Nine hundred twenty-one thousand five hundred eleven.
That's how many SATA shares Strive printed between Aug. 28 and Sept. 4. The company's variable-rate perpetual preferred went from 9,073,914 shares to 9,995,425 — inside seven calendar days, through an at-the-market program, with no single headline event to justify it.
Multiply the increase by the $13 annual rate charged on SATA's $100 stated amount. You get $11,979,643 of new annualized dividend obligation. Call it $12 million, created in a week.
Over the same window, cash and cash equivalents rose $19.1 million, from $183.5 million to $202.6 million. Run the coverage math — cash divided by the current-rate annualized dividend — and the number goes from 18.67 months to 18.71 months. Flat. Marginally better, if you squint.
That flat ratio is the entire bull case in one line. Cash grew faster than the dividend bill. Coverage held.
It's also close to meaningless.
I've been rebuilding these treasury-company flow tables in Python since Strategy first started layering preferreds under the common, and the pattern is always the same: the coverage ratio measures the wrong thing in the right units. It counts dollars against dollars. It ignores what those dollars are backed by, who sits on the other side of the trade, and whether the funding channel that created the dividend stays open long enough to matter.
Speed beats analysis when the graph is vertical. Strive's graph isn't vertical. It's a slow, quiet ramp, one ATM print at a time — and slow ramps are what break preferred holders. Not crashes.
The Compressed Version of the Playbook
If you live in spot ETFs and haven't touched the treasury-company complex, here's the two-paragraph version.
A public company sells equity, converts the proceeds into bitcoin, and sells shareholders a story about monetary debasement. Strategy built the template. Then it built a second layer — a preferred stack in the STRF, STRK and STRC family — that lets the company raise dollar capital without touching common equity, at a fixed or variable coupon, with no maturity date attached.
Perpetual preferred is legally equity. That's the whole appeal. It doesn't sit in the debt column. It doesn't trip covenants. It has no maturity wall, so there's no refinancing cliff to roll. It just has to be paid — in cash, on schedule, indefinitely — or the board suspends the dividend, which craters the share price and shuts the funding channel that keeps the structure alive.
Strive copied the template. SATA is its instrument: variable-rate perpetual preferred, $100 stated amount, board-set rate, currently 13% annualized. The board affirmed 13% on Aug. 13, effective for periods beginning Sept. 1.
Now the accrual mechanics, because they matter far more than the headline rate.
SATA doesn't pay a quarterly coupon on the last day of the quarter. It accrues per business day. For September, the board declared $0.0516 per share on each of 21 business-day payment dates, payable to holders of record at the preceding business day's close. Do the arithmetic: $0.0516 × 21 = $1.0836 per month. Annualize it: $13.00. Exactly 13% of the $100 stated amount, spread across 252 business days.
Two consequences follow, and both get lost in the coverage debate.
First, the dividend liability grows with the calendar, not with the quarter. Every day a SATA share is outstanding, the obligation exists. Every new share adds $13 of annual cash claim at the current rate. There is no quarter-end cutoff where the company gets to breathe.
Second, the $129.9 million figure is a run-rate projection, not a committed liability. Payments depend on eligible shares on each record date. Redeem shares, and the number falls. Issue shares, and it rises. The table is a snapshot of a machine in motion, and the machine's speed is set by the ATM window.
Now the week in question. Strive reported buying 1,375 BTC between Aug. 31 and Sept. 4 at an average price of roughly $79,281 per coin, fees and expenses included. Holdings stood at 24,531 BTC as of Sept. 4.
Multiply it out: 1,375 × $79,281 = $109.0 million of bitcoin purchased inside that five-business-day window.
Hold that number. It's about to matter.
The Reconciliation the Filing Doesn't Do
921,511 new SATA shares. $100 stated amount. That's $92.2 million of gross preferred issuance, assuming the shares printed at par.
Bitcoin purchased: $109.0 million.
Gap: $16.9 million.
And cash went up, not down — $19.1 million higher over the same period.
I ran this reconciliation on a Tuesday night, in Python, because Strive's disclosure doesn't do it for you. The company explicitly states that the filing does not allocate the BTC purchases between specific financing sources. That's standard for these tables. Standard, though, is not the same as neutral, and the arithmetic leaves three possibilities the filing never resolves.
SATA shares may have printed above par. Preferred can trade above its $100 stated amount, and if the ATM sold into strength, gross proceeds exceed $92.2 million. But even at $102 you get $94.0 million. Still short. Some BTC may have been bought with existing cash, with the cash increase arriving from elsewhere — common equity ATM sales, investment income, distributions off the STRC position. Or it's timing: Strive's table includes shares sold by its stated 4 p.m. cutoff that would be issued on the following business day. The Sept. 4 count includes shares that hadn't settled. The cash follows later.
I'm not alleging a funding hole. I'm pointing at something structural: a treasury company sitting on $202.6 million of cash and a $130 million annual dividend run-rate can absorb a $16.9 million reconciliation gap without a single analyst asking a question. That's the blind spot. Not fraud — opacity that gets a free pass because the ratio is flat and the story is good.
Now the part of the balance sheet the coverage ratio never touches.
Strive holds 505,000 shares of Strategy's STRC preferred, valued at $49.364 million on Sept. 4. That's roughly $97.75 per share against a $100 stated amount.
Read that structurally, not as a line item. A bitcoin treasury company is holding another bitcoin treasury company's variable-rate preferred stock, classifying it as a liquid asset, and marking it slightly below par.
STRC is Strategy's variable-rate preferred. So the asset side of Strive's "cash and cash equivalents" story includes a claim on Strategy's own dividend obligation. If Strategy's preferred complex ever wobbles, the numerator of the ratio everyone quotes wobbles with it.
Back in November 2022 I spent two weeks building a real-time solvency list of venture firms while FTX was unwinding. The lesson from that stretch wasn't about any single firm. It was that in a crisis, the assets you believed were diversified are almost always the same trade wearing different tickers. The treasury complex is now financing itself with each other's preferred paper. That's not diversification. It's a mirror — and mirrors don't provide liquidity when the lights go out.
The Coverage Ratio Is a Runway in Disguise
18.71 months. Here's what's inside it and what's outside it.
Inside: cash and cash equivalents of $202.6 million. Run-rate dividends at the current 13% rate, $129.9 million annualized.
Outside: operating expenses, future financing costs, investment income, other liquid assets, and — critically — the fact that the numerator is a nominal dollar balance while the denominator is a board-set rate that can be reset at any meeting.
The exclusions cut both ways, and I'll be fair about that. Leaving out other liquid assets makes the ratio conservative. Leaving out operating expenses makes it generous. Net direction is genuinely unclear. What isn't unclear is that the ratio treats a variable-rate, board-set, perpetually resettable dividend as if it were a fixed bond coupon. It isn't.
Model the cost of capital directly. Every $100 of SATA raised at par costs $13 a year, in cash, with no maturity and no principal repayment. To service that $13 out of bitcoin, Strive needs BTC to appreciate, or it needs to issue more preferred. Issue more preferred, and you add more $13 claims.
That's the reflexivity. The dividend is funded by the same instrument that creates the dividend. It works precisely as long as the equity narrative holds and the ATM window stays open. It works the way a chain letter works — not because anyone designed a fraud, but because it depends on a continuous inflow nobody controls.
Which brings us to the rate — because the rate is the price of the marginal dollar, and the marginal dollar is the marginal buyer.
SATA pays 13% today. Why? Because 13% is what it takes to clear a perpetual preferred issued by a company whose assets are bitcoin, priced for duration risk, credit risk, and realized volatility. If the marginal buyer starts demanding 15%, the board faces a fork: reset the rate upward and accept a bigger run-rate dividend, or let the shares trade below par and accept a dilutive — eventually closed — ATM.
Look at Strategy's STRC print again. $97.75 against $100.
The largest, most liquid name in the complex, the one with the deepest buyer base and the longest track record, has a variable-rate preferred trading under stated amount. That's the tell. The preferred market stopped paying up for par months ago; the issuance tables simply haven't caught up to it yet. I don't read whitepapers; I read order books. And the order book on perpetual preferred is telling you the marginal bid sits below the strike.
So stress it. Three scenarios, all plausible, none priced.
Scenario one: a rate reset. The board lifts SATA to 15% to keep it trading near par. 9,995,425 shares × $15 = $149.9 million of run-rate dividends. Nothing else changes — no new shares, no new bitcoin. Coverage against $202.6 million drops from 18.71 months to 16.2 months. A board vote compresses the so-called stability metric by 2.5 months, and the market won't see it until the next table drops.
Scenario two: the window closes. Bitcoin draws down, the ATM meets resistance, and Strive stops issuing. The share count freezes at 9,995,425. The $129.9 million run-rate becomes a hard cash draw against a $202.6 million pile with no refill. 18.71 months of coverage becomes 18.71 months of runway.
That's the vocabulary problem the market keeps tripping over. Coverage implies the ability to cover indefinitely. Runway implies a countdown. At a $130 million annual draw with no operating cash flow, Strive has runway — it's just labeled coverage.
Scenario three is the one currently in motion: everything keeps working. Bitcoin appreciates, the equity story holds, the ATM stays open, and the share count climbs. At 921,511 shares a week, the SATA base crosses 12 million shares inside a quarter — a $156 million annual run-rate. Then 14 million, and $182 million. Every step looks accretive in the moment, because the bitcoin bought per share looks good. Every step is also a new permanent dollar claim against a volatile asset.
When I published my slippage work on swap routing in 2020, the whole point was that slippage isn't a footnote — it's a cost that compounds invisibly until it dominates the trade. Same structure here, different asset. The dividend creep is slippage. It doesn't show up in any single print. It shows up in the aggregate, eventually, all at once.
The Angle Nobody Is Trading
The consensus read on this filing is: coverage maintained. Cash up $19.1 million, dividends up $12 million, ratio held at 18.7 months. Bullish. Accumulate.
The contrarian read is that the ratio is a vanity metric and the number that actually matters never appears in the table.
That number is SATA's trading price relative to its $100 stated amount.
Nobody in this complex funds themselves forever by issuing preferred at par. They fund themselves while the shares trade at or above par. The moment SATA prints $97, then $95, then $93, the ATM math inverts: you're selling $93 of capital to service $13 of annual dividend, and your real cost of capital climbs even though the stated rate never moves.
Strategy's STRC at $97.75 is the canary. Strive's 505,000 STRC shares are the exposure. The $16.9 million gap between preferred issued and bitcoin bought is the question I'd want answered before calling a $130 million run-rate "covered."
The best news is the news that moves the price. This filing moved nothing. That's exactly why it deserves a second read.
What to Watch Next
Three things, over the next two disclosure windows.
The 4 p.m. cutoff table — specifically whether the SATA share count keeps climbing at roughly 900,000 a week or decelerates. Deceleration means the ATM is meeting resistance and the reflexivity is bending.
The closing print on SATA, and on Strategy's STRC. If SATA trades at a discount to stated amount, the funding engine is sputtering, and the run-rate dividend flips from a growth option into a fixed cost.
And the reconciliation. If bitcoin purchases keep exceeding preferred issuance while cash keeps rising, either the shares are clearing above par — which is fine — or something else is funding the gap. Both answers are tradeable. Only one is comfortable.
The treasury-company complex spent two years proving it could raise capital. The next two quarters will prove whether it can service it. The question isn't whether Strive can cover $130 million a year. It's who holds the other side of that trade when the coverage ratio stops being flat — and whether they know they're holding a mirror.