Ten years is an awkward number to plan around in crypto, because very little in the space has existed that long, and the handful of things that have spent most of that decade being declared dead. Which is exactly why the exercise is worth doing on paper before doing it with money. Whenever I try to design a portfolio I could genuinely leave alone for ten years, the asset selection turns out to be the fast part. The slow part is designing around the person who has to hold it, because the thing that historically kills a decade long crypto position is a sell order placed by its own owner, usually somewhere around month eighteen of a bear market.
Bitcoin has drawn down roughly 75 to 85 percent from its peak in every major cycle, and so far it has recovered to new highs each time. Anyone who held through all of that did extremely well. Very few people held through all of that. So the real design problem is building something that survives its owner's worst weeks, and picking the assets is the smaller half of it. Four parts, all of which fit on one page.
Filter for survival, then stop
The default way people pick crypto assets is to hunt for the most upside. For a ten year hold I would invert that and rank on odds of still existing. A short list of filters, and anything that fails one is out no matter how good the story is. Has it survived at least two full cycles, including a bear where its narrative died and it drew down 80 percent or more? Is its liquidity spread across many venues, or propped up by one market maker on one exchange? Does it keep working if a single company, foundation, or founder loses interest, runs out of money, or gets arrested? Is there any usage left when the price is falling, meaning people using the thing for reasons other than its price going up?
Run those filters honestly and the output is boring. Mostly bitcoin, probably some ether, maybe one or two more positions if you can argue them through every filter with a straight face, and even then kept small. If you end up with twelve names, the filters were too loose. Three to five positions is plenty, and heavy concentration in the most survivable asset is the design working as intended. My working test is that if holding an asset comfortably requires following its news, it has already failed, because you will not follow anything's news for ten years.
The schedule matters more than the entry
Whatever you buy, buy it on a fixed schedule. Monthly is fine and the date does not matter. What matters is that the amount is sized to survive a recession and a bear market arriving together, because over ten years you should assume they will. If a 70 percent drawdown plus a pay cut would force you to stop contributing, the number is too big, so cut it until it is boring.
I would also resist holding cash aside for buying dips. It feels prudent and it almost never works, because the dip either does not come and you sit out years of gains, or it comes attached to news so ugly that you no longer want to buy. The schedule removes that decision, which is the point. The fewer decisions the plan asks of future you, the more likely future you is still executing it in year seven.
Rebalancing is optional and I would keep it minimal. One look per year, on a date you pick now, trimming anything that has grown past its target and moving the proceeds into the core. In many jurisdictions that trim is a taxable event, so check before promising yourself annual rebalancing you will resent. Skipping it entirely and simply capping what you ever put into the non core positions is a legitimate answer too.
Custody that can sit still for a decade
Ten years is a long time to trust an exchange. Even a well run one can change policy, exit your jurisdiction, freeze withdrawals during a crisis, or get acquired by someone you would not have chosen. None of that has to be likely in any single year to be likely across ten. So the core position belongs in self custody, and the setup should be deliberately unclever. A mainstream hardware wallet. A seed phrase backed up on paper or steel in two separate physical locations, neither of which is a photo, a cloud note, or a password manager. Do one full recovery test with a trivial amount before funding it seriously, because a backup you have never restored from is only a hope.
Two things people skip. First, write instructions someone you trust could follow if you die, sealed and stored with your other documents. A portfolio nobody can access has a ten year return of zero regardless of what the assets do. Second, resist clever setups. Hidden passphrases you might forget, multisig with pieces spread across three countries, encodings of the seed only you understand. Over this horizon, self inflicted lockouts plausibly destroy more value than theft does. A small operational balance on an exchange for contributions and the occasional trim is fine. The core stays cold and dull.
Write the drawdown rules before you need them
Treat a 70 percent drawdown as a scheduled event rather than a tail risk. It may not happen, but the plan should assume it will, and the rules for it have to be written now, while you are calm, because the version of you living through it will be looking for permission to sell. Mine would look something like this.
- Price alone is never a sell trigger. I do not sell a core position because it fell, at any depth, for any duration.
- Contributions continue on schedule unless my income changes. A drawdown is a price, and the schedule exists to buy prices.
- I sell only if a survival filter breaks, meaning a protocol failure, a broken security assumption, or usage collapsing for reasons unrelated to price. Even then I wait 30 days from the day I first decide to sell before placing the order.
- No leverage against the position, no lending it out for yield, no posting it as collateral. Most of the permanent losses I have watched people take in a drawdown ran through one of those three.
- During a drawdown I check the portfolio once a month at most. Watching it daily changes nothing except the odds that I break rule one.
The 30 day cooling period is the rule people push back on, and I would keep it anyway. A genuine protocol failure does not require you to be the fastest seller. If the thesis is truly broken it will still be broken in a month, and you will exit with a clear head. If the thesis was fine and you were scared, the delay just saved the whole plan. Cheap insurance either way.
All of this compresses onto a single page, and writing that page is the actual deliverable here. What you own and at what weights. The contribution amount and date. Where the keys live and who can recover them. The conditions under which you would sell. The drawdown rules, word for word. Then sign it, date it, and keep a physical copy with your seed backup instructions, because the moment you need to reread it will not be a moment you feel like digging through a spreadsheet.
None of this guarantees the assets perform over ten years, and nothing can. It just arranges things so that if they do perform, you are still holding them at the end, which sounds trivial and has historically been the harder half of the problem. The page takes about an hour to write, and the best time to write it is whenever the market feels dull, because once it feels urgent you are already too late to be objective.