10.4x Using $305M normalized EBITDA
Forward P/E 13.2x IF Bitcoin stays above $90K
BTC Treasury Value ~$1.2B vs $2.38B market cap
Net Asset Value (BTC only) ~$420M After subtracting all liabilities from BTC
Price/BTC per share ~$4.69/BTC Market values each CLSK share at ~2x its BTC backing

DCF is not meaningful for a Bitcoin miner because:

  • Revenue is entirely dependent on BTC price (unpredictable)
  • No recurring/contracted cash flows (unlike software or subscription businesses)
  • CapEx is lumpy and driven by halving cycle and competition
  • The business has never generated positive FCF

NAV-based valuation: Total BTC (~13,000 x ~$92K) = ~$1.2B + Mining infrastructure (replacement cost ~$1B-1.5B at $1M/MW for 1 GW) + Land/power contracts (option value) - Total debt ($824M) - Operating burn = rough NAV of $1.4B-$1.9B. At 256M shares = $5.47-$7.42/share NAV. Current price $9.30 implies a premium for growth and AI optionality.

Phase 3: Moat Analysis

Moat Assessment: NONE (No Durable Competitive Advantage)

Bitcoin mining is a commodity business with no structural moat. Key analysis:

1. No Pricing Power Revenue per BTC = market price of Bitcoin. CleanSpark has zero ability to charge a premium. Every miner in the world earns the same price per BTC mined.

2. No Switching Costs Investors can easily switch to any other Bitcoin exposure (spot BTC ETFs, other miners, direct BTC purchase). Miners are fungible.

3. No Network Effects More hash rate does not create a compounding advantage. In fact, more global hash rate makes each additional EH/s less productive (difficulty adjustment).

4. Cost Advantages (Temporary, Not Durable) CleanSpark has achieved:

  • Fleet efficiency: ~16 J/TH (industry-leading)
  • Low power costs: $0.04-0.06/kWh
  • ASIC procurement at $21.50/TH (30%+ below market)
  • Vertically integrated operations across 32+ sites

These are real operational advantages but are NOT durable moats because:

  • Any well-capitalized competitor can buy the same ASICs
  • Power contracts can be replicated
  • Efficiency gains from new hardware are available to everyone
  • No intellectual property or patents protect their process

5. Scale Advantage (Moderate, Temporary) At 50 EH/s (~5% of global network), CleanSpark has meaningful scale advantages:

  • Lower per-unit overhead costs
  • Better vendor pricing (ASIC procurement)
  • Access to cheaper capital
  • Regulatory/utility relationships

However, scale in Bitcoin mining does not compound. The Bitcoin protocol automatically adjusts difficulty, ensuring that no single miner's marginal return improves simply from scale.

6. AI/HPC Optionality (Unproven) The power + land portfolio could become valuable for AI/HPC, but this is optionality, not a moat. Competitors (Core Scientific, Applied Digital, Cipher) are pursuing the same strategy. There is no evidence CleanSpark has a defensible advantage in AI data center development.

Moat Rating: NONE

Moat Durability: N/A -- no moat exists Trend: N/A

Phase 4: Decision Synthesis

Management Assessment

CEO: Matt Shultz (returned Aug 2025, previously Executive Chairman since company inception)

  • Co-founder and long-time executive
  • Guided the company from microgrid/energy software to pure-play Bitcoin mining
  • Now pivoting toward AI/HPC platform strategy
  • Previous CEO Zach Bradford explicitly rejected HPC pivot in Q1/Q2 FY2025; Shultz reversed this within months
  • Strategic flexibility could indicate either vision or lack of conviction

CFO: Gary Vecchiarelli (also President)

  • Strong financial acumen, well-articulated capital strategy
  • Led $650M and $1.15B convertible note offerings
  • Pioneered institutional-grade BTC treasury management (covered calls, yield strategies)
  • Digital Asset Management (DAM) team generating meaningful premiums (~12% annualized yield)

Capital Allocation:

  • Mixed track record: aggressive growth funded by massive dilution (29M to 318M shares)
  • Recent improvement: no ATM since Nov 2024, $460M share buyback via convertible proceeds
  • ASIC procurement at below-market prices shows operational discipline
  • BTC HODL strategy paid off handsomely in FY2025 bull market but carries concentration risk

Insider Ownership: 3.5% -- low for management alignment Institutional Ownership: 83% -- high institutional interest

Position Sizing

Recommendation: DO NOT BUY / REJECT

This does not meet Buffett/Graham quality criteria:

  1. No moat (fails moat test)
  2. No consistent profitability (4 of 5 years were losses)
  3. Massive capital intensity (never generated positive FCF)
  4. Extreme cyclicality (beta 3.56, entirely BTC-dependent)
  5. Chronic dilution (shares grew 10x in 4 years)
  6. No dividends, no capital return history
  7. Management pivot raises execution risk

When Would This Be Investable?

For investors willing to accept Bitcoin mining as a speculative asset class (not value investing), CleanSpark would be more interesting at:

Scenario Entry Price Rationale
Deep value (NAV-based) $5.00-$6.00 Below book value, covered by BTC treasury
Signed AI tenant Re-evaluate Would transform risk profile if contracted
BTC < $50K washout Re-evaluate Would test balance sheet resilience
Proven FCF generation Re-evaluate Would need 2+ quarters of positive FCF

Why Leopold Aschenbrenner Might Own This

Aschenbrenner's Situational Awareness LP thesis is about AGI infrastructure. His likely rationale:

  1. Power scarcity: 1+ GW of contracted US power is increasingly valuable as AI/HPC demand explodes
  2. Optionality: CleanSpark's power/land portfolio is a call option on AI infrastructure demand
  3. Bitcoin as compute hedge: If AI drives energy demand up, Bitcoin mining infrastructure benefits
  4. Small position (0.4%): This is a speculative bet, not a core holding

This is a valid speculative thesis but it is NOT a value investment thesis.

Monitoring Triggers

Trigger Action
Signed AI/HPC tenant agreement Re-evaluate thesis fundamentally
2 consecutive quarters positive FCF Upgrade to RESEARCH
BTC price sustained > $120K Earnings become very attractive
Shares outstanding increase > 10% Confirms dilution is not over
BTC price < $50K sustained Balance sheet stress test
Management sells significant insider shares Red flag

Appendix: Key Data Points

Bitcoin Mining Operations (as of Q4 FY2025)

Metric Value
Hash Rate 50 EH/s
Fleet Efficiency ~16 J/TH (deploying 13.5 J/TH units)
Mining Sites 32+ across GA, TN, WY, MS
Contracted Power 1+ GW
BTC Mined (FY2025) ~7,873
Avg Revenue per BTC ~$98,000
Avg Marginal Cost per BTC ~$43,000
BTC Treasury 13,000+
Global Hash Rate Share ~5%

AI/HPC Pipeline

Site Location Power Status Timeline
Sandersville Georgia 250 MW Live (mining), AI-ready 2026 tenant target
Sealy Texas 285 MW Contracted, ERCOT approved 2027 energization
Metro Atlanta Georgia 100+ MW Operational High demand for low-latency
Pipeline Various Multi-GW Early stage 2027+

Capital Structure

Item Amount Terms
Convert #1 (Dec 2024) $650M 0% coupon, $24.66 conversion
Convert #2 (Oct 2025) $1.15B 0% coupon, 27.5% premium, 6.25yr
Coinbase Credit Line $400M capacity BTC-collateralized
Stock Buyback $460M Via Convert #2 proceeds
Net Shares Reduction ~10.9% From buyback

Share Count History

Date Shares Outstanding Change
FY2021 29.4M Base
FY2022 42.6M +45%
FY2023 102.7M +141%
FY2024 216.9M +111%
FY2025 317.8M +47%
Current 255.8M -19% (buyback)

Analysis based on SEC EDGAR 10-K filings (FY2023-2025), AlphaVantage financial statements, 4 quarters of earnings call transcripts (Q1-Q4 FY2025), and company investor relations materials. No analyst reports or price targets were used in this analysis.