Take the Zero Off the Table

It mostly goes one of two ways

A normal stock can drift sideways for a decade and bore you to death. Bitcoin is unlikely to grant you that mercy. Over a long enough horizon it is closer to a binary than to a normal asset: either it succeeds in becoming a major monetary and reserve asset, or it decays toward zero.

That claim deserves an argument, not a stipulation – because the natural skeptical move is to deny the binary outright. Why can't bitcoin just plateau, the doubter asks: settle into a volatile, one-or-two-trillion-dollar "digital gold lite," neither conquering the monetary world nor dying, for decades? The honest answer is that the doubter has a point, and it has to be conceded before it can be answered. A multi-trillion-dollar plateau is genuinely possible – gold itself is the proof, having sat for decades as a volatile, non-monetising store of value worth trillions, neither remonetising the world nor going to zero. So this is not a clean either/or, and pretending it is would be the kind of overclaim this essay is supposed to be against. The plateau can happen.

The defensible claim is weaker, and it is all the asymmetry case actually needs: that a store of value is a coordination game, and coordination games make the dynamics reflexive, so the distribution of outcomes is heavily skewed toward the two tails. Money is worth holding only to the degree others will hold it later – its value is the expectation of other people's expectations, a Schelling point that everyone converges on precisely because they expect everyone else to. A monetary network that is visibly winning pulls in the next holder, the next treasury, the next sovereign, each of whom makes it safer for the one after – adoption compounds adoption. One that visibly stalls runs the same loop in reverse: if it is not gaining as the obvious long-term store of value, the rational move is to migrate to whatever is, and each departure makes the next more sensible. The good tends either to keep gathering the monetary premium toward itself or to bleed it to a rival. So the middle is not impossible – it is just unstable, a ball balanced on a ridge between two valleys, far likelier to roll than to rest. The "muddle through unchanged at today's price" outcome – the one our instincts reach for because it feels safe and moderate – is the thin slice of the distribution, not the fat part. Yet it is the silent assumption underneath most of the criticism: a tacit belief that the thing can just sit, neither vindicated nor dead, at exactly this price, indefinitely. You do not need the binary to be airtight for the bet to be asymmetric. You only need the mass of the probability to sit out at the tails – and a reflexive coordination game is exactly the kind of thing whose probability does.

It is winning a winner-take-most argument,
or it is quietly bleeding out.

This is also why the debates never resolve. Watch any honest discussion of the pro-bitcoin or pro-MSTR case and you will see the same move, again and again: a person evaluates a sub-argument – the volatility, the corporate structure, the regulatory risk – and then, mid-thought, re-imports the whole zero-scenario back into it and concludes "well, but if it all goes to zero, none of this matters." Of course it doesn't. Nothing does, in that branch. Smuggling the zero into every sub-question doesn't make you rigorous; it makes you paralysed. Every argument collapses into the same argument, and you never get to the part where you actually decide anything.

The middle path – sideways forever at today's price –
is the thin slice, not the fat part.

Name it once, then set it aside

The disciplined move is not to deny the zero. The zero is real, and pretending otherwise is how people get hurt. The move is to give it its due exactly once – and then stop letting it contaminate every other thought.

Separate the two questions that everyone insists on tangling together. The first is the survival question: can this asset go to zero, and if it does, do I survive it? The second is the magnitude question: if it does not go to zero, how large is the other branch? These are different questions and they want different tools. The first is answered with position sizing. The second is answered with reasoning about adoption, scale, and time. You cannot answer them at the same time, because you cannot see the upside clearly while you are flinching at the downside in every sentence.

Survival is answered with position sizing.
Magnitude is answered with everything else.

So answer survival first, and answer it honestly. Assume the zero can happen. Size your position so that if it does, your life is unchanged – a sum you can lose completely and still sleep, still retire, still send the kids to school. Once the position is small enough that the zero is survivable, the zero has been paid for. You have bought the insurance. You do not need to re-purchase it in the middle of every subsequent paragraph. Take it off the table, and the second question finally comes into focus.

The order matters, because the two tools can disagree, and the resolution of that disagreement is the whole rigorous core of the essay. Naive expected value, run alone, will happily tell you to bet everything on any wager with positive expectation – and for an asset that can genuinely go to zero, "everything" is how you get wiped out before the good branch ever arrives, never to recover. A single zero ends the game; you do not get the long run if you do not survive the short one. Survival sizing is the constraint that forbids that ruin; the survival-weighted, Kelly-flavoured discipline of the long-horizon view caps the bet first. So the two are not rivals to be averaged – they are lexically ordered, applied strictly in sequence. You impose the survival constraint first; only on what survives it do you then maximise expected value. Size so a single zero is survivable, set that size aside as paid-for insurance, and only then ask whether the convex shape of the remainder is worth holding. This is the argument's spine, and it is self-contained: it needs no faith about adoption, no institutional flows, no precedent – only the two undeniable premises that ruin is irreversible and that expectation rewards convexity, run in the one order that respects both. Get the order right and the apparent paradox – "how can a sober person hold something that might go to zero?" – simply dissolves.

You cannot study the far shore
while staring at your own feet.

Small fixed loss, large open gain

With the zero sized to be survivable, look at the shape of the bet that remains. It is not symmetric, and that is the entire point.

On the downside, your loss is bounded and fixed: you can lose at most your position, and you already sized that position so the loss is survivable. There is no margin call, no clawback, no liability beyond the stake – a flat floor you set yourself. On the upside, if the other branch happens, the payoff is not a tidy ten or twenty per cent; it is multi-fold, and it does not have a tidy ceiling. A small, fixed maximum loss against a large, open-ended gain – that convexity is not a footnote to the thesis, it is the thesis. It is the reason a sober person can hold something this volatile without being reckless: not because the downside is unreal, but because they paid for the downside in advance and what they bought with that payment was the option on the upside.

You pay a fixed, known fare.
What you buy is an uncapped ticket.

If that still sounds like a slogan, walk one concrete case all the way through – numbers chosen only to make the shape legible, not as any forecast. Say you have a hundred thousand dollars and you decide, after sizing for survival, to put five thousand of it – five per cent – into bitcoin, and to never add more. That five thousand is now the entire downside; the other ninety-five thousand was never on the table. Run the two branches the essay says the probability mass sits in. If bitcoin goes to zero, you are left with ninety-five thousand – down five per cent, a bad year, not a ruined life. That is the whole loss, and you knew its exact size the day you bought. Now the other branch: if bitcoin merely does over a decade what it would need to do to become a major reserve asset – call it an eightfold rise, well short of its past runs – your five thousand becomes forty thousand, and the portfolio is a hundred and thirty-five thousand. You risked five to make thirty-five. The loss was capped at the number you chose; the gain was a multiple of it. That is convexity in plain arithmetic: the most you can drop is one rung, and the upside climbs a staircase with no marked top. The calculator below lets you replace every one of those numbers with your own – including handing the "goes nowhere" outcome a third of the probability – and watch whether the shape survives your skepticism.

Fig. 2 – a model, not advice. The middle slider lets you give the “goes nowhere” outcome real weight – the very outcome the essay argues is unlikely. Hand it a third of the probability and watch what the convex tail still does. Every input is yours to argue with.

Sitting out is also a bet

Here is the half of the ledger almost everyone leaves blank. People treat "risk" as if it had only one direction – the risk of losing money you put in. There is another risk, and for an asymmetric bet it is often the larger one: the risk of not being in.

Run the logic forward. If you assign any non-trivial probability to bitcoin not going to zero – not certainty, just a real, honest, double-digit chance – then declining to hold it is not a neutral act. Not holding is itself a position. It is a bet that the failure branch wins. "Cash on the sidelines" feels like the absence of a wager, but against a binary it is a concentrated wager on the zero outcome – the very outcome you were so careful to insure against on the other side. You have, without noticing, taken the most aggressive position available: all of your chips on failure.

Put it in the language of expected value. A modest probability of a very large payoff can dominate a high probability of a small one; that is what convexity means arithmetically. A twenty-per-cent chance at a tenfold return is worth more, in expectation, than near-certainty of standing still – and standing still is what the sideline buys you in the success branch. The asymmetry that protects you on the downside works against you when you are absent: being underexposed to a convex bet is not the safe choice, it is a different bet with its own, quietly enormous, cost.

And there is the human ledger, too – regret, which markets are not supposed to care about but people cannot help caring about. The story you should rehearse is not only "I bought and it went to zero." It is also the other one: bitcoin keeps winning its argument, the asset and the companies built on it appreciate for years, and you watched the entire thing from outside, having concluded that the safe move was to do nothing. Both are real outcomes. An honest accounting weighs the cost of being wrong in each direction. Most people only ever weigh one.

"Sitting it out" is not the absence of a bet.
It is a maximal bet on the one outcome you feared.

MSTR: the same bet, with a lever

Strategy – the company once known as MicroStrategy – is the same binary wearing a second skin. It is a leveraged, convex expression of bitcoin: when the underlying moves, MSTR tends to move further, in both directions. That makes its upside larger and its risks sharper, and it deserves the same honesty in both columns.

So name its own risks plainly, because they are genuine and specific. The stock has often traded at a premium to the net value of the bitcoin it holds, and that premium can compress or vanish, punishing holders even if bitcoin itself holds up. The company carries debt and financing obligations; leverage that magnifies the gains magnifies the strains. It issues shares to buy more, which can dilute existing holders. And in a deep enough, long enough crash, the spectre at the back of every leveraged structure – forced selling – is not impossible. These are not smears; they are the real texture of the second layer, and anyone who waves them away is doing the mirror image of the sin this whole essay is against.

And be honest that the two downsides are not the same animal. With spot bitcoin held outright, the floor is genuinely flat: you can lose your stake and not one cent more, no margin call, no liability beyond what you put in – the loss side really is the band you chose. A leveraged structure does not give you that clean floor. Debt, financing terms and reflexivity mean a deep enough crash can force outcomes the holder never chose – selling at the bottom, a premium that collapses faster than bitcoin falls, a fixed obligation that comes due at the worst moment. Sizing mitigates that tail – held small, even a brutal outcome stays survivable for you personally – but it does not flatten the structure itself; the path can be uglier than a simple "I lost my stake," and pretending otherwise would be the very sin this essay is against.

So the discipline is the same shape but the worst case is heavier. Name the risks. Size the position so that the worst case – a premium collapse on top of a bitcoin drawdown, with the reflexive tail included – is survivable for you. Then, with that paid for, look at the shape. What remains is an even more convex claim on the same success branch: a way to express the binary with more torque, for those who have sized it so the extra torque cannot break them. The companion piece on the man who built that structure with his life – The Steward's Wager – argues that the leverage is deliberate, not accidental. The point here is only the geometry: a second asymmetric bet stacked on the first, with a larger and rougher downside that must be respected and a larger upside that is the reason anyone takes it.

More torque is more of both.
Size for the downside; the upside takes care of itself.

One behavioural footnote, held at arm's length

The case above is decision-theoretic and self-contained: it rests on the skew, the survival-first ordering, and the arithmetic of the worked example, and on nothing else. What follows is not part of it. It is one anecdote, offered deliberately as the weakest kind of evidence, and the argument loses nothing if you skip it.

Here it is, with its provenance worn on its sleeve. The figures were assembled and publicised by Strategy's own chief executive, Phong Le, reading the quarter's 13F filings – a promoter selectively presenting the holders of his own stock, which is close to the weakest possible source for a bullish claim. So treat the numbers as his and provisional. With that flagged: in the first quarter of 2026, as bitcoin fell roughly twenty-two per cent and MSTR roughly eighteen, Le's reading of the filings had thirteen of the top fifteen institutional holders adding rather than fleeing – and, more telling than the headline, the active, discretionary managers (the Capital Group desks) adding into the fall while the rest of the buying was passive index machinery.¹ 13Fs are public; anyone can re-pull the direction without his framing.

Hold it at exactly the weight it deserves, which is slight. Institutions are not oracles – they have been spectacularly wrong, in herds, and "the smart money is doing it" is one of the weakest arguments there is. This proves nothing. It is a single behavioural data point that happens to rhyme with the framing – people with the most to lose treating the drawdown as a discount rather than a verdict – and a rhyme is not a proof. The spine of this essay never needed it, and does not lean on it now. (The companion piece Volatility Is the Toll, Not the Trip makes the same observation from the price chart, with the same caution.)