I want to begin with a number almost nobody paused to read this week, because it tells you more about the state of this market than any price candle ever will.
Strategy โ the company you still probably call MicroStrategy โ has spent the last two years trading at a premium to the value of the Bitcoin it holds. For most of 2024, that premium was the entire story. It was the reason a software company from Virginia could issue convertible debt on terms that would make a normal chief financial officer weep with envy. It was the reason it could sell shares into the open market, take the proceeds, and buy more Bitcoin, in a loop that felt less like corporate finance and more like a perpetual motion machine. The premium was not a side effect of the strategy. The premium was the strategy.
And then, quietly, the premium stopped being a premium.
I have been watching these instruments for eighteen years now, and the one habit that has saved me more money than any model is this: follow the fear, not the chart. The fear here is not that Bitcoin is falling. The fear is that the mechanism itself โ the flywheel that turned equity issuance into an accumulation machine โ is losing its grip. Billions of dollars are flowing back into this market. You can see it in the stablecoin supply, in the spot ETF inflows, in the funding rates on perpetual futures that nobody wants to discuss at dinner parties. The money came back.
The premium did not.
That gap โ billions returning while the premium refuses to return โ is the most honest thing happening in crypto right now, and almost nobody is talking about it, because it does not fit the mood of a bull market. A bull market is supposed to be a rising tide that lifts everything. Instead we are watching a tide come in while a specific class of boats stays stuck in the mud. That is not a rounding error. That is a signal about what kind of market this actually is.
So let me spend the next several thousand words on it, carefully, the way I would audit a contract rather than the way I would write a tweet. Because the difference between those two modes is the difference between understanding this cycle and being carried by it.
The Machinery You Were Never Shown
Before we can talk about why the premium is dying, we have to be precise about what it is. The term floating around is mNAV โ modified net asset value. The definition is simple on paper: take the market capitalization of a listed company, divide it by the net value of the crypto assets it holds, and you get a ratio. If the ratio is above one, the market is paying more for the company than the coins are worth. If it is below one, the market is paying less.
Simple on paper. Brutal in practice.
A crypto treasury company, in its purest form, is a listed equity whose primary business is holding digital assets. Strategy is the archetype. Metaplanet in Japan is the imitator. There are now dozens of them scattered across exchanges in Tokyo, Toronto, Stockholm, and the backwaters of the American over-the-counter market. They are not technology companies. They are not exchanges. They are balance sheets wearing the costume of a business, and their share price is a referendum on whether the market believes the balance sheet can keep growing.
Here is the part that took me a long time to internalize, and the part that most retail investors never internalize at all: the premium is not a valuation of the company's intelligence. It is a valuation of the company's ability to raise capital. That is a subtle but enormous distinction. When a treasury company trades at two times its net asset value, the market is not saying the coins are worth twice as much. The market is saying that this vehicle has figured out how to convert shareholder enthusiasm into additional coins more efficiently than you can by simply buying the coins yourself.
The mechanism runs like this. A company trades at a premium. That premium makes its equity a cheap currency. The company issues more equity โ sometimes through an at-the-market program, sometimes through a convertible bond, sometimes through a preferred share with a carefully engineered coupon. It takes the proceeds and buys Bitcoin. The Bitcoin holdings per share rise. Because holdings per share rise, the market gets excited, and the premium persists or even widens. Which makes the equity an even cheaper currency. Which lets the company issue more. Which buys more coins. Which raises holdings per share again.
This is a reflexive loop, and like all reflexive loops, it depends entirely on the direction of travel. George Soros understood this decades before anyone applied it to crypto: the loop is self-reinforcing on the way up and self-destroying on the way down. The same mechanism that manufactures the premium on the ascent manufactures the discount on the descent. There is no stable middle. There is only acceleration in one direction or the other.
I audited my first Solidity contracts in 2017, sitting up late at night reviewing a multi-signature implementation because I wanted to protect early adopters from centralized points of failure. I found twelve logic flaws that nobody had bothered to look for. What that experience taught me โ the lesson I have carried into every piece of writing since โ is that the danger is never in the part of the system that everyone is watching. It is in the part of the system that everyone assumes is fine. The treasury company premium is exactly that kind of assumption. Everyone watches the Bitcoin price. Almost nobody watches the premium, because the premium feels like a natural property of the company rather than a fragile social agreement.
But it is a social agreement. And social agreements can be withdrawn.
Why the Premium Existed At All
To understand why the premium is disappearing, you have to be honest about why it was ever there. And the honest answer is uncomfortable for anyone who wants to believe this market runs on fundamentals: the premium existed because, for a long stretch of time, there was no better way to get leveraged exposure to Bitcoin inside a traditional brokerage account.
Think about who buys Strategy. It is not the person with a Coinbase account and a hardware wallet. It is the pension fund analyst in Ohio who cannot buy spot Bitcoin directly, the family office in Singapore whose mandate forbids holding unregulated assets, the retail investor in Europe who wants Bitcoin exposure inside their tax-advantaged retirement wrapper. These people wanted Bitcoin. The path of least resistance, for years, ran through an equity that held Bitcoin on its balance sheet. The premium was the toll they paid for that convenience.
And the toll was not small. A two-times premium means you are paying two dollars for one dollar of coin. That is a steep price for convenience, and it only made sense because the alternative โ buying coins directly โ was, for these investors, not actually available. The premium was a tax on regulatory friction. It existed because the friction existed.
The moment you frame it that way, the rest of this article writes itself. Because the single largest structural change in this market over the last two years has been the removal of that friction. Spot Bitcoin ETFs, approved and now deeply liquid, give exactly the same class of investor exactly the exposure they wanted โ inside a brokerage account, inside a retirement wrapper, with regulated custody and a clean fee line โ without paying a premium over net asset value. In fact, they often trade at a tiny discount or at par.
This is the substitution that nobody wanted to name out loud. When IBIT and its siblings reached scale, they did not just compete with the treasury companies. They ate the treasury companies' reason for existing. Why would you pay two dollars for one dollar of coin when a regulated fund will sell you one dollar of coin for one dollar, minus a management fee that rounds to nothing?
You would not. And the market is, slowly and then suddenly, deciding that it will not.
The Flywheel, Running in Reverse
Here is where the mechanism becomes genuinely dangerous, and where I want you to follow the fear rather than the chart.
When the premium is positive, issuing equity is accretive. You sell a dollar of stock, you buy a dollar of coin, and because the stock was trading at a premium, the coins you bought are worth more per share than the dilution you created. Holdings per share go up. Everyone wins. The flywheel spins.
When the premium compresses to zero, issuing equity becomes neutral. You sell a dollar of stock, you buy a dollar of coin, and holdings per share are unchanged. The flywheel stalls. No harm done, but no magic either.
When the premium goes negative โ when the company trades at a discount to net asset value โ issuing equity becomes destructive. You sell a dollar of stock, you buy a dollar of coin, and you have diluted existing holders to buy coins they could have bought themselves more cheaply. Holdings per share fall. The flywheel now runs backward, and it runs backward faster than it ran forward, because the market watches the ratio and reacts to it in real time.
This is the part that the Strategy bulls never want to discuss on a podcast. The strategy is not symmetric. It is a leveraged long with a convex payoff on the way up and a concave collapse on the way down. The premium is the leverage. Remove the premium and you are left with a software company that happens to hold a lot of Bitcoin, and whose equity is now an inferior way to own that Bitcoin.
I have seen this movie before, in a different theater. In the DeFi Summer of 2020, I watched a governance token crash wipe out my own modest savings and the savings of friends in my Beijing study group. I did not retreat into cynicism. I did something more useful: I interviewed thirty affected retail users, one by one, and wrote a series I called "The Psychology of Impermanent Loss," because I wanted to understand why intelligent people kept walking into the same trap. What I learned is that people do not buy the mechanism. They buy the story of the mechanism. They buy the feeling that they have found a machine that turns conviction into money. The machine is real โ until it is not โ and when it stops, the story that sustained it is the last thing to die.
The treasury company premium is a machine of exactly this kind. It turns belief into coins, as long as the belief holds. And the belief is currently being tested by the most ruthless judge in finance: the spread.
The Numbers Nobody Puts in the Deck
Let me get specific, because the generalities are where people hide.
When you compute an mNAV for a treasury company, you are making a series of assumptions that the company rarely spells out. First, you have to decide what counts as net asset value. Do you include only the coins? Or do you include the operating business, the cash, the debt? Do you mark the coins at the current spot price, or at the average price the company paid, or at some blended figure the company prefers? Each choice changes the ratio, and companies are not always transparent about which choice they made.
Second, you have to decide what to do with the debt. Most treasury companies fund their coin purchases with convertible bonds or preferred equity. That debt sits ahead of the common equity in the capital stack. A naive mNAV that ignores the debt will overstate the premium. A rigorous mNAV subtracts the debt first. I have seen decks where the two numbers differ by thirty percent, and the deck always shows the flattering one.
Third โ and this is the one that keeps me up at night โ you have to decide how to treat the dilution from future issuance. If a company has authorized shares it has not yet sold, and a convertible bond that will convert into more shares, then the true per-share coin count is lower than the headline number. The premium is being calculated on a share base that does not yet exist. This is not fraud. It is the ordinary optimism of corporate finance. But it is exactly the kind of thing that turns a comfortable premium into a shock the day the market decides to do the math properly.
I spent years auditing code, and one of the things that code teaches you is that the edge case is the case. The scenario the developer did not consider is the scenario that takes the system down. The same discipline applies here. The premium looks comfortable under the company's preferred assumptions. It looks very different under the assumptions that a short seller, or a newly cautious institutional investor, or an auditor with a legal obligation to be precise, will apply.
And that is the environment we are now entering. Not a crash. A re-pricing. A slow, grinding, unfashionable re-pricing in which the market stops taking the company's word for the premium and starts doing its own arithmetic.
Kalshi and the Forty Billion Dollar Question
The treasury companies are the clearest example of the premium compression, but they are not the only thing happening. Two other capital events deserve attention, and both of them tell us something about where this market's center of gravity is moving.
The first is Kalshi, the regulated prediction market, reportedly seeking a valuation in the range of forty billion dollars. I want to be extremely careful here, because the number is a rumor, and I have learned that the difference between an ask and a transaction is the difference between a wish and a fact. A forty billion dollar target in a negotiation is not a forty billion dollar outcome. It is an opening position. Treat it accordingly.
But even as a rumor, the number is instructive, because it tells you what kind of asset the market thinks Kalshi is. And the answer is not a technology company. Kalshi's real asset is a license. It holds a Designated Contract Market designation from the Commodity Futures Trading Commission, which is the legal permission to offer event contracts to American users. That designation is the moat. It is why Polymarket, for all its on-chain elegance, has spent years unable to serve the largest regulated market in the world.
This is worth pausing on, because it contradicts something the industry likes to tell itself. We like to believe that crypto's advantage is cryptographic. We like to believe that decentralization is a technical property that confers a durable edge. But Kalshi's advantage is the opposite of cryptographic. It is a piece of paper from a government agency. The moat is regulatory, not technical, and regulatory moats are double-edged in a way that code moats are not. A regulatory moat can be revoked by the same authority that granted it. Ask the companies that lost their licenses when the CFTC changed its mind about election contracts, or when the SEC shifted its stance on a category of token.
If Kalshi really is being valued at forty billion dollars, then the market is capitalizing a regulatory privilege at an extraordinary multiple. That is a bet on the durability of a policy, not a bet on the superiority of a product. And policies are made by people who get replaced.
Here is the thing about code as law, and I say this as someone who has spent a decade auditing the gap between the two: in every DAO I have examined, the smart contract upgrade rights ultimately sit with a handful of multi-signature admins. The code is not the law. The code is the surface, and the law is a few people with keys. Kalshi makes this explicit rather than hiding it. Its entire value proposition is that a few people with keys โ regulators โ have decided to let it operate. That is a legitimate business. It is also a business whose foundation is a decision rather than a proof.
Blockchain.com and the Return of the Public Market
The second capital event is Blockchain.com's reported intention to pursue an initial public offering. This one I find more interesting, and more structurally significant, than the Kalshi rumor, because an IPO is not a valuation claim. It is a disclosure event.
When a company goes public, it must file an S-1. It must open its books. It must reveal its revenue, its customer concentration, its regulatory exposure, its litigation, its related-party transactions, and โ most importantly for our purposes โ how it actually makes money. The crypto industry has spent a decade operating in a fog of selective disclosure. An IPO burns off that fog, at least for one company, and the light it casts will fall on everyone else in the sector.
Blockchain.com is an old name. It has been around since the early days, and its business spans wallets, exchanges, and institutional custody. That breadth is often described as a strength, and it is โ up to a point. But breadth in a competitive market is also dilution of focus. In the exchange business, it competes with Coinbase, with Kraken, with the American arm of Binance. In custody, it competes with the institutional primes. In wallets, it competes with a hundred free alternatives. The question an S-1 will force it to answer is which of those businesses actually generates the margin, and which are just expensive marketing.
But the real significance of the IPO is not Blockchain.com specifically. It is the signal. If a major CeFi company can return to the public markets, it means the institutional plumbing โ the auditors, the underwriters, the exchanges, the index providers โ has decided that crypto companies are again acceptable counterparties. That is a rehabilitation of the entire sector's relationship with traditional finance, and it matters far more than any single company's valuation.
There is a cost, though, and it is the cost that the bull market never mentions. When more crypto companies list, the supply of crypto equities increases. More supply, in a market with a fixed amount of capital, means more competition for that capital. The premium that any single treasury company enjoys is partly a function of scarcity โ of being one of the few clean ways to get exposure. Flood the market with crypto equities and you dilute that scarcity. The IPO window, in other words, is not unambiguously good news for the incumbent premium holders. It is good news for the sector and bad news for the moat.
The Shadow DeFi Nobody Calls DeFi
Here is an observation I have not seen made anywhere, and I think it deserves to be made carefully.
The treasury company flywheel is, structurally, a lending market. The company borrows against its crypto holdings โ through convertible debt, through preferred equity, through the various instruments that let it raise cash without selling coins โ and it uses the borrowed capital to buy more coins. That is leverage. That is collateralized borrowing. That is, in every functional sense, a DeFi lending protocol wearing a corporate charter.
And if you accept that framing, then the interest rate models these companies face deserve the same scrutiny I have long applied to Aave and Compound. Because here is a thing I have believed for years and rarely say out loud: the interest rate models in the major DeFi lending protocols are not derived from anything. They are arbitrary curves โ a kink here, a slope there โ chosen by a handful of developers and ratified by governance votes that almost nobody reads. They do not measure the real supply and demand for credit. They produce a number and then the market pretends that number is a price.
The treasury companies have the same problem in a different costume. The coupon on a convertible bond is not a market-clearing interest rate. It is a negotiated number, informed by the company's equity volatility, by the demand for the bond from yield-hungry funds, and by the underwriter's desire to close the deal. When the premium is high, the company can issue converts at a low coupon because the conversion option is valuable. When the premium compresses, the coupon must rise, because the conversion option is worth less. The cost of capital is not a constant. It is a function of the very premium we have been discussing.
This is the reflexive loop again, seen from the financing side. The premium determines the cost of capital. The cost of capital determines the accretion. The accretion determines the premium. Break any link and the whole chain reacts. And the market is currently testing the first link โ the premium โ which means the second and third links are about to move whether anyone is ready or not.
The Layer Two Parallel
I want to bring in one more thread, because I think it clarifies the shape of what is happening, and because it is a place where I hold a contrarian view that I have been sitting with for a while.
When Dencun shipped and blob space became cheap, the entire Layer Two narrative shifted. Rollups that had been charging meaningful fees suddenly found their fee revenue collapsing, because the cost of posting data to Ethereum fell by an order of magnitude. The market celebrated. Cheaper transactions, more users, a healthier ecosystem. All true, as far as it goes.
But here is what I think the celebration missed. The fee revenue that rollups lost was not just a cost to users. It was a revenue stream that funded the rollups' own token incentives, their developer grants, their marketing, their very existence. When that revenue collapsed, the rollups had to replace it โ and the replacement was token emissions, which is to say, the dilution of their own holders. The cheaper fees were paid for, in part, by the people who bought the tokens.
And I will say the thing that I believe will be obvious in hindsight: blob space will saturate. It is already saturating. When it does, the cost of posting data will rise again, and the rollups will face a choice between raising fees โ which they cannot do without losing users โ or eating the cost โ which they cannot do forever. The era of cheap rollup transactions is a temporary condition, not a permanent state. It rests on a supply of blob space that is growing more slowly than the demand for it. Follow the fear, not the chart: the fear is that the economics of every rollup you use are less durable than the user experience suggests.
The parallel to the treasury companies is exact. In both cases, a mechanism is producing a benefit โ cheap fees, or a high premium โ that is funded by something unsustainable. In both cases, the market is celebrating the benefit while ignoring the funding. And in both cases, the moment the funding runs out, the benefit disappears, and the people who were enjoying it are the last to understand why.
If you can see the mechanism, you can prepare for the turn. If you cannot, you will experience the turn as a betrayal, which is the most expensive way to experience anything.
The Accounting Room Where the Lights Are Off
Let me talk about accounting, because it is the least glamorous part of this story and therefore the part where the most important things are hiding.
How should a company account for the crypto it holds? For years, the answer under American rules was: as an intangible asset, marked down when prices fell but never marked up when prices rose. This asymmetry was brutal for treasury companies, because it meant their balance sheets reflected the bad years in full and the good years not at all. A rule change allowing fair value accounting โ marking the coins to market in both directions โ was celebrated as a victory.
It is a victory. It is also a complication. Because fair value accounting means the company's reported earnings now swing with the price of Bitcoin, and reported earnings feed into the multiples that analysts apply, and the multiples feed into the share price, and the share price feeds into the premium, and the premium feeds into the flywheel. The accounting change did not just make the balance sheet more honest. It wired the income statement directly into the reflexive loop.
This is the kind of thing that only becomes visible when you look at the whole system rather than any single line. A company that reports enormous fair-value gains in a bull market looks profitable. That profitability is cited by analysts as justification for a premium. The premium justifies further issuance. The issuance buys more coins. The coins generate more fair-value gains. The gains look like earnings. The earnings justify the premium.
Do you see the circle? It is a closed loop, and at every point in the loop there is a number that appears to be an independent fact but is actually an output of the loop itself. This is not a criticism of any particular company. It is a description of a system, and systems of this kind are fragile in a way that their participants cannot see from inside.
I have audited enough code to know that the most dangerous bugs are not the ones that crash. They are the ones that produce plausible-looking output from corrupted input, so that nobody notices until the error has compounded beyond repair. The treasury company premium is that kind of bug. It produces a plausible-looking number โ the mNAV โ from inputs that are themselves derived from the number. The market looks at the number, finds it plausible, and proceeds. And then one day the number is no longer plausible, and everyone discovers at once that the plausibility was the only thing holding it up.
Who Actually Controls This
I want to name the governance question directly, because it is the thread that ties all of this together, and because it is the question that the industry most consistently avoids.
We talk about these companies as if they were autonomous mechanisms, governed by their strategies and their treasuries. They are not. They are governed by boards and executives, who decide when to issue, how much to issue, what to buy, and how to present the numbers. The flywheel is not a law of nature. It is a series of decisions made by a small number of people, and those people have incentives that do not perfectly align with the holders'.
When a company is trading at a premium, the executives have every incentive to issue more equity, because issuance at a premium is accretive and the market rewards accretion. When the premium compresses, those same executives have every incentive to avoid issuance, because issuance at a discount is destructive and the market punishes dilution. So the mechanism automatically tightens exactly when the company most needs capital โ which is to say, the mechanism is procyclical by design, amplifying the boom and starving the bust.
This is the same structural flaw I have pointed to in DAO governance for years. The code is not the law. The upgrade rights sit with a few multisig holders. The treasury company's strategy is not a law either. It sits with a board that can change it the moment the incentives change, and the incentives change the moment the premium does. The illusion of a mechanical, self-sustaining strategy is exactly that โ an illusion, sustained by a favorable market and dissolved by an unfavorable one.
If you can hold two things in your head at once โ that the strategy is genuinely clever, and that it depends on a social agreement that can be withdrawn โ then you understand this market better than most of the people running it.
The Contrarian Turn
Now let me argue against myself, because that is the only way to know whether a thesis is real.
The bearish reading of everything I have written is that the premium's death is a warning, that the treasury companies are fragile, that the whole structure is a house of cards waiting for a breeze. That reading is available, and it is not wrong, exactly.
But here is the contrarian angle I keep coming back to, and it is the reason I am cautious about the easy conclusion. The disappearance of the premium is not the same as the disappearance of the market. It is the disappearance of a particular way of paying for things โ paying for exposure with a story premium rather than with fundamentals. And the fact that the market has stopped paying that premium does not mean it has become rational. It means it has become differently irrational.
Consider what the capital is doing instead. The billions that came back did not come back into the treasury companies. They came back into the liquid, tradable, easily-exited instruments โ the ETFs, the majors, the perpetual futures. That is not a market that has learned to value fundamentals. That is a market that has learned to value liquidity. It is a market that will pay up for the ability to enter and exit quickly, and will refuse to pay up for the ability to hold a story over time.
That is a real change, but it is not a maturity. It is a shift in what the market is willing to be paid in. The treasury company premium was a long-duration instrument: you paid the premium today in exchange for the expectation of compounding exposure tomorrow. The liquidity trade is a short-duration instrument: you pay for optionality, and you do not commit. A market that prefers the second to the first is not wiser. It is more impatient. And impatience, in the end, is its own kind of bubble.
So the story I want to leave you with is not that the market has become rational. It is that the market has become fast, and the treasury companies are slow, and speed beats slowness at the moment of transition, which is the moment we are in. The premium is not dying because investors have seen the light. The premium is dying because investors have found a faster way to do the same thing, and the faster way does not require them to trust anyone's arithmetic but their own.
What I Would Watch
If I were managing a book โ and I am not, this is not advice, this is the discipline of a person who has watched too many mechanisms break โ here is what I would watch, in order of importance.
The first is the mNAV of the largest treasury companies, computed honestly, with the debt subtracted and the dilution included. Not the number the company shows you. The number you calculate yourself. If it drops below one and stays there, the flywheel is not merely stalled; it is running backward, and the consequences will show up in the equity before they show up anywhere else.
The second is the spot ETF flows, because they are the upstream pressure source. Every dollar that flows into an ETF is a dollar that did not need to pay a premium to a treasury company. If the ETF flows stay strong while the premium stays weak, the substitution thesis is confirmed, and the treasury companies are in structural trouble regardless of what Bitcoin does.
The third is the IPO calendar. Not because any single IPO matters, but because the aggregate supply of crypto equities is the thing that dilutes the scarcity that premium holders depend on. A wave of listings is a slow-acting poison for the premium, even as it is a vote of confidence for the sector.
The fourth is the regulatory posture, and specifically the treatment of event contracts and crypto equities. Kalshi's valuation, if it is real, is a bet on a policy holding. Watch the policy, not the pitch deck.
And the fifth โ the one I would watch most closely, because it is the one nobody watches โ is the accounting. Fair value marks wire the income statement into the loop. If a company's reported earnings are rising faster than its coin holdings, that is not growth. That is a loop, and loops are fragile. Follow the fear, not the chart: the fear is in the footnote, not the headline.
The Takeaway
The billions came back. The premium did not. And the gap between them is the story of this cycle โ a market that has learned to move capital quickly and has forgotten how to pay for patience. The treasury companies built a machine for converting belief into coins, and the machine worked beautifully for as long as belief held. But belief is a social agreement, and social agreements are repriced by the spread, not by the story.
I do not think this is the end of the treasury companies. I think it is the end of the free premium โ the era in which the market paid them for the convenience of existing. What comes next is harder and more honest: companies that have to justify their valuation with something other than the reflexivity of their own share price. Some of them will. Most will not.
And if you can see the mechanism rather than the narrative โ if you can follow the fear rather than the chart, and read the footnote rather than the headline โ then this transition is not a threat. It is the clearest signal this market has sent in years. The money is back. The story is gone. And what replaces the story is the only thing that was ever going to survive: the arithmetic.