I have an investment opportunity for you. I have a bag, into which I have put £100. My offer to you is that you add £3 to the bag, and in exchange I give you 1% ownership of the bag and its contents (retaining 99% myself).
Your first response might be that this seems like a poor deal; your 1% stake in the bag would be worth only £1.03. However, I say, I intend to offer the same deal (scaled up) to a series of investors – each time they pay 3% and get a 1% stake in the bag. Once I have found 56 more investors, the bag will contain £523; your stake has been diluted to 0.6% but is now worth £3.01 so you have broken even. And from that point on, the more investors put in, the more you are in profit.
Your next response might be that this sounds like a pyramid scheme. I shrug and say that the scheme stands on the merits described. You could say that it seems like it’s a much better deal for me than for you — I got to buy in at 1:1 whereas you buy in at 3:1, so I get 3x the return that you do. I shrug and say that that’s because I’m an entrepreneur and it’s my idea.
However I’m not really an entrepreneur so this is not a real investment opportunity. If the opportunity did exist, it would have potentially unlimited returns for an unlimited number of investors if the number of available investors and their money was unlimited; however like any pyramid scheme it would in reality run out of potential investors foolish enough to put their money into the bag and, at that point, all but the early investors would find that they received less money than they put in. It could nevertheless run for a while before reaching that point and be very profitable for the 2nd investor, even at a 3:1 buy in.
MicroStrategy
So let’s look at Strategy. Once upon a time, Microstrategy was a software business. However in its most recent 10Q (ref), it made $90M in revenues on its software business and $3.9B in unrealised gains on its bitcoin portfolio. So there is broad agreement that the software part of the business is vestigial and the company is a kind of bitcoin investment firm at this point.
Strategy could be or could have been a kind of open-ended investment company that owned only Bitcoin. The managerial skill required there would be limited to things like ensuring that the custodial arrangements for $60B of bitcoin were robust (a pretty hard problem on its own, I guess). But I suppose that Strategy’s management noticed that their shares were overpriced for a mere custodial firm, and saw a more active strategy available to them. So Strategy set KPIs to measure management’s performance with that active model — specifically (see their 2025Q3 10-Q starting page 42):
- Bitcoin Per Share: how much bitcoin the shareholders own per (diluted) share of common stock.
- “BTC Yield”: the percentage increase over a specified time in Bitcoin Per Share.
I’m going to put “BTC Yield” (or just “yield”) in scare quotes throughout, since, to quote the footnote from the slides in their earnings webinar (ref Third Quarter 2025 around 4:50), “BTC Yield is not equivalent to ‘yield’ in the traditional financial context.”
How can management affect their KPI, Bitcoin Per Share, and how much “yield” can they generate? Tempting as it is to put all of the algebra in here, the answer is probably easier to understand than the workings (and those that feel so inclined can check my workings over here). Looking at first-order effects only, the change in “BTC yield” over a given period is:
where Assets is the total assets held at the start of the period, Shares is the diluted common shares in issue (so the common stock plus the number of shares if all convertible bonds and preferred shares are converted), Debt is the value-at-issue of non-convertible debt/preferred shares, and PremiumEV is the premium between the company’s enterprise value (the sum of its fully-diluted market cap and its non-convertible debt) and the value of its bitcoin assets. And “I” is the rate of interest payable over the period in question.
The three terms here represent three effects on the “yield”:
- If there is a premium greater than the current leverage (see (2) later), then selling overpriced shares to buy bitcoin increases Bitcoin Per Share.
- Issuing new debt to buy bitcoin always directly increases Bitcoin Per Share.
- Paying interest on the debt (regardless of whether it is paid in stock, or paid by selling stock and remitting dollars) is new stock without new bitcoin and so reduces Bitcoin Per Share.
A quick aside on preferred shares. For this model, all the preferred shares count as debt unless they are convertible. Preferred shares pay a dividend, and must be paid ahead of common stockholders, so the yield on the preferred is like interest on a debt and we can just treat it as debt. Michael Saylor, Strategy’s chairman, is correct in saying (ref Third Quarter 2025 around 1:00:00) that the preferred stock has advantages over conventional debt for the company: it is perpetual (no refinancing risk), the dividend can be paid in stock (no cashflow risk), and the dividend can be delayed indefinitely at management’s discretion (really no cashflow risk). But these are all benefits if the company is struggling for funds; if the company is successful and valued on a going-concern basis, we should assume that it is paying its obligations. After all, the dividends compound if not paid and must be paid eventually; only in liquidation and if the common stock is being wiped out do they ultimately go unpaid, and presumably the company does not expect that to happen. STRK, since it is convertible, counts as equity in this model (as do all the legacy convertible bonds); the other preferred shares are debt.
We can now make some deductions about Strategy’s business model — that is, pursuing “BTC yield”:
1. Selling stock to generate “Yield” is a pyramidal business model
Maybe it was obvious that this would be the case. The PremiumEV for Strategy has been around 100% for much of the last 2 years; if we simplify and suppose that Debt = 0 for a moment (as it was once, since the bonds were all convertible before this year and so count as equity in this model), the company could generate roughly 20% yield per quarter by issuing 20% new stock per quarter (and it roughly did that in 2024). But that requires roughly doubling the number of shares every year, doubling the market cap and the amount of dollars invested. So you need to raise an amount of money that doubles every year from investors buying in at 2:1 to finance returns for the earlier investors… a pyramidal business model. So that cannot continue long term, or even in the medium term.
Strategy as a business is not a pyramid scheme, because it does not control its share price and is a valid bitcoin treasury business when the premium disappears — if we set aside the marketing and disappointment for investors who bought in at 2:1 who lose out when the premium falls. I’m narrowly saying that selling overpriced stock to generate a high “yield” cannot continue medium term, and that as with any pyramidal investment the late investors lose out compared to early investors. Of course, if bitcoin goes up by 30% p.a. as Strategy expects, the late investors will still make money in the end — but would have made more money by buying bitcoin directly.
Setting aside the morality of it, as a character says in Margin Call, “we are selling to willing buyers at the current fair market price.” Strategy can generate returns for current investors by selling stock and buying bitcoin and arguably it should do so as long as the premium exists. It should not have implied that investors can get a large “BTC yield” in the long or medium term, and without that investors should not be buying in at 2:1 or 3:1. But anyone buying into Saylor’s “structured finance company” (ref Third Quarter 2025 around 36:45) — or into any crypto — should be a sophisticated investor and able to calculate the value that they are getting.
Maybe the opportunity is already exhausted. The premium has fallen to +13% recently and it would seem to me odd to ever buy in at 2:1 on an investment if I knew that it sometimes traded at close to 1:1. (Though in fairness I would have just bought bitcoin rather than buy in at a premium in all cases if I believed the investment premise, so clearly the people buying MSTR don’t think at all like I do.)
2. The relevant premium is comparing market cap to bitcoin value.
Interestingly, the EV premium minus the leverage is the same thing as the premium of (fully diluted) market cap vs gross bitcoin assets. That is not a meaningful financial ratio normally (it’s the premium if you pretend that there is no debt), but it happens to be the right value here. That’s because issuing new shares, even at a premium to EV, dilutes existing debt and so decreases the yield provided by the bitcoin bought with debt.
The Financial Times is way ahead of me here. I was wondering why in their articles they drew attention to this strange ratio, with cryptobros flaming them in the comments of every article when they mention it. Yet the cryptobros are not entirely wrong, I think; for now, while the company can issue non-convertible preferred shares (debt, in this model) as well as common stock, it can issue both to maintain its leverage; if you fix the level of leverage by issuing both shares and debt in the right proportions, then the debt-over-assets term from issuing shares and the delta debt term will cancel out.
3. Rising price of bitcoin does not generate “Yield”
Not by itself, anyway. Stockholders’ existing Bitcoin Per Share already entitles them (loosely speaking) to any gain in the (dollar) value of the bitcoin held for them (further increased by the leverage). To generate “BTC Yield”, the company has to do more than sit on appreciating bitcoin.
4. Issuing debt can generate “Yield”
Strategy is expecting a 30% p.a. growth in the dollar value of bitcoin. If we assume that the markets have no trouble with Strategy maintaining its current level of leverage, then it can also issue 30% more debt per year (that’s 30% more debt, so at 20% starting leverage that is 6% of its start-of-year market cap in new debt and new bitcoin), which can be in part used to buy bitcoin and generate “Yield” without raising its leverage. Or around 1.3% “Yield” per quarter.
So that’s why they are issuing preferred stock this year: this is a perfectly fine way to boost their KPI, if bitcoin in the medium term is going +30% p.a..
5. Strategy depends on an increasing price for bitcoin over the medium term
Interest is a drag on “Yield”, and it becomes a serious drag if the price of bitcoin does not grow as fast as the percentage rate of debt issuance. That drag slightly more than a -1% per annum “Yield” today according to Strategy’s website, and will grow unless it allows the leverage to fall (because it would have to replace low-yield convertible debt with non-convertible preferred shares yielding 10%).
In taking on debt, Strategy has moved from a very safe business model — selling shares to buy and hold bitcoin, which is safe if your benchmark is the price of bitcoin — to one that depends on bitcoin gaining value over the medium term. Specifically, to keep the level of leverage and the drag on “Yield” from growing, we need in the medium term for the growth of the price of bitcoin to be greater than the interest on the debt — around 10% for the current preferred stock.
Debt increases Strategy’s direct benefit from future bitcoin price increases, but impairs its ability to generate future “yield”.
A Valid Business Model
Is there a valid business model here? It seems clear that there are a couple of valid business models: buy bitcoin without leverage and act like a bitcoin ETF, as IBIT does; or bet on bitcoin appreciating and buy bitcoin with leverage and maintain that leverage, generating some “BTC Yield” as the price appreciates.
If bitcoin goes up by 30% p.a. over the medium term as Strategy expects and can sustain the current 20% leverage, then it can generate a “BTC yield” of 1.3% per quarter. And once the stock premium is gone, Strategy can sell bitcoin and use the proceeds to buy back its own stock (and to pay preferred dividends), which generates a small amount of “BTC Yield” (provided bitcoin appreciation exceeds the interest drag) without needing to raise more capital.
What about the long term — what if bitcoin gains for now but hits a price ceiling later? That’s not a disaster: Strategy can sell some bitcoin to redeem all of the non-convertible preferred stock, as the terms of the preferreds allow it to do. That is a big one-off negative “BTC yield” to redeem the debt and remove the drag of the interest payments. But it benefited from the leverage in the meantime in this scenario.
So there is a valid business model here that is a natural evolution of Strategy’s current business model, if you agree with Strategy that bitcoin will continue to appreciate in the medium term. It is not enough better than owning bitcoin directly to be worth paying a premium for, as far as I can see (given that individual investors can achieve the same bitcoin exposure by buying bitcoin directly).
At this point, I wondered if the real reason that Strategy introduced Bitcoin Per Share as a KPI was to prepare its cryptobro investors (I assume that the people buying in at 2:1 or higher premiums must be cryptobros) for a switch to eventually selling bitcoin, with continuing to increase Bitcoin Per Share as the carrot — you, the individual investor that hodls, are still getting more bitcoin, at the expense of the fools that sold their stock back to us!
But will Strategy do these things? From Michael Saylor’s “21 rules of bitcoin” (Michael Saylor tweet 29th Nov):
and he has tweeted the same in the past (e.g. “You do not sell your bitcoin” Michael Saylor tweet 27th March). I guess it’s doomed then? Obviously that cannot work: if Strategy promises “BTC yield” and its CEO is saying that he will never sell, then it must grow its pile exponentially (“yield” compounds) until it owns every bitcoin (there is a finite supply) — and that would kill bitcoin. Is this guy serious?
Newsflash, as I was proof reading this post, 2025-12-01: “Strategy said if its mNAV falls below one, the company would sell bitcoin to fund its dollar reserve…” (ref). WTF? The CEO tweeted the opposite 2 days ago. Is Saylor’s X feed just a sewer of AI-generated crap that he feeds to the gullible/faithful? Is this how businesses are run nowadays?
So, what now? Do cryptobros rush for the exits because Saylor was bullshitting them? There is no point buying MSTR now if there is a good chance that it will be available at a discount (relative to bitcoin) later, because the cryptobros fell out with Saylor.
That said, I cannot agree with the doom-laden tone of articles like “Crypto hoarders dump tokens as shares tumble“. None of this is fatal for companies that currently hold tradeable assets with low leverage and are, or at least should be, benchmarked against bitcoin. If the bitcoin price crashes and the stocks of these companies crash, they should still be able to track bitcoin fairly well which is the thing that an asset-holding fund should do. While I am dismayed by the pseudo-religious language of Saylor’s comments like “Volatility is Satoshi’s gift to the faithful,” some of his saner comments are correct: “if you are going to hold the asset, you need a four year time horizon [at a] minimum” (ref).
Even if no investor ever buys their stock again, regardless of discount, a sensibly-managed low-leverage DAT can still sell bitcoin to deleverage and to buy back their own stock if a discount develops. This is value enhancing if the stock trades at a discount (relative to bitcoin), and gives a way for shareholders eager to exit to depart without extreme losses (relative to bitcoin). Keep doing that forever, or convert to an ETF and let the market makers close the discount. That’s an unglamourous end, but not a disaster. A decline in share prices to a level where they reflect the assets held may seem like a disaster to investors that bought in at 2:1 or 3:1 overvaluations, but that was already a loss from the moment that they hit buy.
Disclosure: I do not own any crypto or stocks primarily invested in crypto currently or recently.
Further Reading/Listening
The Bitcoin Treasury Reckoning — a fairly reasonable take I think, though I dislike that it discusses the idea of a sell-off death spiral without the caveat that this would be a short to medium term price movement only.