CLSK · CleanSpark, Inc. · Stock research

Current Thesis

CleanSpark is the last large pure-play Bitcoin miner still valued on an AI/HPC "neocloud" pivot rather than mining economics and that pivot leg has shifted from re-accelerating to stalling. The catalyst bulls paid up for got a named counterparty on 2026-06-08 (Meta reported in talks to lease part of the ~250 MW Sandersville, GA site), which carried shares to ~$17.24 by 2026-06-18. Five weeks on, the lease is still unsigned, June BTC production slipped to 614 (from 671 in May), and the stock slid into early July (2026-07-07). The trade is now a binary waiting room: a signed Meta lease re-rates the multiple, a "talks ended" headline is a down-20% session, and until either prints the tape drifts on Bitcoin beta.

Bullish and bearish views on CleanSpark, Inc.

The model's bull view on CleanSpark, Inc. (CLSK), in brief: The hyperscaler has a name: Meta reported in talks (2026-06-08) to lease part of the 250 MW Sandersville, GA site for AI workloads a concrete upgrade from the vague "hyperscaler lease talks" that ran the stock +37% in May (2026-05-22). The bear view: The pivot is still a negotiation. Every headline says "in talks" or "ambitions," not "signed." Roughly five weeks after 2026-06-08 there is no announced contract, and a "Meta talks ended" print is a down-20% session that stays live until ink dries. Fundamentals deteriorated… Both cases follow in full.

Bull Case

• The hyperscaler has a name: Meta reported in talks (2026-06-08) to lease part of the 250 MW Sandersville, GA site for AI workloads a concrete upgrade from the vague "hyperscaler lease talks" that ran the stock +37% in May (2026-05-22). An investment-grade anchor tenant is what re-rates the multiple.
• Power is the scarce asset: ~1.8 GW contracted against ~808 MW used leaves roughly 1 GW of grid-interconnected, shovel-ready capacity (GA/MS/TN/WY) that AI demand values well above mining margins.
• Sell-side keeps underwriting the pivot through a bad print: Chardan Buy PT $19 (2026-06-08), Macquarie Outperform $22 (2026-05-13), Maxim Buy $22 (2026-05-12); consensus clusters near a ~$20 average across ~16 analysts, range $14–$27.30.
• Two independent 13Fs accumulated into a rising tape: Aschenbrenner's Situational Awareness "significantly boosted" its stake (2026-05-20) and a second fund "sharply boosted" (2026-05-21/22).
• Sector cluster stays bid: Bernstein Outperform on four Bitcoin miners citing >$90B in AI deals across 3.7 GW of power (2026-05-19); CLSK is the laggard chasing IREN/CORZ/WULF/CIFR/APLD.
• Balance-sheet BTC still compounds: June output 614 BTC lifted treasury to 13,924 BTC (2026-07-07), a hard-asset backstop that grows with Bitcoin.

Bear Case

• The pivot is still a negotiation. Every headline says "in talks" or "ambitions," not "signed." Roughly five weeks after 2026-06-08 there is no announced contract, and a "Meta talks ended" print is a down-20% session that stays live until ink dries.
• Fundamentals deteriorated under the story: fiscal Q2 EPS $(1.52) vs $(0.50) est and revenue $136.4M vs $145.4M est (2026-05-11) roughly a dollar EPS miss on top of a revenue miss.
• Momentum is fading rather than building: shares slid on 2026-07-07, and June production of 614 BTC came in below May's 671, softening the one line that prints real cash.
• Positioning is a two-sided trap. CLSK reappeared on most-shorted "squeeze" lists (2026-06-29) after earlier short-seller sweeps (2026-05-28); elevated short interest against the 13F longs produces violent moves in both directions.
• The marquee buyer is hedged elsewhere. Aschenbrenner is simultaneously short NVDA/AVGO/ORCL $8.5B (2026-05-18), so the CLSK long may be one leg of a power-vs-chips dispersion trade rather than a clean equity vote.
• Bitcoin beta still dominates P&L and price. Without a signed HPC revenue line, a BTC drawdown overrides the Meta narrative outright.

Setup & Price Structure

The June Meta-headline bounce topped near $17.24 (2026-06-18) and the early-July tape drifted back toward the ~$16.50 shelf that marked the 2026-06-08 reaction low. That shelf is the line that matters: it separates a base that can hold into a signing from a failed-breakout unwind of the whole May–June move. Overhead sits the $19–$22 analyst cluster (Chardan $19, Macquarie/Maxim $22), with the tape below it. Unusual options prints flagged on 2026-07-10 read as two-way gamma interest, not a directional signal. The structure is a range-bound waiting pattern on a MATURING theme strength is not the setup here; a confirmed signing or a clean higher-low base above $16.50 is. This is a binary to buy on confirmation or fade on a shelf loss, not an accelerating name to chase blind.

Catalyst Calendar (next 30 days)

• Undated / live: Meta Sandersville lease a signing headline (MW/term/revenue disclosed = binary up) or a "talks ended" headline (binary down). The single most important catalyst; no fixed date, live every session.
• ~2026-08-05 (est.): July BTC production and operational monthly update (first-week cadence). Watch for any first HPC/AI booking language, not just the BTC count; June's update (614 BTC) softened vs May.
• ~2026-08-07 (est.): fiscal Q3 FY2026 earnings (fiscal Q2 printed 2026-05-11; Q3 typically lands early-to-mid August). Blackout-aware from late July avoid fresh entries into the print. The test is whether signed HPC/AI revenue finally appears against a still-negative mining EPS.
• Continuous: Bitcoin spot level, the dominant driver of both reported earnings and daily price.

What Would Change Our Mind

• Bullish confirmation: a signed Meta (or peer hyperscaler) lease with disclosed megawatts, term, and annual revenue; a monthly update showing a first contracted HPC booking; or a weekly reclaim of $17.24 on expanding volume that puts the $19–$22 cluster back in play.
• Bearish confirmation: a weekly close below $16.50 — that loses the Meta-news shelf; a confirmed "lease talks ended" headline; a fiscal Q3 print with still-zero signed HPC revenue and a widening mining loss; or Bitcoin breaking its own support and dragging the complex.
• Neutralizing: continued "in talks" limbo with no signing and BTC rangebound keeps this a stand-aside until the base or the deal declares itself.

Correlation Notes

• Bitcoin spot: highest correlation mining revenue, treasury mark, and sentiment all key off BTC; a drawdown overrides the pivot narrative.
• BTC-miner-to-AI peers: IREN, CORZ, WULF, CIFR, APLD trade as a cluster on the "miners monetize power for AI" theme (Bernstein, 2026-05-19); CLSK is the laggard beta chasing them.
• Neocloud / AI-datacenter infrastructure: shares the power-scarcity bid with GPU-cloud names, so an AI-capex wobble hits the whole group at once.
• Dispersion overlay: Aschenbrenner long CLSK / short NVDA-AVGO-ORCL $8.5B (2026-05-18) can decouple CLSK from the chip complex on days the power-vs-silicon trade is in favor.
• Short-interest mechanics: an elevated short base (2026-06-29 squeeze lists) means sharp up-days can be covering rather than fresh conviction; respect a signed-deal move, discount a purely mechanical squeeze.