CleanSpark, one of America’s largest publicly traded Bitcoin miners, produced 593 BTC in August but sold 821 BTC — deliberately shrinking its treasury for the first time in a while. The company’s latest operational update is a masterclass in how modern miners actually make money when Bitcoin trades near $79,300.
By Michael Nguyen | September 9, 2026
The Hook
If you own Bitcoin or any miner stocks, August’s numbers from CleanSpark (NASDAQ: CLSK) deserve ten minutes of your attention. The company released its August 2026 operational update on September 8, and on the surface it looks like bad news: the miner sold more coins than it produced, and its Bitcoin stockpile shrank from 13,931 BTC at the end of July to 13,703 BTC at the end of August.
But dig one level deeper and the picture flips. Mining economics actually improved last month — meaningfully. Understanding why helps you judge every mining company on the market, because the gap between winners and losers in this industry is now enormous.
On-Chain Evidence: The August Numbers
Start with production. CleanSpark mined 593 BTC in August, up slightly from 586 in July — and it managed that increase even though its average operating hashrate (the computing power actually running) slipped from 38.6 to 38.3 exahashes per second. Think of hashrate as how many workers you have on the mining assembly line. Producing more with slightly fewer workers means the machines got more efficient or the environment got friendlier. In this case, it was mostly the environment — more on that below.
The scale of this operation is easy to gloss over, so here it is as a list, straight from the company’s Sept. 8 update:
- 593 BTC produced in August — an average of 19.12 BTC per day, peaking at 20.40 BTC on the best day
- 4,903 BTC produced year-to-date
- 201,269 machines deployed as of Aug. 31
- 50 EH/s operational hashrate (the peak concurrent capacity of installed, energized miners)
- Peak fleet efficiency of 16.07 joules per terahash — a measure of how much electricity each unit of computing power burns
- 808 megawatts of maximum concurrent power use, backed by 1.8 gigawatts of contracted capacity
Now the treasury. CleanSpark ended August holding 13,703 BTC, and here is the detail most headlines will skip: of those, about 3,951 BTC — roughly 29% of holdings — were posted as collateral or recorded as receivables connected to derivative transactions. The company is not just HODLing; it is actively using its Bitcoin as financial machinery.
The Core Conflict: Why Sell More Than You Mine?
The 821 BTC sold in August breaks into three buckets per the update: 77 BTC sold at spot prices, 500 BTC sold through call exercises, and 244 BTC tied to a delta-neutral basis trade — a strategy that profits from differences between the futures price and the spot price while staying neutral to direction. The blended average came out to $65,420 per BTC, a figure CleanSpark calculates using net proceeds plus premiums.
That average is below where Bitcoin trades today, near $79,300, with Ethereum around $2,501 and Solana at about $104. But remember what August looked like: the month opened with Bitcoin lower and rallied hard in the final third. CleanSpark’s sales programs executed throughout the month — they are liquidity tools, not market calls. Miners sell continuously to pay for power, payroll, and growth; a treasury that only goes up is a treasury that eventually cannot pay its electricity bill.
The more important number for the whole industry is hashprice — the revenue a miner earns per unit of computing power each day, tracked by Luxor’s Hashrate Index. Hashprice opened August at $31.63 per petahash per day and closed at $39.33, averaging $34.63 — its highest monthly level since May, according to Luxor’s monthly lookback. That late-month Bitcoin rally is exactly what lifted production revenue for miners without adding a single machine.
Luxor’s analysis also quantified the profitability divide. Fleets running between 14 and 19 joules per terahash — where CleanSpark’s peak efficiency of 16.07 J/TH lands — earned about $87 per megawatt-hour of energy revenue in August, while the industry’s average power cost was estimated near $48 per megawatt-hour. In plain English: modern machines covered their electricity bills with room to spare, while older fleets and expensive power contracts stayed underwater. CleanSpark’s economics benefited, though the company’s release does not disclose its actual average fleet efficiency or power price, so full-month net profitability cannot be confirmed from the update alone.
Market Implications: What This Means for Your Portfolio
If you hold mining stocks, August taught the same lesson the industry has been repeating all year: efficiency and power contracts are the moat. When hashprice rises, everyone benefits — but the efficient fleets keep the spread. When hashprice falls, only the lean operators survive. A miner’s machines-per-employee pitch matters less than its joules-per-terahash and its dollars-per-megawatt-hour.
Second, watch the derivatives activity. Nearly a third of CleanSpark’s treasury is now working capital — collateral and hedging instruments rather than idle coins. That is sophisticated treasury management, and it signals how public miners are evolving from “we buy and hold” into something closer to a Bitcoin-native financial firm. It also introduces counterparty and collateral risk that monthly updates do not fully disclose, so treat aggressive treasury engineering with the same skepticism you would apply to any leverage.
Third, the treasury drawdown is not distress. Selling 228 BTC net in a month against a stockpile of nearly 14,000 coins is a rounding error — and doing it via structured trades that generate premiums is smarter than dumping spot.
The Verdict
CleanSpark’s August update reads like a company in control: slightly more production on slightly less average power, rising hashprice tailwinds captured, and a treasury being put to work rather than left idle. The open question — actual net profitability for the month — will only be answered in the quarterly filings, since the latest operational release omits real power costs and fleet-wide average efficiency.
For regular investors, keep it simple: in Bitcoin mining, the winners are the ones who spend the least to produce each coin. August’s numbers say CleanSpark remains in that group. The bar to watch going forward is whether hashprice holds above the mid-$30s as the industry’s efficiency race keeps accelerating.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
821 sold vs 593 mined at 79k says everything about what post-halving economics look like now. nobody is immune
producing 593 with hashrate down from 38.6 to 38.3 EH? luck adjusted that is basically flat, would not read too much into it
29% of the stack posted as collateral for derivatives is the actual headline here. That is a hedge fund with mining rigs attached.
16.07 J/Th fleet efficiency is elite tbh, most of their competitors would kill for that number
A hedge fund with mining rigs is the honest description of every scaled miner now. The open question is whether that derivatives book is hedging or just leveraged BTC exposure with extra steps.
q2 filing said hedging program and never showed the greeks. that vagueness is doing heavy lifting for the delta neutral read
clsk sells 821 while mining 593 and its strategy. a smaller miner does the same and its distress. scale gets you a very different headline
hard agree. riot trimmed holdings last quarter too and nobody wrote a distress piece. headline writers just love a sold more than mined hook
The 29% of holdings posted as collateral is the real detail. 13,703 BTC sounds like a fortress until you read the fine print on those derivatives.
29 percent posted is standard treasury ops at this point, half the sector does it. what matters is whether the derivatives book is delta neutral or just levered btc exposure in a trench coat
right, and at 16.07 J/TH they can actually survive $79k btc. the collateral game is how you squeeze more juice out of each coin when margins are thin
Sabine the collateral isnt new tho, they flagged the derivatives book in the q2 filing. selling 821 to trim leverage a bit is actually the responsible read
13,703 BTC with a fifth posted as collateral is fortress adjacent, q2 disclosure or not. margin call season tells the real story
selling 821 at 79k while efficiency sits at 16.07 J/TH is just running the business. the part id watch is hashrate dipping to 38.3 while production rose, luck flattered them a bit
luck adjusted flat with hashrate down 0.3 EH is fine but wait for the quarterly efficiency numbers. if 16.07 J/TH is fleet average they are buying S21 pros in bulk
16.07 J/TH fleet average already tells you the S21 pro story. that efficiency capex is exactly why they can sell 821 coins without flinching at 79k btc
they been swapping fleet since the georgia expansion. 16.07 fleet avg mid upgrade means the S19s are basically gone
38.3 EH while producing more btc than july means luck padded the 593. the real tell is selling 821 at 79k to keep the machine running, business not distress
Sold 821 at 79k, still holds 13,703, fleet at 16.07 J/TH. thats just treasury ops, the distress framing is lazy