CleanSpark holds 12,205 BTC on its balance sheet. It also sold 9,400 BTC worth of call options in a single quarter. PowerCompute borrowed $21.89 million against 307 BTC, with a collar that transfers all upside above $93,500 to its lender. USBC, a bank, has pledged 34.1% of its reserves as collateral for credit lines. These are not on-chain events. They are off-chain obligations that determine how many Bitcoin will actually hit the market. And the market is still pricing these companies as passive accumulators.
That gap is the thesis of this article. I have spent fifteen years watching digital assets migrate from speculative retail instruments to institutional balance-sheet tools. In 2024, I led a micro-research team dissecting the first two weeks of spot Bitcoin ETF flows. We found that BlackRock's IBIT and Fidelity's FBTC were not retail-driven; they were rebalancing vehicles for traditional equity funds. That taught me a lesson I now apply to every corporate disclosure: the headline position is not the exposure. The footnotes are. The footnotes in this case are dense with derivatives.
The Context: Post-Halving Financial Engineering
In September 2024, Bitcoin trades near $78,767. The halving in April cut miner block rewards from 6.25 BTC to 3.125 BTC. Revenue per petahash dropped by roughly 50%. Miners face a capital expenditure cliff: new ASICs, rising electricity costs, and a coin price that has not yet fully offset the reward reduction. CleanSpark, PowerCompute, and USBC are not responding by selling coins outright. They are doing something more sophisticated—and more dangerous. They are using options, collars, and collateralized lending to convert their Bitcoin reserves into cash or yield while retaining a carefully managed portion of upside. The accounting looks like a balance sheet optimization. The market interprets it as accumulation. Neither is accurate.
Let me walk through the three structures, because each one distorts the available supply in a different mechanical way.
Core: The Mechanics of Conditional Supply
CleanSpark's Covered Call Overlay
CleanSpark's treasury holds 12,205 BTC. In the same quarter, it sold call options with a strike of $76,383 against a notional of 9,400 BTC. The average spot price at writing was $68,766. That creates a static yield of approximately 11.1%—calculated as ($76,383 – $68,766) / $68,766. The company collects this premium upfront. In exchange, any appreciation above $76,383 belongs to the option buyer. At the current spot of $78,767, those options are in-the-money by $2,384 per BTC. If the owner exercises, CleanSpark must deliver Bitcoin from its reserve. This is not a prediction; it is a contractual obligation. The market sees CleanSpark as a holder of 12,205 BTC. The options reduce that effective float to a variable number that depends on price. At $80,000, the company may need to deliver 9,400 BTC. At $90,000, it delivers the same number, but the opportunity cost is higher. The so-called treasury is, in fact, a partially short call position.
I have seen this pattern before. In my 2020 DeFi Summer work, I deployed automated yield strategies on Compound and Aave. I learned that any covenant that earns yield by selling optionality has a hidden cost: it caps the tail of the distribution. CleanSpark's strategy is not novel in traditional finance. The novelty is applying it to the largest cryptocurrency without recognizing that it changes the supply narrative. The 9,400 BTC figure is quarterly flow, not closing inventory. The company may roll these options repeatedly, which means the constraint is dynamic. Every roll is a new decision to sell more upside.
PowerCompute's Collar Loan
PowerCompute's structure is more explicit. It collateralized 307 BTC and borrowed $21.89 million at 6.5% interest. The loan is wrapped in a collar with three price regimes. If BTC stays between $71,112 and $93,500, PowerCompute retains all upside. If BTC drops below $71,112, the company can either deliver the 307 BTC in full or walk away via a non-recourse default. If BTC rises above $93,500, the lender captures every dollar above $75,000. The 307 BTC is not just collateral; it is a call option sold to the lender above the strike, and a put option sold to the borrower below the floor. The company also paid $3.765 million to unwind a prior collar, folding that cost into the new loan. That is a 19.8% increase over the borrowed amount. The entire structure is a re-financing and leverage replay, not a one-time hedge. The rolling term is set for September 24, 2024—just weeks away. At the current price, PowerCompute sits in the favorable middle band. But the distance to $93,500 is only 18.7%. A modest rally announces the forfeiture of all upside above that level.
This is what I mean by conditional supply. The 307 BTC are not free-moving inventory. They are a collateral pool that can be seized or surrendered at determined price points. The lender has a call option on the Bitcoin. The borrower has a put option. Both sides have embedded incentives to trade on the price. When the market reaches those barriers, the Bitcoin will move—either from the borrower to the lender or into the open market to rebalance. That movement is not a discretionary sale; it is an algorithmic consequence of the contract.
USBC's Pledged Reserves
USBC reported that 34.1% of its Bitcoin reserves are pledged as collateral. It has also secured a credit line against 478 BTC. This is the third variant: the bank does not sell its coins, but it has transferred the right to liquidate them to a creditor if the collateral ratio falls. The market treats these BTC as bank-held. The creditor treats them as a claim. The actual availability for spot purchases is a function of the BTC price and the loan-to-value covenants. The supply is real until it isn't.
Taken together, these three structures produce a systemic effect that the market has not priced. The clean narrative of "corporate treasury = buying pressure" breaks down. Every option, collar, and pledged reserve creates a shadow inventory that tends to be released at high prices, not low ones. Call options at $76,383 are more likely to be exercised when spot rises. Collars at $93,500 are more likely to transfer upside to the lender when BTC rallies. Pledge ratios are more likely to be breached when prices fall. This is a martingale of hidden supply: the probability of forced selling increases as price moves in either direction, but the magnitude of the selling pressure is concentrated in the upper and lower tails.
Contrarian: The Miner Is the New Short
The contrarian angle is not that miners are selling too soon. It is that the entire category of "miner as net buyer" is now a fiction. CleanSpark bought 244 BTC and sold 250 BTC in the same month, according to the data. That is near-market-maker activity, not accumulation. The disclosed numbers—9,400 BTC in calls, 307 BTC in collateral, 34.1% pledged—reveal a sector that has shifted from price-taking producers to delta-neutral volatility sellers. The mining industry has become a shadow derivatives desk. That transformation is not reflected in the valuation models that analysts apply to miners. Most models still use a simple formula: produced coins minus operating costs minus treasury additions. They ignore the conditional liabilities embedded in options and loans. As a result, the true net exposure of a miner like CleanSpark is overstated on the long side and understated on the short side. The balance sheet says hold. The derivatives say sell.
I have seen this disconnect cause catastrophic misinterpretation before. In 2022, I reverse-engineered the TerraUSD collapse and published a report on systemic fragility in algorithmic stablecoins. The surface metrics—hundreds of millions in reserve transparency—looked robust. The mechanics showed that the stability mechanism was a reciprocal spiral of supply expansion and leverage. Miners today are not stablecoins, but the principle holds: when an entire sector relies on structured products to maintain cash flow, the consensus narrative becomes the last thing you should trust. The disclosure is not fraudulent. It is merely complex enough to be opaque. CleanSpark's auditors can point to line items. But the aggregation of these line items across the industry creates a false sense of total exposure. One company's collateral is another's option position. Merging them into a single metric is meaningless.
There is a further twist. The current bull market narrative depends on the assumption that public Bitcoin companies are reducing the free float. MicroStrategy and similar entities buy billions of dollars of BTC and hold. Miners were assumed to follow that path. Now the disclosed data shows that miners are not holding—they are renting their coin to generate yield. Every covered call transfers upside to a counterparty. Every collar loan transfers liquidation rights to a lender. The long-term holder becomes a short-term lessor. This shifts the demand curve from inelastic accumulation to elastic renting. The market does not yet price that elasticity.
The takeaway is not that these strategies are inherently wrong. Covered calls reduce downside volatility. Collar loans provide cheap capital. Pledged reserves increase liquidity. The problem is the narrative mismatch. An investor buying municipal bonds expects semiannual coupons and principal repayment. An investor buying a miner stock expects commodity exposure. What they are getting is a complex amalgam of commodity exposure, embedded short options, and credit risk. The risk is not disclosed in the first-impression metrics. It is hidden in the 10-Q footnotes and the OTC contract files. Survival is the ultimate metric of a robust system. And the system here is the financial architecture of the public Bitcoin treasury complex. It will survive as long as it can service its derivative obligations. But it will survive differently than the stories we tell.
Takeaway: The Supply Is a Function of Price
I am not predicting an imminent collapse. The structures are legal, marked to market, and managed by professionals. But the next time you read that a mining company has increased its Bitcoin treasury, ask a different set of questions: What percentage of that treasury is pledged? What percentage is covered by short calls? At what price will the company be forced to deliver coins? The answers will tell you more about the future supply than any on-chain metric. Bitcoin's hard cap of 21 million is a total supply ceiling. The float is a dynamic derivative of price and contract terms. The industry has built a layer of conditional supply that moves in response to strikes and barriers. This is not a defeat. It is a maturation. But the maturation comes with a cost: the market must now treat miner hodling as a probabilistic obligation, not a static fact. When the price rallies to $93,500, PowerCompute's lender will call the upside. When it drops to $71,112, the collateral will be swapped. Every contract is a promise that changes the supply curve. The question is not whether these promises will be kept. They will. The question is whether the market has the sophistication to protect itself from the noise. If you thought the Bitcoin float was just the sum of all wallets, you have not been reading the footnotes. The conditional supply is coming. You have been warned.