What happens when a public company treats Bitcoin like a temporary tool instead of a permanent banner? Sequans just answered that question in the most blunt way possible. It sold the last 314 coins on its books and said the treasury experiment is over. No leftover stash. No “we still believe” footnote. Just cash, a cleaner balance sheet, and a return to the chips that actually pay the bills.
Why Sequans Shut The Bitcoin Treasury Door
I have watched a lot of firms talk about digital assets as if they were part of the brand. Sequans never quite fit that mold. It is a semiconductor company first. Cellular IoT. Software defined radio. 5G modules. That is the real product line. Bitcoin sat on the side, useful when debt needed a haircut, awkward when revenue wobbled.
On September 24 the firm confirmed the last sale covered every coin still held as of June 30. After that, the crypto line on the balance sheet reads zero. In my view, that is less a market call and more a capital structure call. The company wanted convertible notes gone. It wanted a stronger cash pile. It wanted management attention back on silicon.
From Conviction Speech To Full Exit
Less than a year earlier the tone was different. In November 2025 Sequans sold 970 BTC and cut debt from $189 million to $94.5 million. Holdings fell from 3,234 BTC to 2,264 BTC. The chief executive called it tactical. Long term conviction, he said, had not changed. The firm would be in a better spot to buy more later.
That later never arrived. First quarter 2026 brought another 1,025 BTC sale as revenue slipped and losses widened. By April 30 the stack was down to 1,114 BTC from 2,139 BTC at year end 2025. Of what remained, 817 BTC sat pledged against $35.9 million of convertible debt. Realized losses from those quarter sales hit $11.7 million. Proceeds went mainly to note redemption and an American depositary share buyback.
By May the remaining convertible debt was gone. Management then sketched a wind down of the treasury plan. The last 314 coins closed the loop. No outstanding debt left except government backed research obligations. Cash up. Crypto exposure gone.
With this transition complete, we are focused on capitalizing on the strong momentum across our semiconductor business.
– Company leadership after the final sale
That line is the whole strategy in one breath. Monetize the remaining coins in a measured way. Kill the convertibles. Point every spare dollar at product.
How The Sales Unfolded Over Time
It helps to see the path as a sequence, not a single headline. First came the big November cut. Then the first quarter dump while operations looked messy. Then collateral pledged against leftover notes. Then full redemption. Then the leftover 314 coins. Each step served the same job: shrink leverage and shrink distraction.
| Period | Action | Holdings After |
| Late 2025 | Sold 970 BTC, halved debt | About 2,264 BTC |
| End 2025 | Year-end stack | 2,139 BTC |
| Q1 2026 | Sold 1,025 BTC | 1,114 BTC by April 30 |
| Mid 2026 | Redeemed remaining notes | Wind-down announced |
| September 2026 | Sold final 314 BTC | Zero BTC |
Numbers like these do not need drama. They show a firm that used Bitcoin as working capital with extra volatility attached. When the extra volatility stopped helping, the asset left the building.
Other Public Firms Sold Coins Too
Sequans is not the only name that tapped a Bitcoin pile in 2026. Several listed companies sold slices to pay debt, refill cash, or fund operations. The difference is simple. Most of them kept a large reserve. Sequans kept none.
Empery Digital sold 1,400 BTC for about $87.1 million between early May and mid July. Filings show 1,167 BTC sold in the first half for $80.1 million, with a realized loss of $56.8 million against original cost. After the quarter closed, another 1,635 BTC went out between July 1 and August 6 for $102.2 million. In that window the firm repaid $20 million of borrowings and got back 585 BTC that a lender had held as collateral. Painful on the loss line. Useful on the cash line.
Nakamoto sold roughly 600 BTC plus some derivative positions and booked about $48 million net. Most of that money repaid 45 million USDT on a Bitcoin backed loan. At the end of June it still held 4,467 BTC worth about $261.5 million. That is a trim, not a goodbye.
Smarter Web used 177.89 BTC to repay an $11.7 million convertible in July. Average sale price sat near $65,762. The note was cleared about two weeks before maturity. The firm still held 2,700 BTC and said the long term plan stayed in place.
Strategy, the largest corporate holder, sold around $218 million of Bitcoin earlier in the year to cover preferred dividends and rebuild dollar reserves. Reports later noted authorization for up to another $1.25 billion in sales tied to a broader plan that included share repurchases. Sales ran for four straight weeks into early August, including 1,638 BTC for about $104.7 million between late July and August 2. Holdings still sat above 840,000 BTC. In September the firm bought 950 BTC for $75.7 million and pushed the stack to 846,000 BTC. Sell when you must. Buy when you can. That is a different playbook from Sequans.
MARA Holdings took yet another route. The miner sold about 23,093 BTC for roughly $1.63 billion in the first half to fund operations, investments, and liquidity. It still held 35,577 BTC at the end of June. Mining firms live next to the asset. Selling inventory is not the same as abandoning a treasury thesis.
- Empery sold hard and booked large realized losses, but it did not empty the vault.
- Nakamoto paid down a coin backed loan and kept thousands of BTC.
- Smarter Web cleared a convertible and restated its long term stance.
- Strategy mixed sales with a later buy and stayed the biggest corporate holder.
- MARA funded the operating machine and still carried a sizable reserve.
Perhaps the most interesting contrast is intent. Those firms treated Bitcoin as a core reserve that can be tapped. Sequans treated it as a bridge off a leveraged structure. Once the bridge was crossed, there was no reason to keep standing on it.
What A Zero Bitcoin Balance Sheet Really Means
A zero stack is not automatically a bearish call on the asset. It can just mean the asset no longer matches the firm’s job. Sequans makes radios and IoT silicon. Customers care about 5G eRedCap, design wins, backlog, and whether the next transceiver ships on time. They do not care whether the treasury team is long digital gold.
I’ve found that markets often smear every corporate sale into one story: “companies are dumping Bitcoin.” That is sloppy. Some sales fund dividends. Some repay lenders who hold coins as collateral. Some refill a dollar reserve so preferred coupons do not force a fire sale later. Sequans is the rare case of a complete off-ramp.
The company now says product revenue grew more than 80 percent year over year in the second quarter. Six month product backlog at quarter end was more than three times the year earlier level. Work continues on the 5G eRedCap platform as the cellular IoT market shifts from 4G. A new radio frequency transceiver for software defined radio already booked a first drone design win. Interest is showing up in defense, drone, and space conversations. That is the business the board wants capital pointed at.
Is that the sexier story for crypto headlines? No. Is it the cleaner story for a chip firm that just spent a year wrestling convertibles? Yes.
Debt First, Narrative Second
Convertible notes are a peculiar kind of pressure. They sit in that gray zone between equity hope and debt reality. When the share price cooperates, they can convert and vanish. When it does not, they become a clock. Sequans used Bitcoin sales to stop that clock.
In November 2025 the first large sale halved the note pile. In 2026 more coins went out while operations were still uneven. Collateral sat against the leftover $35.9 million. Then the leftover notes were redeemed. Only after that did the last 314 coins leave. Sequence matters. This was not a panic tweet. It was a multi quarter cleanup.
The firm also ran an American depositary share buyback with some of the proceeds. That is a capital allocation choice people can argue about. Buybacks after realized crypto losses will always look messy on a spreadsheet. They can still make sense if management thinks the stock is cheap relative to the chip pipeline. I am not here to bless that call. I am here to note that the company chose balance sheet shape over treasury branding.
Cleanup sequence in plain terms: Cut the big note balance Sell more coins as losses and revenue stress showed up Unpledge collateral by retiring leftover convertibles Sell the stub position Run the semiconductor plan without a crypto sidecar
The Cost Of Using Bitcoin As A Debt Tool
Realized losses are the part people skip when they cheer a treasury launch. Sequans booked $11.7 million of them in a single quarter of sales. Empery’s first half loss against cost was $56.8 million. Those figures do not prove Bitcoin is a bad asset. They prove timing and leverage can chew the thesis.
If you buy coins with operating cash and no debt against them, a drawdown is an accounting headache. If you buy coins and then pledge them against convertibles, a drawdown becomes a funding problem. Sequans lived in the second world. That is why the exit feels final. The tool had already done the only job management still needed.
In my experience, investors give treasury stories more patience when the core business is throwing off cash. They give less patience when the core business is still proving a product cycle. Sequans is in a product cycle story now: backlog up, 5G transition, new radio wins. Keeping a volatile reserve on top of that would have been a second bet. The board chose one bet.
What This Says About Corporate Bitcoin In 2026
Corporate Bitcoin is not one strategy. It is a pile of strategies wearing the same ticker symbol. Some firms accumulate through every dip. Some sell a slice when preferred dividends come due. Some miners sell production to keep the lights on. Some industrial companies try the asset, use it to fix leverage, and walk away.
2026 made that split obvious. Sales happened. They did not all mean the same thing. Strategy sold, then bought again. Smarter Web sold a sliver and kept the slogan. Nakamoto repaid a loan and stayed long thousands of coins. Sequans sold the stub and changed the subject.
That last path will stay rare. Most boards that adopt a Bitcoin reserve do it for signaling as much as for liquidity. Walking it all the way to zero admits the signal was never the point. Fine. Honesty is underrated in this corner of the market.
- Ask whether the coins are a reserve or a financing patch.
- Ask whether debt sits against the stack.
- Ask whether the operating business can fund itself without sales.
- Ask whether management still wants the branding.
- Only then decide if an exit is a market view or a house cleaning.
Those five questions would have told you Sequans was heading for the door long before the last 314 coins printed.
The Semiconductor Story Waiting Behind The Sale
Once the coins are gone, you have to judge the company as a chip vendor again. That is healthy. Product revenue up more than 80 percent year over year is the kind of number that should have been the headline all along. A backlog more than three times last year’s level is the kind of number that tells you customers are placing real orders, not just touring the booth.
The 5G eRedCap push matters because IoT radios are moving off aging 4G designs. That transition is slow, messy, and full of design-in cycles that take quarters. If Sequans lands those sockets, the treasury chapter becomes a footnote. If it misses them, a zero Bitcoin stack will not save the equity story. Capital allocation cannot invent demand.
The software defined radio transceiver is the wildcard. First drone design win in the second quarter. Talk of defense, drone, and space interest. Those markets pay for performance and reliability, not for a clever treasury slide. I like that the company is pitching hardware into places where a missed packet has consequences. That is a different customer than a meme-driven balance sheet crowd.
Licensing opportunities were also mentioned. Licensing can be lumpy. It can also be high margin when a radio architecture gets designed into someone else’s board. Watch that line. It will tell you whether the “focus on semiconductors” line is operating reality or just a press quote.
Why Some Readers Will Hate This Exit
A slice of the market wanted every public company to become a proxy fund. Any sale feels like betrayal. I get the emotion. I do not share it. A board’s first job is not to cosplay as an exchange traded product. Its first job is to keep the operating company solvent and pointed at customers.
If you bought Sequans because you wanted levered Bitcoin with a chip logo on top, the last sale is a problem. If you bought Sequans because you wanted a small radio vendor with a less scary capital structure, the last sale is the point.
There is also a quieter group that will shrug. They never believed a mid cap semiconductor name needed a digital reserve. For them, November 2025 already told the tale. Everything after that was paperwork.
A treasury asset that exists only to fix yesterday’s leverage is not a treasury strategy. It is a one-time wrench.
That is my read. Call it opinion. The filings read the same way even if you strip the adjectives.
Lessons For Boards Still Sitting On Coins
Do not copy Sequans blindly. Copy the clarity. If the coins are there to retire expensive paper, say so. If they are there as a multi year reserve, fund that reserve with cash you can lose without pledging the stack. Mixing those two motives is how you end up selling into weakness and explaining realized losses on an earnings call.
Collateral is the tripwire. Once a lender holds your coins, your “long term conviction” has a margin schedule. Sequans pledged 817 BTC against leftover notes. That is not a philosophical position. That is a contract. Contracts expire. Philosophies get rewritten after the fact.
Communication matters too. The November line about unchanged conviction aged poorly. Better to say: we are using this asset to reshape the right side of the balance sheet, and we will decide later whether a reserve still belongs here. Investors can live with a maybe. They dislike a slogan that dies in two quarters.
- Separate reserve policy from emergency liquidity policy.
- Keep pledged coins in a different mental bucket than free coins.
- Measure the core business without the mark to market noise.
- Retire the slogan if the slogan no longer matches the cash uses.
What To Watch Next
The next Sequans story is not a price target on Bitcoin. It is backlog conversion, eRedCap design wins, and whether the new transceiver turns one drone socket into a family of sockets. Cash without crypto exposure also means future fundraising, if needed, will be judged as a chip raise, not a treasury raise. That is a cleaner conversation.
For the broader corporate Bitcoin set, watch who sells to pay paper and who sells because the thesis cracked. Those are different tapes. Strategy buying 950 BTC after a stretch of sales is one tape. Sequans printing zero is another. Pretending they are the same headline is how analysis turns into noise.
I keep coming back to a simple test. If you delete the Bitcoin line from the slide deck, does the company still have a reason to exist? For Sequans the answer is yes: radios, modules, and a 5G shift that customers already seem to be booking. That test is why the exit does not read like a crisis. It reads like a firm putting the shiny object back in the drawer.
Will other industrial names follow all the way to zero? A few might, especially those that adopted coins mainly to look modern while they nursed convertibles. Most will not. The ones that raised equity on a treasury brand have a reputation to protect. Sequans never needed that costume as badly. That is why it could take it off.
A Cleaner Balance Sheet Is Not A Magic Trick
Let me put a cap on the optimism. Killing convertible debt and selling the last coins does not guarantee the semiconductor plan works. Growth of 80 percent is easier to print from a small base. Backlog can slip. Defense and drone cycles can stall. Licensing can arrive late. A stronger cash position only buys time and options. It does not ship wafers.
Still, time and options beat a pledged stack and a note that wants cash on a fixed date. That is the unglamorous win. Markets love a myth. Operating companies need a calendar they can survive.
So here is where I land. Sequans used Bitcoin as a wrench, not as a religion. The wrench did its job. The last 314 coins were the leftover bolts in the tray. Now the workbench is the chip business again. That story is less loud than a treasury manifesto. It may also be more honest. And if you have been around this market long enough, you know honesty is the rare asset that does not need a ticker.
