I keep coming back to Harry Browne's permanent portfolio because it is one of the few allocation frameworks that starts from a question I can actually answer. Not which asset will win, but which economic regime we are in. Browne's whole idea was that you cannot know the regime in advance, so you hold one asset that thrives in each of the four he cared about, in equal weight, and you rebalance on a schedule. Stocks for prosperity, long bonds for deflation, gold for inflation, cash for the tight-money recession. Twenty five percent each. No forecasting required.
The reason it is worth revisiting now is that crypto did not exist when Browne wrote this, and the obvious lazy move is to bolt it onto the gold sleeve and call it a day. I think that is wrong, and it took me a while to work out why. So this is less a pitch for a specific allocation and more a way to reason about where a volatile, still-young asset actually belongs in a framework built to survive regimes rather than predict them.
What each sleeve is actually doing
The trap is to think of the four assets as four bets. They are not. Each one is insurance against a specific way the world can go wrong for the others. Stocks pay you in growth and get destroyed by inflation shocks and recessions. Long bonds pay you when growth and prices both fall, which is the deflation scenario nothing else likes. Gold is the odd one, because it is not really an inflation hedge in the way people say. It is a hedge against monetary disorder, against negative real rates, against the sense that the currency itself is the problem. Cash is the boring one that quietly wins when money gets expensive and everything with duration or leverage gets repriced down.
Hold that framing, because it is the only honest way to place crypto. The question is not whether Bitcoin goes up. The question is which regime it protects you in, and whether that protection overlaps something you already own.
Where crypto actually behaved like gold, and where it did not
Here is the uncomfortable part. Over the short window we have, crypto has mostly behaved like a high-beta growth asset, not like gold. When liquidity was easy and risk appetite was high, it ran with the most speculative corner of the stock market. When central banks tightened and cash suddenly paid a real yield, crypto sold off hard alongside long-duration tech, not the other way around. That is roughly the opposite of what you want from a gold replacement, because gold is supposed to be the thing that holds up while your growth sleeve is bleeding.
There have been stretches where the gold story showed up. Periods where confidence in a specific currency or banking system cracked and Bitcoin caught a bid as an exit rather than a bet. Those episodes are real and they are the whole reason the monetary-hedge argument is not nonsense. But they are episodic, and most of the time the correlation to risk assets dominates. So if you slot crypto into the gold sleeve, you are quietly cutting your true monetary hedge in half and replacing it with something that tends to fall at the same time your stocks fall. That defeats the purpose of the sleeve.
The cleaner reading is that crypto is its own regime bet. It pays you in the scenario where a specific fiat regime loses credibility and capital wants a neutral, non-sovereign rail, and it also pays you in the plain-old-liquidity-boom scenario. It is part debasement hedge, part growth call option. That combination does not map onto any of Browne's four slots. It deserves its own sleeve, and it deserves a small one.
A modified five-asset version
So instead of four equal quarters, I think about five sleeves with crypto carved out deliberately rather than smuggled into gold. A version I find defensible looks roughly like this. The exact numbers matter less than the logic, and you should scale the crypto weight to how much drawdown you can actually stomach.
- Stocks, roughly 25 percent. Broad, cheap, global. Your prosperity sleeve. Unchanged.
- Long-term bonds, roughly 20 percent. Your deflation and hard-recession hedge. Trimmed slightly to make room.
- Gold, roughly 20 percent. Kept meaningful and kept separate from crypto, because its job is to work when crypto is falling with everything else.
- Cash or short bills, roughly 25 percent. The tight-money sleeve, and the dry powder you rebalance from. I do not shrink this, because it is the thing that lets you buy the others after they crash.
- Crypto, roughly 10 percent. Mostly the largest, most liquid names. This is the debasement-plus-growth call. Sized so that a total loss is survivable and an eighty percent drawdown is annoying rather than portfolio-ending.
Notice what the crypto sleeve did to the rest. It came out of bonds and gold, not out of cash, because cash is doing structural work that crypto cannot replace. And gold stayed large on purpose. If you only take one thing from this, let it be that crypto and gold are not substitutes. They are two different insurance policies that happen to both get called precious by people who do not hold either.
The rebalancing rules are the whole point
The allocation is easy. The discipline is where people quietly blow themselves up, and crypto makes the discipline harder because the sleeve moves so much that it constantly wants to become a bigger share of your net worth than you agreed to. Two ways to rebalance, and I lean on the second.
- Calendar rebalance. Once or twice a year, sell whatever grew past its target and buy whatever shrank below it, back to the weights above. Simple, mechanical, hard to argue yourself out of. The cost is that you can let a sleeve drift far between checks.
- Band rebalance. Set tolerance bands and only act when a sleeve breaches them. For the stable sleeves I use tight bands, roughly a quarter of the target weight in either direction. For crypto I use a wider band, because a ten percent sleeve that swings to fifteen has not broken anything, but one that swings to twenty and I have not trimmed is now dictating the whole portfolio's mood.
The rule that actually saves you is the trim rule on the way up. When crypto rips and the sleeve doubles, you sell it back to target and move the proceeds into whatever got cheap, usually bonds or cash. This feels terrible every single time, because you are selling the thing that is working. That feeling is the mechanism. Browne's portfolio makes money precisely by forcing you to sell strength and buy weakness on a schedule you set while you were calm, so you do not have to make the call while you are euphoric or terrified.
The failure mode I have watched real people hit is the reverse. Crypto runs, they let it ride because trimming feels dumb, the sleeve balloons to a third of the book, and then the drawdown that always comes takes out years of gains from the rest of the portfolio. They did not have a crypto problem. They had a rebalancing problem, and the crypto sleeve just happened to be the loudest place it showed up.
How to actually run it
Keep it boring. Write the five target weights down. Write your bands down. Pick a rebalance cadence and a hard trim rule for the crypto sleeve, something like never let it sit above 1.5 times its target without cutting it back. Then mostly do nothing between triggers. The point of the whole structure is that you make the hard decisions once, in advance, and then execute them on autopilot when the regime you cannot predict finally shows up.
If crypto turns out over the next decade to behave more like gold and less like a leveraged growth bet, this framework absorbs that fine, because you can slowly widen the sleeve as the evidence comes in. And if it stays a high-beta debasement call, the small deliberate weighting means it can do its job in the good regime without owning your outcome in the bad one. Either way you are not forecasting. You are just making sure that whichever regime turns up, you already own the thing that was supposed to work in it.