The Bitcoin Treasury Model Is Splitting: Satsuma's Liquidation and the Rise of Permanent Capital

Satsuma Technology is dead. The shareholder vote hit 90% in favor of liquidation on July 21, 2026. The company held exactly 668 bitcoin, worth around $43.5 million at the time. This is what a pure-play Bitcoin treasury looks like when the market stops paying premiums. The mathematics are brutal. A public vehicle whose sole asset is Bitcoin and whose sole strategy is holding is nothing more than a wrapped coin with extra legal fees. Satsuma just proved that the wrapper is worth nothing if the model cannot produce yield. The failure is not an isolated event. VanEck's Matthew Sigel has observed that many pure-play Bitcoin treasury companies are now abandoning the accumulation strategy. Strategy (MSTR) is facing real pressure as its leveraged credit model collides with a changing market structure. But this is not the end of the Bitcoin treasury concept. It is the beginning of a split. Two successor models are emerging from the wreckage. The first is the credit model, perfected by MSTR — borrowing low, buying Bitcoin, and issuing equity at a premium to extend the flywheel. The second is the permanent capital model, represented by new vehicles like Orange Juice and Tether-backed Twenty One Capital. These two structures are as different as a hedge fund and a corporation, but both are answers to the same problem: how do you accumulate Bitcoin without depending on market sentiment? The timing reveals how fast the ground is moving. Orange Juice and Twenty One Capital announced their structures within roughly a week of each other. That is not a coincidence. Both vehicles emerged in the shadow of Satsuma's failure, and both are designed to avoid the exact flaw that killed it. The old model was elegantly simple. Raise equity, buy Bitcoin, watch the stock trade at a premium to net asset value. The premium was the alpha. Investors loved it because it delivered leveraged Bitcoin exposure without the need to custody anything. The company loved it because the premium allowed it to issue new shares at a price that was accretive to its Bitcoin-per-share metric. In a bull market, the flywheel spins. In a bear market, it reverses. When the premium compresses, new share issuance becomes dilutive. When issuance becomes dilutive, the company slows its purchases. When purchases slow, the growth narrative dies. And when the narrative dies, the premium dies harder. This is the reflexivity trap that Satsuma could not escape. Satsuma had no leverage, no credit, and no yield. It was the purest expression of the old model. And the market told it to liquidate. That is a signal, not a noise. The 90% vote means shareholders have concluded that a non-yielding Bitcoin holding company is a negative carry asset. They prefer the underlying Bitcoin directly, or they want a business model that enhances returns through operations. The vote was a democratic verdict on a model that had no answer to the question: what does this company actually do, other than hold coins? MSTR, by contrast, is a balance sheet arbitrage vehicle. It issues convertible notes at interest rates that have historically been near zero, uses the proceeds to buy Bitcoin, and then uses its equity premium to do it again. The key metric is BTC yield: the growth rate of Bitcoin per diluted share. In a rising market, this works exceptionally well. The premium on MSTR stock allows it to issue shares at a price that implies a lower actual BTC cost per share, then use that capital to add to the treasury. The result is a continuously rising Bitcoin-per-share ratio. But there is a hidden cost. The interest on the notes is real, and the eventual maturities are real. If Bitcoin enters a prolonged sideways market, the premium evaporates, and the company faces a choice: issue equity at a discount or liquidate some of its Bitcoin. Both are wealth destruction. This is not a theoretical risk. I have been in this game long enough to know that leverage that looks free in a bull market becomes crippling in a consolidation phase. My own playbook, developed after the 2022 Terra collapse, relies on monitoring the correlation between a treasury company's equity premium and its debt service obligations. When the premium starts moving independently of Bitcoin price, I treat that as a warning sign. Let us put some numbers on the credit model to show why it is so seductive and so fragile. Suppose MSTR holds 300,000 BTC and has 150 million diluted shares. That is 0.002 BTC per share. If the stock trades at two times the value of its underlying BTC, the market is paying $200,000 for a share backed by $100,000 in Bitcoin. If the company issues five million new shares at that premium, it raises $1 billion. At $100,000 per BTC, that buys 10,000 BTC. The new share count is 155 million, and the Bitcoin balance is 310,000. The new BTC per share is 0.0023, a 15% BTC yield. Now imagine the premium falls to 1.2 times. A new issue of the same size raises only $600 million, buying 6,000 BTC. The BTC yield drops to 5%. And if the stock trades at a discount to net asset value, issuance destroys value. At that point, the model stops working entirely. The market is not pricing MSTR on its BTC per share; it is pricing the probability that the premium stays high enough to sustain the flywheel. Now consider the permanent capital model. Orange Juice is a newly formed vehicle designed to hold Bitcoin forever while generating operating cash flows from other businesses. The goal is to use those cash flows to buy more Bitcoin without ever tapping the equity markets. This is a fundamental structural break from the old model. Instead of relying on the stock price to fund accumulation, the vehicle relies on actual income. Think of it as Berkshire Hathaway with a Bitcoin treasury. If the operating business produces $50 million in free cash flow, that cash can be deployed into Bitcoin without diluting a single shareholder. No interest payments. No issuance risk. The accumulation is funded by real economic output, not by capital market arbitrage. Lyn Alden and Jeff Booth are backing this design, which gives it credibility among the Bitcoin intellectual class. The name itself, Orange Juice, is a deliberate homage to a permanent compounding vehicle. The structure is designed for a 50-year time horizon, not a 50-day trading cycle. Twenty One Capital is the more aggressive variant. It is backed by Tether. This is where the scale changes. Tether's USDT issuance generated billions in profits during the last cycle. If those profits are channeled into Bitcoin acquisition, Twenty One Capital could become the largest buyer of Bitcoin without ever issuing a public share. That is a structural asymmetry that MSTR cannot match. MSTR's cost of capital is the interest rate on its convertible notes plus the equity dilution from its ATM programs. Twenty One Capital's cost of capital is essentially zero, because Tether's profit engine is a cash printer. The market is starting to price this. The compression in MSTR's premium is not a signal of Bitcoin weakness. It is a signal of model risk. The market is asking a simple question: why should I pay a half-billion-dollar premium for a company that has to borrow money to buy Bitcoin, when I can invest in a vehicle funded by a stablecoin oligopoly? This brings us to the contrarian view. The rapid rise of the permanent capital model is not a guaranteed victory. There are three large blind spots. First, permanent capital is a promise, not a protocol. A company that generates cash flow can also waste it. The management team has discretion over how the cash is allocated. If they make bad acquisitions or if the operating businesses fail, the model collapses. Satsuma failed because it had no cash flow. Orange Juice could fail because its cash flow is mismanaged. The difference is that the failure takes longer and is harder to detect. I have always believed that yield without protocol is just delayed loss. In this context, the protocol is the governance structure that ensures cash flows actually go into Bitcoin rather than into executive compensation or bad deals. Second, Tether's backing is a double-edged sword. Twenty One Capital's advantage is Tether's profit engine. But that engine carries reputational and regulatory risk. Tether has been subject to multiple investigations, and its balance sheet has historically been a black box. If Tether faces a compliance breach or a run on USDT, its ability to fund Twenty One Capital evaporates. Investors in Twenty One Capital are not just exposed to Bitcoin; they are exposed to the solvency of a stablecoin issuer. That is a trust assumption, not a code assumption. I trade the ledger, not the hype cycle. On this ledger, Tether's liabilities are still opaque. Third, the market may be over-rotating. The pure-play model failed because it lacked yield. But the credit model is not inherently broken; it is cyclical. MSTR's premium is highly correlated with Bitcoin's price momentum. If Bitcoin resumes its upward trend, MSTR's model will work again, and the premium will expand. Satsuma's liquidation is not a death sentence for MSTR. It is a warning that companies without any yield mechanism will be penalized. The permanent capital model is a response to that, but it is not proven through a bear market. The only test that matters is how each model behaves when Bitcoin drops 60% and stays down for a year. Satsuma would have died. MSTR might survive if it has enough liquidity to service its debt. Orange Juice might survive if its operating business is solid. But none of them will survive if managers panic and make capitulatory decisions. So what does this mean for MSTR? It means the market is shifting its attention from the asset side of the balance sheet to the liability side. The number of Bitcoin held matters, but the cost of holding that Bitcoin matters more. Investors are starting to calculate the interest expense per Bitcoin, the dilution per share, and the cash flow yield of the entire vehicle. This is a healthy maturation. The next bull run will reward treasury companies that can accumulate Bitcoin without destroying shareholder value, and it will punish those that rely solely on the kindness of the equity markets. I have applied the same filter to treasury companies that I used in 2017 when I audited 50 ERC-20 whitepapers. The question then was: does this token have a revenue model? The question now is: does this treasury company have a cash flow engine? If the answer is no, the equity is just a leveraged Bitcoin future with no expiry. The market has started to figure this out. The winners will be the ones that generate yield, either through credit operations, options sales, or actual business profits. The losers will be the ones that simply hold coins and hope. Volatility is the tax on undiscerned capital. The market is finally learning how to pay it. As we move into the next phase, watch the liability structure, not just the asset headline. Ask whether the entity can accumulate Bitcoin in a bear market without issuing equity at a discount. Ask whether the operating cash flow is sufficient to service debt. Ask whether the management team is disciplined enough to keep the protocol intact. The market pays for clarity, not complexity. The treasury models that survive will be the simplest to understand and the hardest to break. That is the takeaway. The shift from pure-play to permanent capital is not a rejection of Bitcoin. It is a rejection of lazy capital. The market is demanding a reason to own the wrapper. If you cannot explain where the cash flow comes from, you are holding a delayed loss.