A hundred dollars is a promise. STRC has stopped keeping it.
Strategy's floating-rate perpetual preferred โ the "Stretch" line in a family that now runs from STRK through STRF, STRD and STRC โ has been trading below its $100 par value. The company's answer, disclosed this week, was a conditional daily dividend.
Daily. Not quarterly. Not monthly.
Set the two facts beside each other and the shape of the thing resolves. An instrument engineered to sit at par is sitting under par. The issuer has responded by raising the frequency of the payment that is supposed to hold it there.
That is not generosity. That is maintenance.
I have spent enough years reading capital structures the way I read Solidity โ line by line, hunting for the clause that was written to be skipped โ to know the load-bearing word in any disclosure is usually the smallest one. Here it is "conditionally." Four syllables that quietly convert an obligation into an intention.
The source material behind this story is thin. Two facts, no sourcing, no figures. That thinness is itself informative. When a company announces a dividend, the market wants three numbers: how much, how often, and out of what. Only the middle one was answered.
The instrument, and the machine it feeds
Start with what STRC actually is, because most coverage of Strategy still treats the company as a single ticker. It is not.
MSTR common is a leveraged call option on bitcoin with a software business bolted on. The convertibles are cheap-duration financing with an equity kicker. The preferred series is something else entirely: a claim that ranks ahead of common in the capital stack, pays a coupon, and โ critically โ never matures.
Perpetual preferred has no maturity date. There is no principal return event. The only exit is the secondary market. Par value of $100 is a benchmark, not a guarantee. It anchors liquidation preference and redemption math, and it gives the instrument a psychological center of gravity that income buyers can underwrite without doing original credit work.
The "Stretch" design appears to do more than that. If the name means what it implies, the coupon floats inversely to price: below par, the rate rises; above par, the rate falls. A self-correcting anchor. The point of such an anchor is legibility โ to make a crypto-adjacent credit instrument look like something a cash manager, an insurance mandate, or a short-duration income fund can hold without explaining bitcoin to a risk committee.
That design carries an assumption nobody writes down. Anchors hold only while the thing they are attached to holds.
Now widen the frame to the machine. The bitcoin treasury company model runs on a flywheel: a premium of market value over net asset value, an issuance into that premium, a purchase of bitcoin with the proceeds, an expanded narrative, a sustained premium. Each turn makes the next turn cheaper. The premium is not a valuation opinion; it is a financing subsidy, and the subsidy is the product.
Strategy defined this model. Metaplanet localized it to Japanese capital markets. A hundred smaller imitators are attempting it with less access, worse pricing, and no software business to point at during earnings calls. Every one of them is running the same reflexive engine, and every one of them will discover that reflexivity is symmetric.
The coupon that chases its own price
A daily dividend on a perpetual preferred is unusual to the point of being a signature. Almost every listed preferred in the US market pays quarterly. Some pay monthly. Daily accrual exists in money market funds, in repo, in overnight facilities โ instruments where the holder's mental model is a balance that grows, not a security that trades.
That is the engineering intent. Daily payment manufactures the perception of continuous cash flow. It compresses the interval between owning and being paid to nearly zero. For a holder deciding between a Treasury bill and a bitcoin treasury company's preferred, the daily cadence is a psychological bridge.
Frequency is not solvency. Paying more often does not create more to pay with. What daily payment does do is raise the operational and cash-smoothing burden on the issuer and โ this is the part that rarely gets said โ it raises the cost of stopping. A quarterly payer that suspends a dividend delivers one shock. A daily payer that suspends delivers three hundred and sixty-five small ones, then a narrative collapse.
The daily dividend is a liquidity illusion: it makes a perpetual claim feel like a deposit while leaving the duration risk exactly where it was.
The arithmetic of a promise
Preferred dividends are a cash obligation ranking ahead of common. Strategy's operating business โ enterprise analytics software โ generates revenue that is real, recurring, and nowhere near the scale of its annualized preferred obligations across the STR stack. That gap does not make the structure fraudulent. It makes it dependent.
Cash to pay a preferred dividend can come from three places: operations, asset sales, or new issuance. Operations contribute a fraction. Selling bitcoin would break the narrative the entire capital structure exists to serve. Which leaves issuance โ the same capital markets currently pricing STRC below par.
State it plainly, because the plain version is the one that matters. The coupon is not being paid by the business. It is being paid by the next buyer of paper.
That is a legitimate model while the window is open. It is a fragile model when the window narrows, and a terminal one when it closes. The technical term for a structure that pays existing obligations out of new issuance is not "yield." It is duration-matched dependence.
Accounting complicates this in a way that flatters the story. Under fair-value treatment of digital assets, bitcoin appreciation flows straight into reported earnings, which makes the income statement look like it is generating the capacity to service debt. It is not. Mark-to-market gains are not distributable cash, and a dividend is a cash event. The number that matters sits three statements away, in the financing activities section, and almost nobody reads that far.
Conditional is a legal term, not an adjective
"Conditionally" is doing enormous work in a nine-word announcement, and nobody should let it pass.
In preferred stock documentation, conditionality points to a small number of specific mechanisms, each with a different consequence for the holder. If the series is non-cumulative, unpaid dividends do not accrue โ they evaporate, and the holder's claim is extinguished by board discretion. If the board retains a deferral right, the cash-flow timeline becomes a management decision rather than a contractual certainty. If payment may be made in kind โ shares instead of cash โ then the dividend is real on the income statement and dilutive on the cap table.
Each of these is disclosed. None of them is announced. The press release says conditional; the certificate of designation says what conditional means. Those are different documents, and only one of them is legally binding.
In early 2017 I spent three months manually auditing the contracts behind a DAO protocol called EthicChain. Twelve critical reentrancy vulnerabilities. Four million dollars of user funds exposed to a single unguarded external call. What struck me afterward was not the bug. It was that the bug lived in a function everyone had read and nobody had understood. The failure was not technical. It was attentional.
The same failure mode governs preferred stock. Holders read the yield. Almost nobody reads the waterfall. Audit the algorithm, not just the code โ and in a structured preferred, the algorithm is the payment priority, the cumulation clause, and the deferral right. Those three sentences decide whether you own a bond or a hope.
Who actually holds this paper
Here the analysis has to leave the balance sheet, because the balance sheet will not explain behavior.
The buyers of a daily-pay preferred are not speculators. They are income desks, family offices, wealth advisors managing retired clients, corporate treasury managers operating under mandates that prohibit equity volatility. They were sold a bond-like thing with an equity-like story attached. Daily accrual told them it was safe. The $100 anchor told them where it would trade.
This is the sociological core of the bitcoin treasury trade, and it is the part the industry least wants to discuss. The model did not scale by converting bitcoin skeptics into believers. It scaled by packaging bitcoin exposure into instruments skeptics could hold without admitting what they were holding. Convertibles for the volatility funds. Preferred for the income funds. Common for everyone else.
That is a translation layer, and translation layers are honorable work โ I spent four years doing it between institutional desks and protocol teams, and I believe in it. But translation has a failure condition. When the underlying changes character, the translation must change with it. A preferred share that behaves like equity while being sold as fixed income is a mistranslation, and the people holding it will be the ones who discover the error.
A structure that requires the holder not to understand it is not a structure. It is a sales channel.
The thermometer, not the price
Here is the analytical move most coverage misses. MSTR common is the wrong instrument to watch.
Common equity in a leveraged bitcoin vehicle is a call option. It rises on volatility, on narrative, on reflexivity. It can look healthy while the credit underneath deteriorates, because optionality is insensitive to small shifts in default probability.
Preferred is a credit instrument. It is sensitive to exactly what common ignores: coverage, refinancing access, and the market's estimate of the issuer's willingness to pay. When the credit instrument breaks par while the equity holds, the market is transmitting information the equity cannot carry.
Watch the whole matrix, not the single line. If STRK, STRF and STRD follow STRC below their anchors, this is not an instrument-specific quirk. It is a term-structure repricing of Strategy credit โ the market renegotiating the price of the entire financing stack at once.
Then follow the thread upstream. Strategy has been the most identifiable marginal buyer of bitcoin in public markets. Its purchases are financed by the same window that prices its preferred. Financing friction does not stop the flywheel; it slows it. Slower accumulation weakens the narrative that supports the mNAV premium, and a narrower premium raises the cost of the next issuance. The mechanism is reflexive in both directions. That symmetry is the whole risk.
The contrarian test
The consensus reading of STRC breaking par is distress. Test it before accepting it.
Consider that this market has no history. There was no listed bitcoin treasury preferred before 2024. No comp set, no spread history, no basis for asserting that $100 was ever the correct clearing price. Par may have been a fiction from the first trade, and the market may simply be the first participant honest enough to say so.
Consider also the rate environment. A perpetual with a floating coupon, issued by a single-asset company with no operating cash flow to cover it, should trade below par in a higher-for-longer regime. That is not a signal of imminent failure. That is a spread doing its job.
So where is the blind spot? It sits in the assumption that the label matters. Distress and correct repricing produce identical consequences for the holder: exit at a discount, coupon must rise, refinancing gets more expensive. The narrative distinction is irrelevant to the cash flow and irrelevant to the calendar. What matters is not whether we call it a crisis. What matters is when the next issuance has to clear, and at what price.
The second blind spot is attentional. Everyone is watching Saylor and the bitcoin chart. Almost nobody is watching the indentures, the coverage ratio, and the payment-in-kind clause. The risk in this trade is not drawn on a price chart. It is printed in a filing most holders will never open.
I have made this mistake before, in a different form. After Terra collapsed in 2022 I withdrew for six weeks and analyzed fifty failed DeFi protocols โ not for technical flaws, but for cultural hubris. The pattern was identical across all fifty. Nobody lost money because a yield was too high. They lost money because an anchor was presented as a mechanism when it was only a mood.
What to watch
Three signals, in order of weight.
STRC's price against $100 โ whether the break deepens or repairs. The payment method disclosed in the next 8-K โ cash, or in kind. And whether the rest of the STR stack follows, because a single line below par is a data point and a whole curve below par is a verdict.
Bitcoin may well go up. Over a long enough horizon, the case is defensible. But the question this instrument poses is narrower and harder: does the structure survive the wait? A perpetual has no maturity to force an answer. It has only a coupon, a promise, and a market that has begun to price both.
In an algorithmic age where machines generate infinite claims about value, the only scarce asset left is verifiable human intent โ and a $100 anchor that does not hold is a machine talking to itself. Trust no one, verify the solitude. Speed kills. Precision saves.