Under the ledger, the interesting number in TD Cowen's revised note on Smarter Web was not the 90% implied upside. It was one word: MORE.
MORE is a preferred series. And a preferred series inside a bitcoin treasury vehicle is not a branding detail. It is a change to the liability side of the balance sheet.
Since the start of this drawdown I have been running a cohort screen on listed bitcoin treasury vehicles — market capitalization divided by the market value of held BTC, adjusted for the fully diluted share count. The median premium in that cohort has compressed by roughly a third. Vehicles that printed above 2.0x modified net asset value in the fourth quarter are now trading between 1.1x and 1.4x. Two names have crossed below 1.0x, which means every share they issue destroys value per share instead of accreting it.
That is the environment in which a small-cap treasury company announces a preferred IPO. Not a growth story. A financing story.
Context
A bitcoin treasury company has one job: convert capital-market access into bitcoin per share. The mechanism is reflexive. When the share price exceeds the per-share value of the held coins — when modified net asset value sits above 1.0 — issuing equity buys more bitcoin per share than it dilutes, and the accretion reinforces the premium that made the issuance cheap. George Soros described the loop decades before Michael Saylor industrialized it.
The loop runs backward with equal force. Below 1.0x, issuance is destruction. The flywheel becomes a grinder.
TD Cowen's argument is that a MORE preferred IPO widens Smarter Web's financing options and improves capital efficiency. That may be true in a narrow, structural sense. It is also a claim with no arithmetic attached — no coupon, no liquidation preference, no conversion terms, no redemption schedule, no share count. Sell-side research is a product, not a disclosure document, and I do not treat it as data. Due diligence is the armor against narrative hype, and a price target is narrative with a letterhead.
The preferred instrument itself is not novel. Strategy has been running a ladder of them — STRK, STRF and their siblings — for two years, selling yield to buyers who want bitcoin exposure without bitcoin's drawdown profile. Those instruments clear at Strategy's scale because they are large, liquid and index-adjacent, and income funds can size a position in them. A five- or nine-figure preferred issue from a small-cap issuer reaches a different buyer entirely: a retail account chasing a coupon, priced to clear at a spread wide enough to compensate for illiquidity.
My method for vehicles like this is deliberately narrow. I read prospectuses, share-count filings and custody-adjacent wallet flows. Everything else is commentary.
Core
Start with the arithmetic, because the arithmetic is where capital efficiency either survives or dies.
Take a representative vehicle: 100 million common shares, 2,000 BTC on the balance sheet, bitcoin at $60,000. Net asset value is $120 million, or $1.20 per share. Assume the market prices it at 1.4x, which gives a $168 million market cap.
Now issue $50 million of preferred stock and buy bitcoin with the proceeds. The company acquires 833 BTC. Holdings rise to 2,833 BTC. Bitcoin per share rises from 0.00002 to 0.0000283 — a 41% increase in coin exposure per common share. That is the number the press release will lead with, and it is real.
What it is not is free.
The preferred carries a fixed claim. Assume an 8% coupon, generous and probably optimistic for a small-cap issuer without an index mandate. That is $4 million per year in cash the company must find before a single common shareholder sees anything. The preferred also sits ahead of the common in liquidation.
Now run the scenario grid. If bitcoin rises 40% to $84,000, the coin position is worth $238 million; subtract the $50 million preferred claim and the common equity is worth $188 million, or $1.88 per share. Without the deal, those same holders would have $1.68. The leverage worked.
If bitcoin falls 40% to $36,000, the coins are worth $102 million; subtract the same $50 million and the common is worth $52 million, or $0.52 per share, against $0.72 without the deal. A 40% asset move produced a 57% equity move in both directions. The leverage is close to symmetric, which is exactly what you expect, because it is leverage. What is not symmetric is the coupon. That $4 million is owed in the rising case and in the falling case, and it is owed precisely when the asset base just fell by 40%.
That is the correct frame for MORE: balance-sheet leverage with a cash cost attached, not capital efficiency. In a market where operating cash flow at a treasury vehicle is close to zero by design, the coupon has to be met from the same well that funds the bitcoin purchases. Paying a dividend out of further issuance is sustainable while the premium holds and corrosive the moment it does not.
Coverage is the number nobody publishes. A $50 million preferred at an 8% coupon requires $4 million a year. If the vehicle holds $170 million of bitcoin after the raise and generates no operating revenue — which is the industry norm — coverage is zero from operations and depends entirely on price appreciation or further capital raising. That is not a defect unique to Smarter Web. It is the design. The dividend is a claim on the flywheel itself, not on a business.
Which instrument is rational depends on one ratio. Above roughly 1.2x modified net asset value, common issuance is accretive even after dilution and should win on simplicity alone. Between 1.0x and 1.2x, accretion is thin enough that a fixed-coupon instrument priced below the equity's implied cost can make sense. Below 1.0x, every instrument destroys value and the correct action is to stop issuing and let the balance sheet breathe. A preferred IPO announced into a compressed cohort tells you which regime management believes it is in, or which one it needs investors to believe it is in.
I ran a version of this exercise in 2022, modeling the outflow cascade from Celsius and Three Arrows. The lesson from that quarter was not that leverage is bad. It was that leverage with a fixed payment schedule and a mark-to-market asset base converts a price decline into a solvency event on a timeline nobody models in advance. Two billion dollars of stablecoin redemptions did not kill those books. The schedule did.
There is a clean institutional precedent here, and it is not in crypto. Mortgage REITs run the identical structure: a mark-to-market asset book financed with fixed-cost liabilities, marketed on spread capture, and valued on a premium-to-book multiple that expands when spreads are wide and contracts toward or below book when they are not. The sector has produced a durable high-yield product alongside a recurring cycle of equity destruction. Nothing in the structure is fraudulent. Everything in it is sensitive to the direction of one variable, and the fixed payment is due in every state of the world.
The flow evidence says something different from the headlines. Treasury-vehicle purchases have become a recognized bid in the bitcoin market, and analysts routinely describe them as structural support. Trace the custody-side movements instead of the press releases and the picture shrinks. A $50 million purchase by a vehicle this size is a rounding error against a single day of spot volume on the major venues. Its impact is not measured in coins. It is measured in sentiment, and sentiment is the only input a small issuer actually controls.
Attribution has limits worth stating plainly. Treasury-vehicle coins sit with custodians, and custodian clusters are shared infrastructure — one address cohort serves hundreds of clients. I can observe aggregate custodial rebalancing and the timing of large settlements. I cannot name the buyer from the chain alone. What the chain gives me is timing and magnitude, and the discipline of refusing to infer more than that.
I applied connected-component clustering to Ethereum wallets in 2021 and found fifteen addresses holding 12% of a flagship NFT collection, a supply concentration the community was describing as organic distribution. The same technique maps onto the shareholder registers of thinly floated treasury vehicles. When the free float is small enough, a premium to net asset value can be manufactured with less capital than it takes to sustain a large-cap equivalent. The premium that makes the flywheel spin may be a liquidity artifact rather than a statement about capital access. That distinction matters, because a premium that exists by accident does not survive contact with a large seller.
The competitive frame is unsentimental. Strategy holds the lowest cost of capital in the category, the broadest institutional ownership, the deepest preferred market and index membership that funnels passive flows whether or not anyone is paying attention. Metaplanet has a domestic capital market largely to itself. Everything below that tier, Smarter Web included, competes for the same dollars with a worse coupon, a thinner float and a shorter track record. Category winners in capital markets do not split evenly. The cheapest issuer absorbs the demand.
I built inflow models for the spot bitcoin ETFs in 2024 and watched a first-hundred-days average of roughly $450 million per day into a single issuer's trust confirm that institutional flow behaves as a supply-shock variable rather than a sentiment variable. The logic inverts here. When the marginal buyer of a treasury vehicle is a retail account reacting to a price target, the flow is reflexive, and reflexive flow reverses.
Contrarian
The obvious reading of MORE is that it adds leverage into a drawdown, and that reading is probably right. The less obvious reading is more interesting: a preferred IPO is often a confession.
A company that can issue common stock accretively at 1.4x modified net asset value would do so. It is cheaper, simpler and carries no coupon. Moving to preferred, with a fixed obligation and a senior claim, usually signals that the accretion math on common equity is thinner than the marketing implies, that the common buyer base is saturated, or that management wants to protect the premium from the dilution that would compress it. None of those signals is fatal. All of them are information.
The audience has its own blind spot. A price target is not a forecast. It is a communication instrument with a twelve-month horizon, revised most often in response to price rather than in anticipation of it. In 2017 I audited vesting schedules and inflation models across three token sales and calculated that more than 60% of supply would reach the market within two years. The data was right and irrelevant for six months, because targets and narratives both move after price. Patterns emerge only when chaos is organized, and a single analyst revision is not organization. It is noise with a title.
Which leaves the only document that matters. Code is law, but intent is the evidence — and for a preferred issue, the terms are the code. A modest coupon paired with a conversion feature tells you management expects the flywheel to keep turning and is comfortable selling part of that upside to a new class of holder. A high coupon with hard redemption and no conversion tells you management is managing a liquidity constraint and calling it capital efficiency.
Takeaway
The signals worth tracking are specific. Watch the prospectus for the coupon, the liquidation preference and whether conversion or redemption exists. Watch bitcoin per share against the dividend obligation to see whether coverage is structural or funded by further issuance. Watch the weekly modified net asset value across the small-cap treasury cohort, because the first name that issues common below 1.0x is your reversal tell. And watch the custody-side movements, because that is where intent leaves a footprint.
Ledgers don't lie. The question is whether the buyers of MORE will be reading them, or reading the target price.
