Strategy STRC Buyback Nears $100 After $951M Purchases

Almost a billion dollars later, the preferred that once slumped toward seventy dollars is knocking on the door of par again. That is not a small market hiccup. It is a company using cash, common-stock proceeds, and even a slice of Bitcoin sales to pull its own Stretch preferred back toward a tight $99-to-$100 band. I have watched a lot of treasury stories get dressed up as strategy. This one is unusually blunt about the goal.

Why The Stretch Preferred Suddenly Matters Again

Strategy has spent about $950.8 million buying back nearly 9.96 million shares of its Variable Rate Series A Perpetual Stretch Preferred Stock since July 20. The latest weekly filing covering September 8 through September 13 added another 1,420,467 shares for $139.3 million, paid from the USD cash balance. No new Stretch was issued through at-the-market programs in that window. No Bitcoin changed hands either.

Add the eight weekly disclosures from late July through mid-September and the picture is consistent. Management wants Stretch near its $100 stated amount. The board even lifted the preferred-securities repurchase authorization from $1 billion to $2 billion on September 8. After the most recent tranche, roughly $1.05 billion of that capacity was still unused.

STRC closed the latest completed U.S. session on September 16 at $97.07 after an earlier intraday print above $99. The bounce from nearly $70 looks dramatic on a chart. It also raises the question every preferred holder eventually asks: is the price now a product of real demand, or of a buyer who can still write very large checks?

The Weekly Buyback Trail In Plain Numbers

The campaign did not start loud. It started with a modest $25 million clip. Then the tickets got bigger, almost week by week, as the discount stayed wide enough to look accretive.

That last line brings the running total to 9,961,554 shares and about $950.8 million. Independent market tallies put company activity near 18 percent of total Stretch trading volume over the same stretch. Eighteen percent is not quiet support. It is a heavy footprint.

I find the pacing more interesting than the headline sum. Management has said it likes to buy more aggressively when the discount is deep and to taper as the security approaches $100. The weekly prints mostly follow that script. Early tickets were smaller. August and early September were the heavy weeks. The most recent check was still large, but not the peak.

Par Is Not A Promise, Even If It Looks Like One

August investor materials spell out the toolkit. Dividend-rate decisions. Repurchases. Liquidity management. Issuance policy. All of it is meant to keep Stretch near the $100 stated amount. The same materials warn that the target is not a price guarantee and that the policy can change.

When Stretch gets back to par, we’ll issue more.

– Company leadership, discussing capital plans

That sentence is the whole business case in one breath. Stretch is not just a listed preferred. It is a funding pipe. If the shares trade near par, new issuance can raise cash without looking like a distressed sale. If they languish at a discount, that pipe clogs. New capital for Bitcoin purchases becomes harder to tap through this particular security.

The current coupon is a 12 percent annualized rate, or $0.50 for each full semi-monthly period on $100 of stated value when the dividend is declared. The board has already declared $0.50 payments for the periods ending September 30 and October 15. Management has also said it does not currently intend to recommend changing that 12 percent rate until Stretch shows sustained trading near $100. Existing policy is similarly not to sell new Stretch below the $100 stated amount. Both are statements of intent, not contracts.

Perhaps the most interesting aspect is how tightly issuance and buybacks now sit next to each other. Buy below par to retire $100 of stated value for less than $100 and shrink future dividend load. Issue near par later to refill the war chest. It is neat on a whiteboard. Markets are rarely that tidy once the company steps back from the tape.

Where The Cash Actually Came From

Buybacks need a source. This program had three. Common-stock sales. Bitcoin sales. Then the cash pile itself.

Market reconstruction of the July 20 to mid-September window attributes roughly $765 million, or about 80 percent of repurchase spending, to common-stock issuance, including money that first sat in the cash pool before being redirected into Stretch. Another $161 million, give or take, came from Bitcoin sales. The rest was already sitting in dollars.

The Bitcoin-funded slice can be tied to specific weekly reports. During July 27 through August 2, the company sold 1,638 BTC for $104.73 million and sent $52.3 million of those proceeds into Stretch repurchases. The other $52.4 million covered preferred dividends. A week later it sold 1,690 BTC for $108.6 million at an average net price of $64,262 and said all of that sale funded Stretch buybacks. Those two allocations total $160.9 million, which matches the broader $161 million estimate.

Later weeks leaned on common-stock proceeds more directly. One week sent $132.2 million of those proceeds into Stretch. The next sent $136.4 million. The week after that sent $151.8 million. For the first two September reporting periods, $315.6 million of Stretch purchases came from the USD cash account. That balance slipped from $1.61 billion on August 30 to $1.44 billion on September 7 and $1.30 billion by September 13. A separate USD reserve still stood at $5.10 billion.

• Common-stock issuance funded the bulk of the campaign.
• Two disclosed Bitcoin sales supplied about $161 million.
• September tickets pulled more heavily from existing dollar cash.
• The unused common-stock repurchase authorization stayed untouched.

There is a separate $1 billion authorization for buying back the common stock. It remained unused through the latest filing. That contrast is worth a pause. Preferred paper got the checks. Common equity was the funding source, not the target.

Bitcoin Stayed Parked At 845,050 Coins

For the two consecutive periods ending September 7 and September 13, Strategy reported no Bitcoin purchases and no Bitcoin sales. Holdings therefore stayed at 845,050 BTC, acquired for an aggregate $63.73 billion at an average cost near $75,412 per coin.

The last reported purchase in this window came during August 24–30, when the company added 4,603 BTC for $369.7 million at an average $80,318. That same week it also allocated $151.8 million from common-stock issuance to Stretch. In other words, it was buying coins and buying its own preferred in the same reporting period.

Across July 20 through September 13, Stretch repurchases consumed about $950.8 million. New Bitcoin purchases in that same span were $369.7 million. Preferred support took more than twice the dollars directed at fresh coins. That ratio will bother some long-term Bitcoin holders. It will look perfectly rational to anyone who treats Stretch as the next funding valve.

I’ve found that treasury debates often collapse into a false choice: stack coins or tidy the capital stack. This episode is both. The coin stack is huge and, for now, frozen. The preferred stack is being cleaned up so it can be used again.

Why Buying Below One Hundred Can Still Be Accretive

The economic argument is simple enough to put on a napkin. Stretch carries a $100 stated amount. Buy it at $90 and you retire a hundred dollars of preferred claim for ninety. Future dividend obligations shrink. If the company later issues new Stretch near par, it has effectively recycled cheap paper into a cleaner funding tool.

That math only works if two things hold. First, the buybacks cannot become the only bid. Second, investors have to believe the 12 percent rate and the near-par issuance rule will last long enough to matter. Take either belief away and the security starts trading like a high-yield instrument with policy risk attached.

In my experience, preferred markets punish mixed signals faster than common equity does. Income buyers want a clean story: coupon, call features, issuer behavior. They do not want a weekly referendum on whether management still likes the product. That is why the “not a guarantee” language in filings deserves more attention than most people give it.

The ETF Footprint And The Liquidity Question

Earlier in the summer, Stretch had already become a large holding inside three major U.S. preferred-stock funds while still trading below par. Combined exposure across those products was cited around $756 million at the time. That matters because fund flows can amplify a rebound or a fade. If the security holds near par, more passive demand can follow. If it slips again after company buying cools, those same holders may not want a growing discount in a product they treat as income ballast.

Liquidity is the unglamorous part. A company that represented close to a fifth of trading volume cannot step away without leaving a hole. Analysts have already asked the obvious question: how does Stretch trade when official purchases decline? One portfolio manager put it bluntly enough that it stuck with me. The price is the product. If Stretch cannot live near par without constant official support, it stops being an efficient funding instrument.

Company filings warn that repurchases and dividend policies may fail to keep the security near $100.

That warning is not decoration. It is the legal version of “we will try.” Investors who treat the $99-to-$100 band as a floor should read that sentence twice.

What The Bigger Authorization Really Changes

The original June authorization capped preferred-security purchases at $1 billion and named Stretch as first priority when management judged buybacks accretive. On September 8 the board doubled the envelope to $2 billion, including amounts already spent. After the September 8–13 tickets, $1.05 billion remained.

Capacity is not a spending plan. Policy still lets the company increase, reduce, suspend, or stop purchases depending on price, liquidity, available capital, and other market conditions. The next weekly decision is therefore open on purpose. If Stretch holds in the high nineties, tickets may shrink. If it slips, the remaining billion-plus can still show up on the tape.

There is also a sequencing issue. Leadership has linked new Stretch issuance to a return toward par, and linked that issuance to the ability to raise funds that could go into Bitcoin. So the buyback is not an isolated capital-allocation hobby. It is maintenance on the machine that is supposed to buy more coins later.

1. Retire discounted Stretch and cut future preferred dividends.
2. Let the market test whether the security can sit near $100.
3. Resume issuance near par if the test holds.
4. Point fresh preferred proceeds back toward the Bitcoin treasury if that remains the plan.

Step two is the fragile one. Steps three and four do not happen if step two fails.

How Income Buyers Should Read The Coupon Story

Twelve percent on a perpetual preferred is not a sleepy bank coupon. It is a high advertised yield attached to an issuer whose common equity and Bitcoin reserve can swing hard. That combination attracts two very different crowds. One crowd wants the income and treats the Bitcoin treasury as extra collateral in spirit, even if the preferred claim is structurally different. The other crowd treats Stretch as a trading instrument whose fair value is a function of Bitcoin beta, issuance risk, and buyback intensity.

Those two groups do not need the same price. Income buyers can live with $97 if they trust the dividend calendar. Traders may need $99 to $100 because that is where the official narrative lives. When those groups disagree, spreads get noisy. We already saw that in the slide toward $70 and in the sprint back toward $99.

The declared $0.50 payments for the periods ending September 30 and October 15 keep the near-term cash story simple. The harder story is the rate review. Management does not currently intend to recommend a change until Stretch shows sustained trading near $100. “Sustained” is doing a lot of work in that sentence. A single tick above $99 is not a regime.

Common Stock As The Hidden Engine

It is easy to talk about Stretch in isolation. That would be sloppy. A large share of the repurchase cash started as common-stock issuance. The company sold equity, parked dollars, then used a portion of those dollars to buy preferred. In some weeks the allocation was explicit in the same reporting period.

That loop has fans and critics. Fans say it is just balance-sheet plumbing: issue the more flexible security, retire the one that is offside, keep the Bitcoin reserve intact. Critics say it is circular. Dilute the common to defend a preferred that exists so the company can later raise more money, possibly to buy more Bitcoin, which then supports the common. You can believe either story. You cannot pretend the loop is not there.

The unused $1 billion common-stock repurchase authorization is the quiet tell. If management wanted a simple “return cash to equity holders” headline, that facility was sitting there. It chose the preferred instead. That choice reveals the current priority. Clean up Stretch. Keep the funding option alive.

What Happens If Official Buying Fades

Assume Stretch holds $97 to $99 and company tickets shrink. Two outcomes are plausible. Genuine income demand fills the gap and the security oscillates in a tight band. Or the bid thins, the discount reopens, and the funding pipe looks expensive again.

A reopened discount would not be a moral failure. It would be a market. Perpetual preferreds reprice when rates move, when issuer policy looks less certain, or when the bid that was 18 percent of volume steps aside. The company can always return with the remaining authorization. It can also stand down. Filings leave both doors open.

I’d argue the next six reporting weeks matter more than the last eight. The last eight proved management would spend. The next six will show whether anyone else wants the paper at these levels without that much official company flow.

A Practical Checklist For Readers Tracking The Story

You do not need a spreadsheet full of macros. You need a short list you actually check.

• Weekly Stretch repurchase size and average price versus $100.
• Whether new Stretch is issued, and at what price relative to stated amount.
• USD cash and reserve balances after each reporting window.
• Any Bitcoin purchases or sales, even small ones.
• The stated dividend rate and whether language around “sustained” trading near par changes.
• Remaining preferred repurchase capacity against the $2 billion cap.

If those six items stay boring, the official plan is working. If two or three of them break at once, the narrative will move faster than the press releases.

The Broader Treasury Lesson Hiding In One Security

Public companies that hold large Bitcoin reserves eventually bump into the same design problem. Coins do not pay a coupon. Operating cash may be thin. Equity issuance can look painful. So they invent hybrid paper that promises income to one set of investors while preserving upside optionality for another. Stretch is one version of that invention.

The invention only works if the hybrid paper has a home in the market. A perpetual preferred that cannot live near its stated amount becomes a tax on future issuance. A preferred that can live there becomes a tap. That is why almost a billion dollars went into a security most casual crypto readers had never heard of six months earlier.

Is that elegant? Sometimes. Is it fragile? Also sometimes. Capital structures that depend on a tight trading band need either deep natural demand or a sponsor willing to keep buying. Right now Stretch has had the sponsor. The test of natural demand is still incoming.

Reading The Latest Close Without Overfitting One Session

A $97.07 close after a $99.03 intraday high is not a failure. It is a security that tagged the official zone and then drifted. Intraday highs get screenshots. Settles get remembered. The $97.01 to $97.95 range on September 16 looks like digestion, not collapse.

Still, digestion after heavy official buying is not the same as independent price discovery. Give it time. If Stretch can spend several weeks in the high nineties with shrinking company volume, the $100 project looks healthier. If it needs another $139 million week to stay there, the project is still on company life support.

That distinction is the whole article, if I am honest. The dollars are public. The weekly share counts are public. The remaining authorization is public. What is not public yet is whether Stretch can stand up without the buyer that just spent $950.8 million trying to teach the market where par is supposed to be.

Final Thoughts Before The Next Filing

Strategy has made its preference clear. Stretch should trade between $99 and $100. The board doubled the repurchase budget to keep that preference funded. Bitcoin holdings sit unchanged at 845,050 coins after two quiet weeks. Cash is lower than it was in late August because some of it went into preferred paper. The common-stock buyback sleeve is still unused.

None of that guarantees the next print. Policies can change. Markets can ignore a target. A 12 percent coupon can look generous one month and merely adequate the next if risk premia move. The honest stance is to watch the authorization, the weekly tickets, and the issuance line, not to treat $100 as a covenant.

If Stretch does settle near par and new issuance returns, this whole campaign will look like expensive but successful maintenance. If the discount reopens once official flow fades, the same campaign will look like a very large bid that could not create a market. We do not know which ending we are in yet. That uncertainty is not a bug in the story. It is the story.