CleanSpark, Inc.

Overview

This is a Druckenmiller-style top-down, reflexivity-driven assessment of CleanSpark Inc. (CLSK) as a bitcoin miner executing a high-stakes pivot into AI/HPC data-center infrastructure. The analysis focuses on macro liquidity, power scarcity, capital-cycle dynamics, market positioning, and asymmetric risk/reward rather than traditional value metrics.

Macro Context

CLSK sits at the collision of two macro regimes. First, the AI infrastructure buildout is in a secular acceleration phase: power is the binding constraint, and grid-connected megawatts are re-rating from commodity mining capacity into strategic compute infrastructure. Financing conditions for high-grade AI data-center assets are described by management as constructive, with recent data-center financings oversubscribed and pricing around 6%. Second, bitcoin mining is in a cyclical profit recession following the 2024 halving. Hashprice has fallen toward roughly $27-31/PH/s/day, and only the most efficient miners are generating cash. Geopolitically, U.S. customs/tariff risk on Chinese-origin mining hardware and grid/power regulation add sector risk. The dominant macro force here is not generic Fed policy but a power-first capex cycle in which contracted, energized capacity is the scarce asset.

Company Position in Macro Landscape

CleanSpark is simultaneously a beneficiary of the AI/power scarcity theme and a victim of depressed bitcoin mining economics. It controls approximately 1.8 GW of contracted power and has signed a 20-year, $6.6 billion triple-net lease at Sandersville for 175 MW of critical IT load, with expected deliveries beginning Q4 2027 and approximately $330 million in average annual NOI. It also has a letter of intent covering its 885 MW Texas portfolio. However, the current income statement is still a bitcoin miner: Q3 FY2026 revenue was $138.0 million, down 30.5% year-over-year; net loss was $239.8 million; Adjusted EBITDA was negative $113.0 million. Cash was $202.6 million, bitcoin holdings were $814.9 million, total long-term debt was $1.8 billion, and stockholders' equity was only $0.8 billion. CLSK is therefore best viewed as a leveraged call option on AI infrastructure, with a deteriorating mining bridge.

Reflexivity Analysis

The positive reflexive loop is clear: a signed investment-grade tenant validates the Sandersville site and power strategy, lowering financing costs and making future portfolio-level deals more likely. If Texas converts from exclusivity to a definitive lease, the company begins to be re-rated from a marginal bitcoin miner to a long-duration digital-infrastructure landlord. That could reduce the cost of equity, attract infrastructure capital, and accelerate management's ability to monetize additional megawatts. The negative loop is also active: falling bitcoin prices cause mark-to-market losses, shrink equity, weaken operating cash flow, and raise the perceived risk premium. With the stock below its 50-day and 200-day moving averages and below the 2030/2032 convertible conversion prices of roughly $14.80 and $19.16, dilution and hedging overhangs are live. Market positioning is mixed: a prominent 13F showed one fund adding 648% in Q1 2026 at an implied cost near $8.51, while other institutions exited. Analyst targets cluster around $16-27, well above the current $12.09 price. The negative reflexivity currently dominates, but the lease creates a potential cusp for reversal if financing and Texas milestones arrive.

Competitive Position & Disruptive Threats

CleanSpark has one of the most efficient public mining fleets, with approximately 16.2 J/TH, 47.3 EH/s of hashrate, and about 4.66% of global hash. Its cost discipline and relatively low SG&A/stock-based compensation compared with peers are real advantages. In AI/HPC, however, it is a second mover behind Core Scientific, IREN, TeraWulf, Cipher, and Hut 8, which already have AI revenue or more advanced builds. CLSK's moat is contracted power, site optionality, community relationships, and operational uptime. Its main disruptive threat is that AI competitors outbid it for power or that execution failures delay lease cash flows. AI data-center supply is also expanding rapidly, and a future supply overbuild could compress lease yields. Tariff/regulatory exposure and reliance on a confidential high-investment-grade tenant add uncertainty.

Asymmetric Risk/Reward

At $12.09, the payoff is not yet cleanly asymmetric. The stock is down from a 52-week high of $23.61 and above a 52-week low of $8.00. If Sandersville and Texas are fully financed and de-risked, the AI portfolio could support a valuation far above the current $3.1 billion market cap, and analyst targets imply meaningful upside. However, per-share upside is diluted by roughly 104 million shares underlying the convertible notes and large performance/restricted equity awards. On the downside, a bitcoin break below $60,000, failed project financing, or lease milestone slippage could push the stock toward $8 or below; book value is only $2.97. The best asymmetry likely appears either on a capitulation flush toward $8-10 or on confirmed financing/news that carries the stock back above the $14.80-15.00 conversion zone.

Key Risks

Investment Details

Overview

This report applies Joel Greenblatt's Magic Formula framework to CleanSpark, Inc. (CLSK). The Magic Formula ranks stocks on two simple quantitative measures: earnings yield (EBIT/enterprise value) to find cheap stocks, and return on capital (EBIT/(net working capital + net fixed assets)) to find good businesses. CLSK is examined using recent financial data through fiscal Q3 2026, with particular attention to normalized earnings after adjusting for bitcoin mark-to-market volatility.

Business Quality Assessment

Operationally, CleanSpark is one of the more efficient bitcoin miners: approximately 16.2 J/TH fleet efficiency, 47.3 EH/s average hashrate, and power costs around $0.052/kWh in fiscal Q2 2026. However, a good Magic Formula business must generate high returns on capital, and CLSK currently does not. In fiscal Q3 2026, revenue was $138.0 million, cost of revenues excluding depreciation was $85.5 million, producing only a $52.5 million gross profit before overhead. After professional fees, payroll, G&A, disposal losses, and indirect tax expenses, core operating profit before depreciation and bitcoin fair value changes was roughly breakeven to slightly negative. Once depreciation of $111.0 million is included, normalized EBIT was approximately negative $116 million for the quarter. For the nine months ended June 30, 2026, normalized EBIT after adding back bitcoin fair value losses and impairments was still deeply negative. Return on capital is therefore negative. Historically, CLSK generated strong operating returns during bitcoin bull markets, but those returns have not been durable through the post-halving hashprice downturn. The emerging AI/HPC leasing strategy may eventually create high-return contracted cash flows, but it has not yet produced revenue and requires massive additional capital. Under the Magic Formula, this is not currently a good business in the quantitative sense.

Valuation Analysis

CleanSpark is not cheap on current or normalized earnings. Using the June 30, 2026 balance sheet: market capitalization is approximately $3.10 billion; total debt including current portion is approximately $1.78 billion; cash is approximately $202.6 million; and bitcoin holdings plus bitcoin collateral receivable are approximately $814.9 million. Excluding bitcoin as a liquid asset, enterprise value is approximately $3.87 billion. If one did not subtract bitcoin, EV would be approximately $4.68 billion. Normalized EBIT is negative; a rough trailing figure is approximately negative $300 million, producing a negative earnings yield around -8%. Even if one adds back all bitcoin fair value losses, current mining operations do not generate enough EBIT to produce a positive earnings yield. Compared to a 10-year Treasury yield in the mid-single digits, a negative earnings yield is not attractive. The bull case relies on future AI/HPC cash flows: the Sandersville lease is expected to generate about $330 million of annual NOI after completion in late 2027. Against the current EV, that implies roughly 11.7x future stabilized NOI, but the project still requires an estimated $1.75 billion to $2.1 billion of landlord capital, and there is no guarantee that Texas portfolio exclusivity converts into contracts.

Magic Formula Ranking

Normalized Earnings Analysis

Reported GAAP earnings are distorted by bitcoin fair value accounting. For example, fiscal Q2 2026 included a $224.1 million loss on the fair value of bitcoin, and fiscal Q3 2026 included a $116.3 million loss. These are non-cash mark-to-market swings and should be added back when evaluating normalized operating earnings. However, even after adding back bitcoin fair value losses, impairments, and other one-time items, EBIT remains negative because depreciation is high and mining economics are weak. In fiscal Q3 2026, gross profit was about $52.5 million; cash operating expenses including professional fees, payroll, and G&A consumed approximately $53.2 million, leaving essentially no pre-depreciation operating profit. Depreciation of $111.0 million then drove EBIT deeply negative. Sustainable owner earnings are therefore not present. The AI/HPC lease may change the earning power of the company, but it is not yet in service and should not be treated as current normalized earnings.

Why The Market Is Wrong

The market is not clearly mispricing CLSK from a Magic Formula perspective. The stock is down sharply from its 52-week high, but that does not make it cheap on current earnings. The market may actually be ascribing significant option value to the AI/HPC pivot. The signed $6.6 billion Sandersville lease is a real strategic milestone, and Texas exclusivity creates large optionality. However, those cash flows are years away, materially capital-intensive, and subject to execution risk. The contrarian value case would require the market to be overly pessimistic about future AI cash flows, but the current price already embeds substantial AI expectations relative to current negative EBIT. In Greenblatt terms, the market is not wrong about the present: this business is not currently high-return or cheap.

Key Risks

Conclusion

The Magic Formula approach seeks companies that are both good and cheap. CleanSpark currently fails both tests quantitatively: it has negative normalized EBIT, negative return on capital, and substantial net debt. The AI/Sandersville lease is a promising longer-term transformation, but it requires enormous capital and will not produce cash flows until late 2027 at the earliest. A disciplined Greenblatt investor would not buy this stock today because it would not clear a positive-EBIT screen. Existing holders are exposed to AI optionality rather than investing in a high-return, undervalued business. Until the company demonstrates sustainable positive operating earnings, the Magic Formula verdict is to avoid or sell.

Overview

A DeepFuckingValue-style deep value and narrative analysis of CleanSpark, Inc. (CLSK) at $12.09. The market is pricing CLSK as a cash-burning Bitcoin miner with a risky AI pivot. This report asks whether the bear consensus is overdone given a signed $6.6B 20-year data center lease, a scarce 1.8 GW contracted power portfolio, a liquid Bitcoin treasury, and a potential re-rating as AI infrastructure execution becomes real.

The Bear Case

Consensus hates CLSK because it still generates 100% of revenue from Bitcoin mining while carrying an AI infrastructure story. Q3 FY2026 revenue fell 30.5% year-over-year to $138.0 million, net loss was $239.8 million, and adjusted EBITDA swung to -$113.0 million from +$377.7 million a year earlier. GAAP losses are worsened by non-cash Bitcoin mark-to-market losses, including $116.3 million in Q3, plus $111 million of depreciation and amortization. Mining economics are weak: hashprice is near post-halving lows, network difficulty remains elevated, and energy costs around 5.2 cents/kWh leave little cash flow cushion. The balance sheet now carries $1.8 billion in long-term debt, mainly converts, and the Sandersville AI project may require $10-12 million per MW for 175 MW -- roughly $1.75-2.1 billion -- before first revenue in Q4 2027. Bears also point to execution risk, no contracted AI cash flow today, potential tariff exposure up to $130 million, and dilution risk from the convertible notes.

The Bull Case

CLSK is a second-mover using an energy-native operating history to convert scarce, grid-connected power into AI/HPC infrastructure. The July 2026 Sandersville lease is a 20-year triple-net agreement with a high-investment-grade global technology tenant: $6.6 billion of contracted revenue, expected average annual NOI of approximately $330 million at near-100% contribution margins, and extension options to $11.6 billion. The same tenant has an LOI and exclusivity arrangement covering CleanSpark's entire Texas portfolio of up to 885 MW. That is substantial embedded optionality relative to a $3.1 billion market cap. The company also holds about $814.9 million of Bitcoin, $202.6 million of cash, and $400 million of untapped credit lines, against only $155.8 million of current liabilities. The convertible notes are low or zero coupon and not due until 2030/2032. Meanwhile, CLSK still operates a highly efficient mining fleet at roughly 16.2 J/TH, meaning a Bitcoin recovery would immediately improve cash generation. A high-conviction institutional filer increased its CLSK stake by 648.4% quarter-over-quarter to $104.47 million, and analyst price targets cluster from $16 to $27.

Fundamental Deep Dive

Sentiment & Technical Setup

Catalyst Analysis

The main catalysts are: 1) Sandersville project financing and possible tenant disclosure; 2) conversion of the Texas 885 MW LOI/exclusivity into a definitive lease; 3) commercialization of additional sites such as Washington, Tennessee, and Wyoming; 4) a Bitcoin/hashprice recovery restoring mining cash flow; 5) analyst re-rating as AI infrastructure revenue becomes more visible; 6) continued institutional accumulation after the 648% conviction add; and 7) potential short squeeze if negative sentiment is wrong. The biggest near-term catalyst risk is that AI lease cash flow does not begin until Q4 2027, so the market needs to see financing and execution milestones.

Key Risks

Conclusion

At $12.09 with a $3.1 billion market cap, CLSK trades below its 50-day moving average of $14.39 and near its 200-day average, providing a contrarian technical setup. The market is pricing the company as a distressed Bitcoin miner while giving little credit to a signed $6.6 billion lease with a high-investment-grade tenant, $1 billion of cash and Bitcoin liquidity, and 885 MW of Texas optionality. If Sandersville delivers, it could produce approximately $330 million of annual NOI at near-100% margins, enough to re-rate the entire equity story. The GAAP losses are ugly, but they are heavily influenced by non-cash Bitcoin marks. The real risks are execution and funding, not imminent insolvency. For risk-tolerant investors with patience, the asymmetry is compelling: the balance sheet and lease provide a floor, while AI infrastructure execution and Bitcoin recovery provide the upside. This is a speculative BUY sized for volatility.

Overview

This is a Peter Lynch-style analysis of CleanSpark, Inc. (CLSK) as of August 16, 2026, using recent Q3 FY2026 financials, the Sandersville AI data center lease, and Lynch's stock-category, PEG, balance-sheet and story criteria.

The Two-Minute Story

CleanSpark owns power-connected data center sites. It used them to mine bitcoin efficiently, and now it is signing long-term leases to convert some of those sites into AI data centers. On July 14, 2026, it announced a 20-year, $6.6 billion triple-net lease with a high-investment-grade technology tenant for 175 MW at Sandersville, Georgia, with delivery expected in Q4 2027 and a letter of intent covering up to 885 MW in Texas. The simple idea: cheap, grid-connected power is scarce; if CleanSpark can build and lease its portfolio, it could replace volatile bitcoin mining income with long-term, high-margin rent checks. The catch: the AI cash flows are still years away, and the current quarter lost money because bitcoin prices fell.

Stock Category

Do You Understand This Business?

The core is understandable: secure electricity, build data center infrastructure, mine bitcoin today and convert some sites to AI computing tenants over time. An average investor can grasp the logic. The edge is harder: CleanSpark has an energy-operations background and low-cost mining operations, and it secured one major lease. However, the actual AI data center buildout involves specialized engineering, utility negotiations, and financing that are not simple consumer-level businesses.

PEG Ratio Analysis

Lynch's Checklist

Tenbagger Potential

A 10x would take the stock from about $12 to $120, implying roughly a $31 billion market cap. That would require CleanSpark to convert a large share of its 1.8 GW portfolio into leased AI data centers, deliver several Sandersville-style contracts, and finance the builds without massive dilution. That is possible only in a sustained AI infrastructure boom with excellent execution. It is not realistic to count on 10x from here given current losses, high debt, and the fact that the first major lease cash flows are not expected until Q4 2027.

Key Risks

Conclusion

This is a classic Lynch-style conflict between story and numbers. On one hand, CleanSpark owns scarce power-connected sites and has signed a $6.6 billion lease with an investment-grade tenant, which validates the asset-play logic. On the other hand, it is losing money, has negative EPS, no positive PEG, and a leveraged balance sheet. Lynch generally avoided paying for future stories unless he had a strong edge and could explain the business simply. Here the simple story is believable, but the current financials do not provide a margin of safety. Therefore, HOLD is appropriate: it is not safe enough to buy aggressively, but the signed lease and asset portfolio are valuable enough that an outright sell is also not warranted.

Overview

This report applies William J. O'Neil's CAN SLIM methodology to CleanSpark, Inc. (CLSK) as of August 16, 2026. The analysis focuses on earnings growth, technical strength, institutional sponsorship, and market direction, using recent SEC filings, earnings releases, and peer comparisons to determine whether CLSK qualifies as a buyable market leader under CAN SLIM rules.

Financial and Business Overview

CleanSpark is a Bitcoin mining company transitioning into an AI/HPC data center developer. For fiscal Q3 2026 ended June 30, 2026, revenue was $138.0 million, down 30.5% year-over-year from $198.6 million. Net loss was $239.8 million, or -$0.89 per share, compared with net income of $257.4 million, or $0.90 per share, in the prior-year quarter. Adjusted EBITDA was -$113.0 million versus +$377.7 million. The company ended Q3 with $202.6 million in cash, total bitcoin-related HODL value of $814.9 million, total debt of approximately $1.8 billion, total liabilities of $1.9 billion, and stockholders' equity of only $761.3 million. TTM EPS is -$3.78 and forward EPS is -$1.42, so the business is deeply unprofitable on a GAAP basis. The main positive is a signed 20-year, $6.6 billion triple-net lease at Sandersville, Georgia with a high-investment-grade global technology tenant, expected to deliver approximately $330 million in annual NOI, but income is not expected until Q4 2027. The company currently earns effectively all revenue from Bitcoin mining, and AI/HPC revenue remains negligible.

Market Position & Competitive Advantages

CleanSpark has genuine operational strengths: it operates one of the most efficient public mining fleets at approximately 16.2 J/TH, has a history of low all-in mining costs among peers, and controls about 1.8 GW of contracted power across the U.S. Its Sandersville lease validates the strategic pivot toward AI infrastructure, and the related Texas portfolio exclusivity covering up to 885 MW provides a large optional growth pipeline. Management has emphasized modular construction and strong community relationships. However, the competitive position is constrained by negative earnings, heavy debt, Bitcoin price exposure, and execution risk in building AI data centers. Compared with peers, CLSK has a less negative operating margin but is no longer a relative stock-market leader.

Stock Performance

CLSK trades at $12.09 as of August 16, 2026, up 4.95% on the day but still below the 50-day moving average of $14.39 by about 16% and below the 200-day moving average of $12.77 by about 5%. The stock is 48.8% below its 52-week high of $23.61 and 51.1% above its 52-week low of $8.00. Average daily volume is approximately 23 million shares, indicating active trading, but the stock's inability to hold above key moving averages suggests distribution rather than accumulation. The technical picture is weak: lower highs, a declining 50-day average, and a close below the 200-day line.

CAN SLIM Analysis

Key Risks

Conclusion

Under CAN SLIM, CLSK fails the most important criteria. Current quarter EPS is deeply negative, there is no annual earnings growth, and the stock is a relative laggard trading below both the 50-day and 200-day moving averages. The new AI/HPC lease is a legitimate N catalyst, but O'Neil's method requires strong current and annual earnings before considering a purchase. Institutional sponsorship is mixed, and supply/demand appears to favor distribution. For an investor following CAN SLIM, this is not a buyable stock. Existing holders should consider a defensive sell discipline or avoid adding until earnings turn positive and the stock demonstrates true market leadership. Aggressive speculators may view the AI pivot as an option, but that is outside the CAN SLIM framework.

Overview

This is a Warren Buffett-style intrinsic value analysis of CleanSpark, Inc. (CLSK) as of August 16, 2026. It focuses on business understandability, competitive moat, management quality, financial strength, intrinsic value versus market price, and margin of safety rather than short-term price action.

Business Understanding

CleanSpark is a Bitcoin mining company transitioning into an AI/HPC data center developer. It owns and operates data centers in Georgia, Tennessee, Mississippi, Wyoming, and Texas, with about 1.8 GW of contracted power. In July 2026, it signed a 20-year, $6.6 billion triple-net lease for 175 MW at Sandersville with a high-investment-grade global technology company; the same tenant has a letter of intent and exclusivity over up to 885 MW of Texas capacity. The infrastructure side is understandable—power access, cooling, uptime, fleet efficiency, construction—but the dominant current revenue is Bitcoin mining, which is a commodity business tied to Bitcoin price, network difficulty, and halving cycles. That makes near-term economics highly cyclical, volatile, and not comfortably within a conservative circle of competence.

Economic Moat Analysis

CleanSpark has a narrow and largely unproven moat. In mining, it has a cost advantage: industry-leading fleet efficiency around 16.2 J/TH, low power costs near $0.052/kWh in some periods, and among the lowest cash costs, interest costs, and stock-based compensation per Bitcoin among public peers. However, mining itself has no pricing power; any advantage can be competed away. The more interesting moat candidate is in AI/HPC data centers: scarce grid-connected power capacity and a signed 20-year triple-net lease with an investment-grade tenant. Once built, that lease would have meaningful switching costs and long-duration cash flows. But until the data halls are delivered and generating rent—expected no earlier than Q4 2027—it is an option, not an operating asset. There is no brand moat, network effect, or proprietary technology comparable to a true Buffett-quality franchise.

Management Quality

Management has shown operational discipline in mining, conservative disclosure of contracted power, and relatively low stock-based compensation and interest costs per Bitcoin compared with peers. The company repurchased approximately $460 million of stock in fiscal 2026, but it did so using proceeds from 0% convertible notes, increasing leverage rather than using free cash flow. It pays no dividend. The CEO and CFO frame the strategy around capital stewardship and execution, and the Sandersville lease with a high-credit tenant is a significant commercial achievement. Still, the move into large-scale AI data center development is outside management's historical core competency and remains unproven. There are also outstanding class action and derivative legal matters, which require monitoring.

Financial Strength

Financial strength is weak on current GAAP metrics. In Q3 fiscal 2026, revenue was $138.0 million, down 30.5% year over year, with a net loss of $239.8 million and adjusted EBITDA of negative $113.0 million. TTM EPS is approximately -$3.78. Book value per share is $2.97, but stockholders' equity fell from $2.18 billion in September 2025 to $761 million by June 2026. Total liabilities were $1.9 billion, including $1.8 billion of long-term debt, against only $202.6 million of cash and $814.9 million of Bitcoin. Debt-to-equity is over 2x, interest coverage is negative, and operating cash flow was negative in recent periods. The Bitcoin and cash provide a liquidity buffer, but that buffer depends on Bitcoin prices and future financing rather than consistent free cash generation.

Intrinsic Value Assessment

There are no current owner earnings. TTM net income is deeply negative, depreciation is heavy, and the AI buildout will require substantial additional capital. The market value is about $3.1 billion, versus stockholders' equity of $761 million and a price-to-book ratio near 4.1. The bull case rests on the Sandersville lease: management expects roughly $330 million of annual net operating income at nearly 100% margin over 20 years. Discounted at 8%, a $330 million annuity is worth roughly $3.2 billion before subtracting landlord capital costs of perhaps $1.75-$2.1 billion and before execution risk. The Texas portfolio adds real optionality. But those cash flows are years away, and the existing mining business is currently loss-making. At $12.09 per share, the market price already reflects a great deal of successful execution. A conservative fair value requires large haircuts to the lease ramp, financing costs, and mining recovery. I cannot identify a meaningful margin of safety at this price.

Key Risks

Investment Details