Have you ever watched a company that built its public identity around Bitcoin quietly send more cash to a preferred-stock ticker than to the coin itself? That is the odd feeling hanging over Strategy’s latest weekly snapshot. Between September 14 and September 20, the firm spent about $174 million buying back STRC and only $75.7 million adding 950 Bitcoin. The ratio is roughly 2.3 to 1. It is not a collapse of the treasury thesis. It is a reminder that the machine now has more than one job.
What The Latest Cash Split Actually Shows
The week was not empty. Management deployed $307.1 million across three buckets: STRC repurchases, Bitcoin, and the standing cost of preferred dividends plus debt interest. Nearly 57 percent of that disclosed outflow went to STRC. Bitcoin took about 25 percent. Dividends and interest took the rest, roughly 19 percent. Those percentages matter more than the headline, because they show how the firm is using cash it already had rather than fresh share issuance.
No sales went through the at-the-market programs for common stock or the various preferred lines during the week. In plain language, the company did not print new paper to fund the activity. It spent from the till. That is a different posture from the years when almost every Bitcoin lot seemed to arrive with a matching equity raise. I’ve found that investors often miss this shift. They still talk as if every coin is financed by dilution. This week, that story does not hold.
The STRC Ticket In Hard Numbers
Strategy bought 1,751,480 STRC shares at an average of about $99.34. The security is variable-rate Series A perpetual Stretch preferred stock with a $100 stated amount. It trades on the same major Nasdaq venue as the common shares. After the latest batch, roughly $876 million of room remained under the digital credit securities repurchase authorization. A separate $1 billion authorization for common stock buybacks is still sitting there, unused in this particular week.
This was not a one-off. In the prior window, from September 8 to September 13, the company had already spent $139.3 million on 1,420,467 STRC shares at an average near $98.06. Stack the two weeks and you get $313.3 million aimed at STRC. Bitcoin purchases across those same two windows total only $75.7 million, because the first of those weeks included no coin buys at all.
Buying back a perpetual preferred is not the same as buying the asset that made the company famous. It is a bet that shrinking the dividend base is worth more, right now, than adding another sliver of coins.
That last point is the one I keep circling. STRC has no maturity date. As long as the shares stay out there, the dividend conversation never fully ends. Each share retired lowers the future claim on cash. That is boring corporate finance. It is also why the repurchase can look more urgent than another 950 coins when the preferred is trading a hair under par.
Bitcoin Is Back, Just Not In First Place
The company did resume buying. It paid $75.7 million, fees included, for 950 BTC at an average of $79,670. Holdings moved from 845,050 to 846,000 coins. The disclosed aggregate cost sits at $63.80 billion, or about $75,416 per Bitcoin across the whole stack. The newest lot landed above that lifetime average and below the spot prints that followed when Bitcoin traded above $85,000 after bouncing from September lows near $75,000.
At $85,000, the 846,000 coins would be worth about $71.91 billion on paper. That is mark-to-market, not cash in the bank. The number moves every hour. Still, it is the figure people use when they argue the treasury is “in the money” versus cost. Fair enough. Just do not confuse paper value with spendable liquidity. Strategy has already shown, earlier in 2026, that it can sell coins when the capital plan calls for it.
Before this 950-coin add, the last purchase sat on August 31: 4,603 BTC for about $370 million at $80,318. That ended a pause of roughly ten weeks. The firm has also sold. During the week ending August 3 it sold 1,638 BTC for $104.73 million while raising $290.6 million through common-stock sales and buying back $81.2 million of STRC. Holdings briefly sat at 842,138 coins after that sale. Purchases since then rebuilt the stack to the new 846,000 total.
Why Preferred Stock Became A Two-Way Street
Earlier in the year, preferred issuance was framed as a way to raise money for more Bitcoin while giving income-focused buyers a dividend product. Common equity and convertible debt were no longer the only taps. That was the pitch. Recent filings show cash walking the other way. The company is now retiring STRC instead of leaning on it as a permanent funding hose.
There is a logic to the reversal. A perpetual preferred is a long claim. Variable rates can reset. The stated amount sits near $100, and the market has been willing to trade close to that level. Buying below or around par can be a clean way to shrink a liability-like stream without waiting for a call that may never feel urgent. In my experience, markets reward that kind of housekeeping only after the fact. During the week it just looks like “they spent more on STRC than on Bitcoin,” which is true and incomplete at the same time.
- STRC is perpetual, so the dividend clock does not expire on its own.
- Repurchases reduce the share count that can keep claiming cash.
- The buybacks used existing cash, not new ATM proceeds.
- Bitcoin buying resumed, but it was the smaller check.
- Common-stock buyback authority remains unused in this window.
Perhaps the most interesting aspect is how cleanly the firm separated roles for cash. USD Cash is the flexible pool. It can buy coins, refill the reserve, repurchase securities, or cover ordinary corporate needs. The USD Reserve is narrower. It exists to support preferred dividends and interest. Mixing those two stories is how people get lost in the footnotes.
Cash Balances After The Week’s Checks Cleared
USD Cash dropped from $1.30 billion on September 13 to $1.05 billion on September 20. The decline of about $250 million lines up with the $174 million STRC buyback plus the $75.7 million Bitcoin lot. The USD Reserve slipped from $5.10 billion to $5.04 billion after $57.4 million went out for preferred dividends and debt interest. Nobody should be shocked. That is what a reserve is for.
The remaining cash is still large by most corporate standards. It is not infinite. If STRC buybacks keep running at a similar clip and Bitcoin purchases return to the old rhythm at the same time, the flexible cash line will keep leaking. Management can always reopen the ATM window. It chose not to this week. That choice is the signal. When they start selling paper again, the story changes.
| Use of cash | Amount | Share of $307.1M |
| STRC repurchases | $174 million | About 57% |
| 950 Bitcoin | $75.7 million | About 25% |
| Dividends and interest | $57.4 million | About 19% |
Look at that table twice. The Bitcoin line is real. It is just not the dominant line. If you only read social posts, you would think the company did nothing but stack coins. The filing is colder. It shows a treasury that is now managing a capital structure, not only a coin count.
How The Different Tickers Sit In The Stack
For anyone holding the securities, the weekly disclosure is a map of who owns what claim. Common shareholders own the residual equity. Preferred holders own dividend terms, payment priority inside the structure, and a different risk profile. The filing lists the common shares plus STRC, STRF, STRK, and STRD as registered classes on the same Nasdaq Global Select Market. Same exchange. Different contracts with the company.
That matters when people treat every ticker as a leveraged Bitcoin note. They are not identical. A variable-rate perpetual preferred does not move one-for-one with the coin, and it should not. Its job is income and capital structure, not maximum torque to spot. When the company buys it back, common holders can cheer the smaller preferred load. Preferred holders who wanted to keep the paper may feel less thrilled. Both reactions can be rational.
I keep coming back to a simple question. If the preferred was issued to fund more coins, why retire it while the coin bid is only modest? One answer is price. Near $99 against a $100 stated amount, the repurchase is tidy. Another answer is signaling. After two quiet weeks with no Bitcoin and no share sales, restarting the coin bid while still favoring STRC says the firm can walk and chew gum. It also says the gum, this week, was the preferred.
The Pause, The Restart, And The Sale In Between
Context helps. Strategy had gone two weeks without buying or selling Bitcoin and without issuing shares, using cash instead to cut STRC. Then the latest window restarted coin purchases. The STRC bid never really paused. That is why the two-week combined figure looks so lopsided. Bitcoin had a zero in week one and a $75.7 million print in week two. STRC had large prints in both.
The August 3 week is the other bookend. Coins were sold. Common stock was sold. STRC was still being retired. That combination is the grown-up version of the treasury model. Use the asset when you need liquidity. Use equity when the window is open. Use preferred buybacks when the paper is cheap enough relative to the claim it represents. None of that fits on a bumper sticker. It does fit a 8-K.
Does a 950-coin add after a ten-week drought change the long-run identity of the firm? Not really. The stack is still enormous. The cost basis is still the number people argue about on every dip. What changed is the mix of weekly priorities. Identity is the pile of coins. Operations are the checks written each week. Those two can drift apart for a while without the thesis breaking.
What “2.3 Times More” Does And Does Not Mean
Ratios travel well on timelines. They also flatten the story. Spending 2.3 times more on STRC than on Bitcoin this week does not mean the company abandoned the asset. It means the preferred market offered a use of cash that management ranked higher for seven days. Next week the ranking can flip. That is allowed. Companies are allowed to have more than one good idea in a month.
Still, I would not shrug. If the STRC program keeps absorbing the bulk of discretionary cash while Bitcoin buys stay small, the market will start asking whether the treasury is in accumulation mode or liability-management mode. Those modes can coexist. They do not feel the same to holders who bought the common shares as a Bitcoin proxy. Expectations are sticky. Filings are not.
- Read the cash split before you read the coin count.
- Check whether ATM programs were used. This week they were not.
- Separate USD Cash from the USD Reserve. They are not twins.
- Watch the leftover repurchase capacity on STRC and on common shares.
- Compare the latest Bitcoin lot to both cost basis and later spot prices.
That checklist is unglamorous. It is also how you avoid getting bounced around by a single ratio. The 2.3 figure is useful. It is not a philosophy.
Cost Basis, Spot Price, And The Temptation To Scorekeep
The lifetime average near $75,416 is now a totem. Every purchase above it invites a groan. Every purchase below it invites a cheer. The latest $79,670 lot sits in the groan camp versus the average and in the cheer camp versus later prints above $85,000. Scorekeeping like that is human. It is also a bit theatrical. A 950-coin ticket on an 846,000-coin pile barely moves the blended cost. The psychology moves more than the math.
Mark-to-market value around $71.91 billion at $85,000 is a headline number. It does not pay the preferred dividend. It does not retire STRC. It does not refill the reserve. Only cash and, when chosen, coin sales do those jobs. The August sale already proved the company will use Bitcoin as a funding tool when the plan says so. Anyone treating the stack as permanently locked should revisit that week.
A treasury can be both a long Bitcoin position and a working balance sheet. Pretending those roles never collide is how narratives get sloppy.
I’ve watched this movie in other industries. A firm becomes famous for one asset. Then the capital structure grows around that fame. Suddenly the weekly news is about preferreds, reserves, and interest. Fans call it mission drift. Accountants call it Tuesday. The truth sits in the middle. The mission is intact if the stack keeps growing over quarters. The mission is noisier if the weekly cash goes to cleaning up paper first.
Dividends, Interest, And The Quiet Third Line
The $57.4 million that left the reserve does not trend. It should not. It is the cost of the structure the company chose. Preferred dividends and debt interest are the price of having raised money without, in those moments, selling more common stock. Ignore that line and the 57/25 split looks like the whole story. Include it and you see a firm that already committed cash to stay current on its own inventions.
That is why STRC buybacks have a double effect. They spend cash today. They can reduce the cash that must leave the reserve tomorrow. Whether that trade is brilliant depends on the rate, the remaining life of the obligation, and what else the cash could have bought. Bitcoin at $79,670 was the alternative this week. Management picked more STRC anyway. You can disagree with the ranking. You cannot say the ranking was hidden.
What Common Holders Should Watch Next
Three switches will tell you more than any single week. First, does the ATM stay closed while buybacks continue? That is a cash-first stance. Second, does Bitcoin buying scale back toward the old multi-thousand-coin clips, or stay in the hundreds? Size is the tell. Third, does the STRC authorization keep getting used near par, or does the bid fade if the preferred cheapens or richens?
There is also the unused common-stock repurchase authorization. A billion dollars of room is not a promise. It is an option. If common shares ever look, in management’s eyes, like a better retirement target than STRC, the mix will shift again. Do not assume the current ranking is carved in stone. This company has already rewritten the weekly script more than once in 2026.
Liquidity in USD Cash is the constraint that makes all of this real. $1.05 billion after the week is comfortable, not endless. A few more $174 million preferred lots plus a return to $300 million-plus Bitcoin weeks would chew through that line fast. Then the ATM conversation returns. Dilution versus buybacks versus coin adds is a triangle, not a straight line.
A Fair Reading For People Who Only Want The Coins
If you hold the common shares because you want leveraged exposure to Bitcoin, this week was mixed. The pile grew. Good. The growth was small. Less good if you were hoping for another 4,000-coin splash. Cash went to preferreds first. That can still help you if it lowers a long-dated claim. It does not give you the dopamine of a giant buy. Both things can be true in the same paragraph.
If you hold STRC for the dividend design, the repurchase is a market event, not a love letter. Some holders are happy to sell near $99. Some wanted to keep collecting. The company is allowed to shrink the class. That was always in the toolkit once the repurchase program existed. $876 million of remaining capacity means this chapter is probably not finished.
Weekly posture in one glance: Flexible cash down to $1.05B Reserve down to $5.04B Coins up to 846,000 STRC still the larger check ATM window closed for the week
That block is the whole mood. Accumulation did not die. Housekeeping got a bigger budget. Markets hate split narratives. Balance sheets produce them anyway.
The Human Side Of A Very Corporate Week
People get attached to simple stories. This firm became a cultural object because it kept buying Bitcoin when that looked reckless to traditional desks. Simple stories age. Once you add perpetual preferreds, a dedicated reserve, multiple Nasdaq tickers, and a formal repurchase program, you are no longer just a buyer of coins. You are a financial company with a famous inventory.
Is that a betrayal of the original plot? Only if you thought the plot was “never spend a dollar on anything else.” That was never a real corporate plan. It was a slogan. Slogans do not pay variable dividends. I say that as someone who still thinks the size of the stack is the main reason the equity trades the way it does. You can believe that and still admit that $174 million toward STRC is the news this week, not the 950 coins.
Will the next filing flip the ratio? Maybe. Bitcoin above $85,000 after a $79,670 purchase gives the team a decent print to point at. STRC near $99 gives them a decent print to point at too. When both trades look “fine,” the allocation becomes a statement of priority rather than a distress signal. That is the reading I lean toward. Not panic. Not victory lap. Priority.
So where does that leave a reader who has to decide what to watch on the next 8-K? Watch the ratio again, sure. Then watch the funding source. Cash-financed buys and buybacks tell a different story than ATM-financed ones. Watch the reserve, because a shrinking reserve with a growing preferred load would be a worse picture than the one we got. Watch whether 846,000 becomes a round number they defend with larger lots, or just another waypoint.
The company spent 2.3 times more on STRC than on Bitcoin in a single disclosed week. It also added coins after a pause, paid its structure, and left hundreds of millions of repurchase capacity on the table. That is not a mystery novel. It is a capital-allocation memo with a famous asset in the background. The asset still dominates the identity. This week, it did not dominate the checkbook. If that tension bothers you, you are paying attention. If it does not bother you at all, you may be ignoring the part of the story that now moves the cash.
