I have held bitcoin since late 2013. My partner knows this. She knows the rough value. She knows the Ledger sits in a drawer and that the thing matters.
What she does not know is how to access any of it.
If I were hit by a bus tomorrow — the cliché everyone uses and nobody plans for — she would find a small black device she cannot unlock, and some words she does not know exist, hidden in places she has never been told about. The bitcoin might as well not exist.
I have known this for years. I have done nothing about it.
This essay is about why. Not about the tools, which exist and are good. Not about the services, which I’ll mention later. About the psychological architecture of a bitcoiner who can self-custody a seven-figure stash across multiple locations with no single point of failure, yet cannot bring himself to sit his partner down for a thirty-minute conversation.
Three reasons I never taught her
She’s not interested
I have tried, in the way you try when you already expect the answer. I’ve mentioned how the wallet works, how the seed phrase is the actual money, how the device is just a signing machine. I get the look. The look that says: this is your thing, not mine.
She’s not wrong. It is my thing. She didn’t buy bitcoin in 2013. She doesn’t read the mempool. She doesn’t care about block subsidy halvings. She trusts me to handle it the way I trust her to handle things I don’t understand and don’t want to learn.
The problem is that trust has an expiry date, and it is the day I stop being alive.
I’m afraid she’ll lose it
This is the one I don’t say out loud. I’ve spent years designing a custody setup where no single point of compromise can drain the wallet. The seed is split. The hardware is separated. The locations are chosen deliberately. It is, by my own evaluation, a good setup.
And then I imagine handing half of that setup to someone who has never used a command line, who does not distinguish between a wallet and an exchange, who would — with the best intentions — write the seed phrase in the Notes app and text it to her sister for safekeeping.
I am not mocking her. She is smarter than I am in a dozen domains I cannot touch. But operational security is a discipline, and she has no reason to have acquired it. The fear is not that she’s incapable. The fear is that competence in this particular domain takes years of accumulated paranoia, and I cannot transfer paranoia in a conversation.
So I transfer nothing.
I’m still young
This is the weakest reason and the one that has carried the most weight. I’m in my forties. I am healthy. The actuarial tables say I have decades. The urgency of inheritance planning is, statistically, low.
But the actuarial tables are averages. They describe populations, not persons. The bus does not check your life expectancy before it runs the red light.
Every bitcoiner who died without a plan also thought they had time. I don’t know their names because their families didn’t recover the coins and nobody wrote about it. The survivorship bias runs in the wrong direction: the inheritance failures are silent. You only hear the success stories, and there are almost none.
I have known this for years too. Still here. Still haven’t sat down with her.
The security-inheritance paradox
Here is the thing nobody talks about at bitcoin conferences. The better your self-custody, the worse your inheritance plan.
Think about what good security means: no single person, no single location, no single device holds enough information to move the funds. You distribute trust across geography, hardware, and time. You make it so that even if someone finds one piece — the Ledger, one shard of the seed, the PIN — they get nothing. The system is designed to resist exactly one thing: unauthorized access by anyone who is not you.
Your family is “anyone who is not you.”
Every precaution you take against a thief is a precaution you take against your widow. Every location you hide a seed backup is a location your children don’t know about. Every passphrase you memorize is a passphrase that dies when you do.
The industry has a word for when funds become permanently inaccessible: lost. We use the passive voice. Bitcoin “is lost.” As if the coins wandered off on their own. As if there wasn’t a person who locked them away and then failed to leave a key for the people who needed it most.
I don’t want my bitcoin to be lost. I especially don’t want it to be lost to the people I acquired it for in the first place.
Inheritance is not a training problem
My first instinct — the instinct I suspect most holders share — was to think of this as an education problem. If I could just teach my partner how bitcoin works, how wallets work, how seed phrases work, the problem would be solved.
It would not be solved. It would be relocated.
Teaching your partner to use a Ledger means your partner now needs to remember how to use a Ledger. Teaching them to recover from a seed means they need to understand BIP39, derivation paths, and which wallet software to use. Teaching them operational security means they need to maintain the same level of paranoia you maintain, indefinitely, starting from zero.
You have not solved the problem. You have created a second point of failure — a person who now holds critical knowledge but has no practice using it, no context for updating it, and no muscle memory for the operational discipline it requires. When the crisis comes — when you are dead and they are grieving and the estate attorney is asking about “digital assets” — they will be doing something they learned once, years ago, under no stress, and must now execute perfectly under maximum stress.
That is not a plan. That is a hope.
The reframe that changed my thinking: inheritance is a system design problem. The right system requires zero ongoing knowledge from anyone except the person who builds it. The partner, the children, the executor — they should need to follow a procedure, not understand a technology. Open envelope. Call this number. Present this document. The system does the rest.
I am not there yet. But I now know what “there” looks like.
Five principles I’m building toward
I haven’t implemented all of these. I’m writing them down because articulating the target matters more than pretending I’ve arrived. Some of these are active. Some are aspirational. I’ll be honest about which is which.
1. Don’t require ongoing learning
Anyone in the inheritance chain — partner, adult child, executor, attorney — should not need to remember anything beyond the location of a sealed envelope or a phone number. If the plan requires them to recall a PIN, update a firmware, or distinguish between a seed phrase and a passphrase, the plan has a training dependency. Training dependencies decay. People forget. Firmware changes. The wallet app you taught them to use gets discontinued.
The system must be legible to someone who has never touched a hardware wallet and never will.
I have not achieved this yet. My current setup requires knowledge only I hold.
2. Default-on access, default-off custody
The partner should not hold keys in day-to-day life. Keys in untrained hands are keys at risk. But the partner should know how to trigger the inheritance flow — a phone call, a letter, a visit to a specific professional. The distinction matters: they don’t need to hold the bitcoin. They need to start the process that releases it.
This is how every other asset works. Your spouse doesn’t need to know the vault combination at the bank. They need to know the bank exists and the estate attorney has the paperwork.
3. Trust plus multisig beats pure DIY
I have resisted this one for years. The cypherpunk in me says: not your keys, not your coins. Trust no one. Hold everything yourself.
The inheritance planner in me says: if “no one” includes my family after I die, then pure self-custody is a time bomb with a dead man’s switch that doesn’t exist.
A collaborative custody service — Casa, Unchained, Nunchuk, or whatever emerges next — holds one key in an M-of-N multisig. You hold the majority. They can’t move your funds alone. But when you die, your family contacts them, presents legal proof, and initiates a recovery that doesn’t require your family to understand elliptic curve cryptography.
You give up a sliver of sovereignty. You gain the ability to die without destroying your family’s wealth. I think that is a trade worth making. I haven’t made it yet, but I think it’s the right architecture.
4. Test the system once a year
A backup that has never been tested is not a backup. It is a wish.
Schedule an actual fire drill: once a year, walk through the entire inheritance procedure as if you died yesterday. Hand the sealed envelope to your partner. Watch them follow the instructions. See where they get stuck. Fix the instructions. Repeat.
If the sealed envelope says “call this number” and the number is disconnected, you want to find that out during the drill, not during the funeral.
I have never done a drill. That should tell you how far I still have to go.
5. Document outside the wallet
A plain-English operations manual for the executor, the attorney, and the partner. Stored physically — printed paper in a fireproof safe or a safe deposit box. Not on the wallet device, not in a cloud note, not in an app.
The document answers three questions: What exists? Where is it? Who do I call?
It does not contain seed phrases or private keys. It contains the information needed to start the process. The keys themselves are handled by the multisig architecture in principle 3, or by the sharded backup with a professional recovery path.
I have a version of this document in my head. I have never written it down. The irony is not lost on me.
Where I am now
Let me be specific, because honesty is cheaper than credibility repair.
I hold most of my bitcoin on a Ledger hardware wallet. The seed is split across multiple locations. No single location is sufficient for recovery. A small amount sits on Binance — the operational float I use for occasional transactions.
By the five principles I just articulated, my current setup fails on every count:
- Ongoing learning? Required. Only I know how to reassemble the pieces.
- Default-on access? No. My partner has no trigger mechanism.
- Professional cosigner? No. Pure DIY. Single-person dependency.
- Annual drill? Never done one.
- Written documentation? Exists only in my head.
I am the single point of failure in a system I designed to have no single point of failure. The irony is architectural, not just personal.
What I’m evaluating now: a migration from pure self-custody to a collaborative multisig arrangement. The services I’m looking at — Casa, Unchained, Nunchuk — each approach it differently. I haven’t chosen one. I am between Casa and Unchained because both offer an inheritance protocol that my partner could trigger without needing to understand the underlying cryptography. Nunchuk’s model leans more toward self-sovereignty, which appeals to me technically but may not serve the inheritance goal.
I’m also looking at the legal layer — whether a bitcoin-literate estate attorney should hold a key, whether a trust structure changes the custody architecture, whether my children (one of whom studies mathematics at university and could actually understand this) should be in the key hierarchy now or later.
These are open questions. I don’t have answers. I have the questions, and I have the principles, and I have the uncomfortable knowledge that every day I don’t act is a day the system remains broken in exactly the way I’ve spent four thousand words describing.
What comes next
I’ll publish the architecture I land on — the specific setup, services, and legal structure — once I’ve implemented it and tested it through at least one annual drill. I won’t recommend what I haven’t run.
If you’re reading this and recognizing yourself — the Ledger in the drawer, the seed in the hiding spot, the partner who knows you have bitcoin but couldn’t access it if you vanished — you don’t need me to tell you what to do. You already know. You’ve known for years.
The question is whether you’ll keep knowing and doing nothing, or whether you’ll treat this as what it is: a system design problem with a deadline you can’t predict, affecting people you can’t replace.
I’m writing this essay as much for myself as for you. I’ll let you know when I’ve solved it.