In thirteen years of holding bitcoin, I have sold once. It was 2021. I sold roughly thirty percent of the position, and the proceeds became the house my family lives in.
Everything else I bought in 2013 is still where I put it.
People hear this and assume I have a strategy. A thesis, an allocation model, a rebalancing discipline — something with a name. I don’t. What I have is a series of accidents that worked out, one rule I’ve never had to use, and an intuition I cannot defend in writing. This essay is an honest inventory of all three, because I think the honest version is more useful than the strategy I could pretend to have.
The allocation nobody designed
Bitcoin is about a third of my family’s total wealth.
I want to be precise about how that happened, because the number sounds like a decision and it wasn’t. Nobody sat down at a kitchen table and concluded that thirty-something percent of everything we own should sit in a single volatile asset. No advisor would sign off on it. The conventional guidance for an asset like this is one to five percent — a satellite position, sized so that going to zero stings but doesn’t scar.
We are at six to thirty times the recommended dose. And we arrived there the only way anyone arrives there honestly: I bought a meaningful amount in 2013, the rest of our assets grew at the pace normal assets grow, and bitcoin grew at the pace bitcoin grows. The allocation wasn’t designed. It compounded into existence.
This distinction matters more than anything else in this essay. There are two kinds of concentrated positions: the ones people build, and the ones people inherit from their own past. If you are deciding today to put a third of your family’s wealth into bitcoin at today’s price, you and I are not in the same situation, and almost nothing about my position transfers to yours. My concentration is a consequence. Yours would be a bet.
What the rest looks like
The other two-thirds is spread across the boring things — real estate, an operating business, equities, cash. No single one of them dominates. If you drew our family balance sheet as a pie chart, bitcoin would be the one large slice and the rest would look like the diversified portfolio of a moderately careful person.
In my head, bitcoin sits in a specific slot: it is the family’s venture capital.
That framing does real work. Venture capital is the part of a portfolio that is allowed to behave badly. It can draw down eighty percent without triggering a family meeting. It can be illiquid, misunderstood, and impossible to explain to relatives. In exchange, it is the only slice with a credible claim to multiplying. Every family with assets has some version of this slot, even if it’s empty. Mine happens to be filled with the venture bet that won — and because it won, it outgrew the slot it was assigned.
I never promoted it out of that mental category, though, and I think this is one of the few things I’ve done deliberately right. Bitcoin’s job in our treasury is not to pay for groceries, fund retirement, or anchor the family’s sense of security. Its job is to be the asymmetric bet. The day I start treating it as the foundation of everything is the day a price crash stops being volatility and starts being a family emergency.
Why I sleep fine
The question I get from people who learn the allocation: how do you sleep with a third of everything in an asset that regularly halves?
Three honest answers.
The first is cost basis. I bought in 2013. The price I paid is, from today’s vantage point, a rounding error. When your entry is that low, a fifty percent drawdown doesn’t take you below your cost — it takes you back to a previous year’s high. I have watched this asset lose half its value more times than I can count, and not once was I underwater. Fear needs a foothold, and a 2013 cost basis doesn’t offer one.
I’m aware of what this means: my calm is not a character trait. It is a feature of my entry price. In early 2015 I watched bitcoin fall back to a hundred-something dollars — down more than eighty-five percent from the peak, the kind of drawdown that ended most people’s involvement with the asset permanently — and I wasn’t worried. I’d like to tell you that was conviction. Some of it was. But conviction is cheap when the position is still above water. The holders who bought the 2013 top and sat through 2015 underwater — those people earned their conviction. I rented mine at a discount.
The second is the eighty percent test. Here is the exercise I actually run, and the one I’d suggest to any family holding a concentrated position: assume the asset drops eighty percent tomorrow and never recovers. Walk through the next five years of your family’s life. What changes?
For us, the answer is: nothing operational. The house is paid for. Daily life runs on cash flow from the business, not from the portfolio. The kids’ education doesn’t depend on a price chart. An eighty percent crash would vaporize a spectacular amount of paper wealth and change nothing about what happens on Tuesday.
If your answer to the eighty percent test involves selling assets to cover obligations, your allocation is too large — not as a percentage, but in absolute relation to your life. The percentage was never the real variable. The real variable is whether the asset’s worst behavior can reach your family’s actual life.
The third answer is the firewall, which deserves its own section.
The firewall
Bitcoin does not participate in our living expenses. Not partially, not occasionally. There is no flow from the wallet to the checking account. Groceries, tuition, travel, the car that needs replacing — all of it runs on ordinary income from ordinary sources.
This is the closest thing I have to an actual treasury policy, and I’d take it over any allocation percentage. The percentage is a snapshot; the firewall is a structure. A family that holds five percent in bitcoin but taps it whenever cash runs short has a worse setup than a family at thirty percent that never touches it. The first family will be forced to sell at the bottom eventually — not because they’re undisciplined, but because expenses don’t schedule themselves around market cycles. The second family gets to make every sell decision voluntarily, on their own clock.
That’s what the firewall actually buys: it converts every future sale from a forced transaction into a chosen one. In thirteen years, I have never once sold because I needed to. Which brings me to the one time I sold because I wanted to.
The one sale
In 2021 I sold about thirty percent of the position and bought a house with it.
I won’t dress this up as a masterstroke of cycle timing. The truthful sequence is that the position had grown to where a meaningful fraction of it equaled a house — the full, permanent, paid-off kind — and the family needed one. So one asset became another asset. Bitcoin became the place my kids come home to.
Was it optimal? Almost certainly not. Bitcoin today is worth multiples of what I sold it for, and if I’d held those coins the math says we’d be wealthier. I know the math. I also notice that I have never once stood in the kitchen and felt the opportunity cost. The house refuses to be compared to a price chart. It converted out of the speculative column entirely, into the column of things that are simply ours — and I’ve come to think that’s what a family treasury is actually for. Not maximizing terminal wealth. Funding the irreversible purchases that define a family’s life, at the moments they matter, without debt and without selling anything under pressure.
One sale in thirteen years. It bought a home. I can defend that record to anyone, including myself at three in the morning, which is the audience that counts.
And I’ll note what the sale proved that holding never could: the position is real. Paper wealth that has never been spent has a hypothetical quality to it — you believe the number, but you’ve never tested whether the world honors it. Converting a slice of 2013 conviction into a physical building the family lives in closed that loop. The remaining position is not a theory. I’ve seen what it converts into.
The thirty percent rule
I have exactly one rule for the position: if it grows to be too large a share of everything, I trim it back to thirty percent.
I’ll tell you what’s solid about this rule and what’s vapor. Solid: the destination number. Thirty percent is the level at which the venture slot stays the venture slot — large enough to matter enormously if the thesis keeps playing out, small enough that the other two-thirds of the balance sheet can absorb a catastrophe. The trim, when it happens, would go where the house went: into the permanent column.
The vapor: the trigger. “Too large” is not a number. Is it forty percent? Fifty? I haven’t defined it, and the rule has never fired — the allocation has hovered around a third, with the 2021 sale doing one trim’s worth of work for house-shaped reasons. So I hold a rule that has never been tested, with a threshold I’ve never specified. I’m aware that this might mean I don’t have a rule at all. I might just have a story that makes the concentration feel managed.
But I keep the rule anyway, vague trigger and all, because of what it does to the kind of question I’ll face someday. Without a rule, “should I sell some bitcoin?” is an emotional question, asked at the worst possible moment — euphoria or panic, the only two times the question feels urgent. With a rule, it’s a mechanical question: what’s the allocation, what’s the threshold, execute or don’t. I’ve watched enough people make wealth decisions inside strong emotions to know that the entire job of a treasury policy is to make sure no decision is ever made there.
The intuition I can’t defend
Underneath everything I’ve described — the hold, the firewall, the unfired rule — sits a single load-bearing belief, and I owe you honesty about it: I think the price goes up. Over any stretch of years that matters, I simply believe it goes up. I believed it in 2013 when I bought because I liked new things. I believed it in 2015 at a hundred and fifty dollars. I believe it now.
I cannot fully defend this. I can recite the supply schedule and the adoption curve like everyone else, but I’d be reverse-engineering a justification for something that lives in my gut, not my spreadsheet. Thirteen years of being right have only made the belief stronger, which should worry me more than it does — that’s exactly what survivorship feels like from the inside. Every holder of every asset that eventually died believed the same thing I believe, right up until the end.
So here is how I hold the contradiction. My intuition says bitcoin keeps going up. My treasury is built for the possibility that my intuition is wrong. The firewall, the diversified two-thirds, the eighty percent test, the refusal to let bitcoin fund daily life — none of that is necessary if I’m right. All of it is necessary if I’m not. I think of it as the difference between believing and betting the family on the belief. I’m allowed the first. I’m not allowed the second.
My wife, for what it’s worth, doesn’t engage with any of this. She doesn’t ask about the allocation, doesn’t track the price, doesn’t weigh in on the rule. I’ve written before about what that means for inheritance — it’s a real problem and it’s mine to fix. For treasury purposes it means every decision described in this essay was made by a committee of one. That has been efficient. Efficient and fragile are not opposites.
What transfers
If you hold bitcoin and you’re responsible for a family, here is my attempt to separate what’s portable in all this from what’s merely mine.
Not portable: the thirty-three percent. My allocation is an artifact of a 2013 entry. Copying the percentage without the cost basis means copying my balance sheet without my margin of safety. If you’re allocating fresh capital today, the boring one-to-five percent guidance exists for people whose conviction hasn’t been rented at a discount.
Portable: the firewall. Whatever your allocation, build the wall between the position and your family’s operating life. No flows from the wallet to the checking account. Every sale a chosen one. This costs nothing and it is, I’ve come to believe, the actual difference between families that survive holding a volatile asset and families that get shaken out at the bottom.
Portable: the eighty percent test. Run it annually. If the worst-case drawdown reaches your real life — if it forces a sale, delays a tuition payment, changes a plan — the position is too big in the only sense that matters, whatever percentage it happens to be.
Portable: deciding the trim rule before you need it. Pick your number while the question is boring. The rule I keep with its undefined trigger is, I admit, half-finished homework. But even half-finished, it has done its job for thirteen years: it stands between my emotions and the sell button, and it means the question “should I sell?” has a procedure instead of a mood.
Portable, with a warning: spending some of it on something permanent. The house did more for my relationship with this position than any drawdown survived or milestone crossed. It proved the wealth converts. If the position has grown beyond what your family will plausibly need, consider letting a slice of it become something irreversible and real. Not because it’s optimal — it isn’t — but because a treasury that has never funded anything is just a scoreboard, and families don’t live on scoreboards.
One sale in thirteen years. A rule that has never fired. An intuition I can’t defend, inside a structure built to survive its failure. That is the complete, unimpressive truth of how a third of our family’s wealth is managed.
I notice it doesn’t sound like much of a strategy. I’ve stopped thinking that’s a weakness. Most of what passes for strategy in this asset class is a forecast wearing a suit. What I have instead is a set of arrangements that don’t require me to be right — and after thirteen years, that’s the only kind of strategy I trust.